Remote Sensing

Remote Sensing is the science of learning about Earth's surface and atmosphere without touching it. Instruments on satellites, aircraft and the ground collect light, infrared radiation and microwaves that have interacted with the surface or air, and analysts turn those measurements into maps, images and numbers. The event rewards one skill above all: taking a measurement from a sensor, a satellite or a map, working out what physical process produced it, and stating a defensible answer with units.

The two divisions use the same toolkit on different problems, and in 2027 they carry different themes. Division B is Observable Earth: reading maps and images, measuring distance, area, elevation and slope, classifying land cover, and recognizing geohazards and Earth-surface processes such as glaciers and rivers. Division C is Climate Change: how sensors collect data, what a fixed list of climate satellites shows about clouds, gases, temperature, sea level and energy flux, and how a planet's energy balance sets its temperature and how fast it changes.

Event format

The binder is your most valuable preparation: both divisions allow any content in any form, so it can hold formulas, satellite and sensor tables, map-reading pages and worked examples. If the event runs lab stations, nothing may be removed from the binder during the event, so build it assuming you cannot rearrange it. Bring the protractors and rulers; map-distance, perimeter and slope questions are much faster with real tools. There are no separate note sheets: everything you want to consult has to be on the binder rings.

Strategy

Split the work by skill. The usual split is physics and calculations for one partner and imagery, maps and data for the other. In Division C, instrumentation and physics (about 25%) and energy balance (about 20%) are mostly algebra that a calculation-minded partner can own, while interpretation (about 35%) and climate processes (about 20%) reward the partner who knows the satellites. In Division B, interpretation is 65 to 75% of the points, so both partners must be fluent with maps, band combinations, land cover and the look of geohazards in each kind of image; the concept partner should also know which equations Division B may ask you to compute (only three). Both partners should know the whole instrument list, because data questions name an instrument and expect you to know what it measures.

Build the binder in layers.

1. Constants and equations: the wave relation, Wien, Stefan-Boltzmann, Kepler, radar range, ground sample distance, NDVI and the energy-balance formula, each with units and a note of which division may calculate it.
2. Sensor table: active or passive, wavelength region, what it measures, a typical use.
3. Satellite table: for each platform in your division's list, its instruments, orbit and what it is known for, plus (Division B) the Landsat band chart and band combinations.
4. Division B map page: scale conversions, area from pixel counts, slope from contours, display types (natural color, false color, thermal, radar, elevation, classified) with one labeled example each, and drainage-pattern sketches.
5. Division C climate page: gas sources and lifetimes, ENSO diagrams, the energy budget figure, and the three heating-rate models.
6. Worked examples: your own solved problem per formula.

Manage the clock. Scan the test, bank the quick recall items, then do calculations, then explanations. For "explain why" questions, write one sentence naming the mechanism and one tying it to the data; points go to quality, not length.

Always write units and a direction. Many lost points are a correct number with no unit, a slope computed with the wrong contour interval, or a percentage with the wrong base.

What the rules cover

The scope below follows the 2027 rules manual: Division B is titled Observable Earth (section 3 of the Division B rules) and Division C is titled Climate Change (sections 3.c to 3.f of the Division C rules). Neither division marks any item as State-only or National-only, so everything applies at every level. The percentages are the stated weights.

Division B: Observable Earth

Division C: Climate Change

Where each rules topic is taught

Use this map to jump from a rules item to the section that teaches it.

Light, radiation and the spectrum

Almost every remote sensing measurement is a measurement of electromagnetic (EM) radiation, which travels through vacuum at the speed of light. This chapter is the physics toolkit for the rest of the guide: what light is, how its wavelength, frequency and photon energy are linked, how hot objects emit it, and what the atmosphere and surfaces do to it before a sensor records it. Division C gives this area about a quarter of its points; Division B asks most of it conceptually.

Wave basics and photon energy

Light is a transverse wave of oscillating electric and magnetic fields that also behaves as a stream of particles, photons. The wave picture explains refraction, diffraction, interference (the basis of InSAR) and polarization; the particle picture explains why each photon carries a fixed energy, why detectors count photons, and the photoelectric and Compton effects. Tests ask both: "light is both a wave and a particle" is a standard correct option.

Useful conversions: 1 nm = 10⁻⁹ m, 1 µm = 10⁻⁶ m = 1,000 nm, 1 eV = 1.602 × 10⁻¹⁹ J. To convert wavelength to wavenumber, $\tilde{\nu}\,(\text{cm}^{-1}) = 10{,}000/\lambda\,(\mu\text{m})$, so 15 µm is 667 cm⁻¹ and 10 µm is 1,000 cm⁻¹. When light enters glass or water its speed and wavelength drop but its frequency stays the same, so the frequency (and photon energy) is the property that identifies the light.

The electromagnetic spectrum region by region

The electromagnetic spectrum runs from gamma rays through X-rays, ultraviolet, visible light, infrared and microwaves to radio. Boundaries are conventions and books differ by a few percent, so give the range the question implies and do not argue over the last digit. Use the table below for "which region is 2 µm in?", ordering questions and "which region is used for…?" questions.

ReferenceElectromagnetic spectrum reference· 19 rows

Ranges are approximate. Photon energy from $E = 1240/\lambda$ (nm). Regions are listed from shortest to longest wavelength, which is also highest to lowest frequency and energy.

19 rows

Region	Wavelength	Frequency	Photon energy	Remote sensing notes
Gamma rays	Below 0.01 nm (10 pm)	Above 3 × 10¹⁹ Hz	Above about 100 keV	Absorbed by the atmosphere; only gamma spectrometry from low aircraft (soil radioactivity)
Hard X-rays	About 0.01 to 0.1 nm	About 3 × 10¹⁸ to 3 × 10¹⁹ Hz	About 10 to 100 keV	Higher energy than soft X-rays; absorbed by the atmosphere
Soft X-rays	About 0.1 to 10 nm	About 3 × 10¹⁶ to 3 × 10¹⁸ Hz	About 0.1 to 10 keV	Absorbed high in the atmosphere; used for the Sun, not Earth's surface
Far and extreme UV, UV-C	10 to 280 nm (UV-C is 100 to 280 nm)	About 10¹⁵ to 3 × 10¹⁶ Hz	4.4 to 124 eV	UV-C is completely absorbed by O₂ and ozone
UV-B	280 to 315 nm	About 1.0 × 10¹⁵ Hz	3.9 to 4.4 eV	Mostly absorbed by stratospheric ozone; causes sunburn
UV-A	315 to 400 nm	About 0.8 × 10¹⁵ Hz	3.1 to 3.9 eV	Mostly reaches the ground; used with neighboring UV-B wavelengths for ozone and SO₂, and for aerosol retrievals (OMI, TROPOMI)
Violet	About 380 to 450 nm	About 670 to 790 THz	2.8 to 3.3 eV	Shortest visible wavelength, highest visible frequency and energy
Blue	About 450 to 495 nm	About 600 to 670 THz	2.5 to 2.8 eV	Most scattered by air (haze); penetrates clear water best with green
Green	About 495 to 570 nm	About 530 to 600 THz	2.2 to 2.5 eV	Vegetation reflects a small peak; the Sun's spectrum peaks here, near 500 nm (answer "green" if asked for a color)
Yellow	About 570 to 590 nm	About 510 to 530 THz	2.1 to 2.2 eV	—
Orange	About 590 to 620 nm	About 480 to 510 THz	2.0 to 2.1 eV	—
Red	About 620 to 700 (750) nm	About 400 to 480 THz	1.65 to 2.0 eV	Longest visible wavelength; chlorophyll absorbs strongly
Near infrared (NIR)	About 0.7 to 1.3 µm	About 230 to 430 THz	1.0 to 1.8 eV	Reflected sunlight; healthy leaves very bright; water very dark
Shortwave infrared (SWIR)	About 1.3 to 3 µm (some books 1.1 or 1.4 to 3)	100 to 230 THz	0.4 to 1.0 eV	Reflected sunlight; leaf and soil water, minerals, snow versus cloud, CO₂ and CH₄ bands
Midwave infrared (MWIR)	About 3 to 8 µm (window 3 to 5 µm)	38 to 100 THz	0.15 to 0.4 eV	Mix of reflected and emitted; fires and hot targets (about 4 µm)
Longwave or thermal infrared (LWIR, TIR)	About 8 to 15 µm (window 8 to 14 µm)	20 to 38 THz	0.08 to 0.15 eV	Emitted by Earth; surface and cloud-top temperature
Far infrared	About 15 µm to 1 mm	0.3 to 20 THz	1.2 meV to 0.08 eV	CO₂ 15 µm band and water-vapor rotation band; mostly opaque from the ground
Microwave	About 1 mm to 1 m	300 GHz to 300 MHz	1.2 µeV to 1.2 meV	Radar, passive microwave radiometers, altimeters; passes through cloud
Radio	Longer than about 1 m (to kilometers)	Below about 300 MHz	Below 1.2 µeV	Broadcasting and communication; ionosphere reflects some bands

The "optical" range of remote sensing is roughly 0.3 to 14 µm: reflective wavelengths (visible, NIR, SWIR, up to about 3 µm) carry reflected sunlight, and emissive or thermal wavelengths (beyond about 3 µm) carry radiation the surface emits because of its temperature. Passive land imagers mostly use the visible and reflective infrared; thermal sensors use 3 to 5 and 8 to 14 µm; microwaves serve radar and microwave radiometers. Gamma rays, X-rays and most UV are absorbed by the atmosphere, so they are useless for imaging Earth's surface from space. Infrared itself is subdivided as near, shortwave, midwave, longwave and far infrared, in that order of increasing wavelength.

Of the Sun's energy arriving at the top of the atmosphere, roughly half is infrared, about 40% visible and under 10% ultraviolet; radio and X-ray contributions are negligible. Earth's own emission falls mostly (about 90%) between about 4 and 40 µm, with a tail into the far infrared (sometimes called longwave radiation, as opposed to the Sun's shortwave radiation below about 4 µm).

Some animals sense wavelengths humans cannot. Bees and many insects and birds see ultraviolet (roughly 300 to 400 nm), and many flowers carry UV-absorbing "nectar guide" patterns that are invisible to us but obvious to a bee; UV photography reveals them. Pit vipers sense thermal infrared with heat-sensitive pit organs. These are the natural versions of extending vision beyond the visible band, which is what multispectral sensors do.

Blackbodies and thermal emission

Every object above absolute zero emits radiation. An idealized emitter that absorbs all incident radiation, at every wavelength and angle, and emits the maximum possible for its temperature is a blackbody: absorptivity 1, emissivity 1, albedo (reflectivity) 0. Its spectrum depends only on temperature, not on its material, size, density or the light falling on it. The best laboratory approximation is a cavity with a small hole; stars, including the Sun, are close to blackbodies; most natural surfaces (water, vegetation, soil) behave nearly as blackbodies in the thermal infrared.

Two laws describe blackbodies. Both are on the Division C list as calculations; in Division B, Wien's law may be calculated but Stefan-Boltzmann is qualitative only. Wien's displacement law says the wavelength of peak emission is inversely proportional to temperature: hotter objects peak at shorter wavelengths. The Stefan-Boltzmann law says the total power emitted per unit area grows with the fourth power of absolute temperature. Doubling a surface's absolute temperature halves its peak wavelength and multiplies its emitted power by 16.

Planck's law and the shape of the curves

Planck's law gives the full spectrum: the spectral radiance a blackbody emits at each wavelength for a given temperature. The curve rises steeply from zero at short wavelengths, peaks, and falls off slowly at long wavelengths. Planck curves for different temperatures never cross: a hotter blackbody emits more at every single wavelength. Wien's law is where the curve peaks (set its slope to zero) and the Stefan-Boltzmann law is the total area under it (integrate over all wavelengths and directions), so both are consequences of Planck's law. Planck's law is not linear in temperature; only its long-wavelength tail (the Rayleigh-Jeans approximation) is. Neither division lists Planck's law as a calculation; know how the curves shift with temperature, and treat the formula and worked example below as background for reading Planck curves.

The frequency form of Wien's law peaks at a different place: $f_{max} \approx 5.879\times10^{10}\ \text{Hz/K} \times T$. For 1,050 K that is $6.17\times10^{13}$ Hz. Because the curve per unit frequency and per unit wavelength are shaped differently, $c/f_{max}$ is not equal to $\lambda_{max}$. Some tests convert a peak frequency to a wavelength and then apply $b/T$; if you do that, say so.

Why classical physics failed. Classical theory gave every possible vibration mode of the radiation the same average energy, $kT$ (equipartition). Because the number of modes rises steeply with frequency, the predicted radiance grew without limit at short wavelengths, the "ultraviolet catastrophe". Planck's quantization requires a whole quantum $hf$ to excite a mode, so high-frequency modes are rarely excited (average energy $hf/(e^{hf/kT}-1)$, far below $kT$ when $hf \gg kT$), and the curve turns over.

Emissivity, Kirchhoff's law and brightness temperature

Emissivity $\varepsilon$ is the ratio of what a surface emits to what a blackbody at the same temperature would emit. It is dimensionless and runs from 0 to 1. A graybody has the same emissivity less than 1 at all wavelengths; a selective radiator has emissivity that varies with wavelength (quartz sand dips near 8 to 9.5 µm, which is how ASTER maps silicate rocks). Kirchhoff's law of thermal radiation says that at thermal equilibrium a surface's emissivity at a wavelength equals its absorptivity at that wavelength: good absorbers are good emitters. For an opaque surface, absorptivity + reflectivity = 1, so $\varepsilon = 1 - \rho$, and in general absorptivity + reflectivity + transmissivity = 1. A perfect reflector (a "white body") has $\rho = 1$, $\alpha = 0$, $\tau = 0$ and emits nothing.

A thermal sensor measures radiance, not temperature. The temperature a blackbody would need to produce the measured radiance is the brightness temperature (also called radiant temperature). The object's true thermometer temperature is its kinetic temperature. Because real surfaces have $\varepsilon < 1$, brightness temperature is lower than kinetic temperature; integrated over all wavelengths, $T_{rad} = \varepsilon^{1/4}\,T_{kin}$. Two objects at the same kinetic temperature can look different in a thermal image if their emissivities differ, which is why land surface temperature products need an emissivity map, and why low-emissivity bare soil carries the largest retrieval error.

The strict Stefan-Boltzmann law $M = \sigma T^4$ describes a perfect blackbody; a real (gray) body emits $M = \varepsilon\sigma T^4$ with $0 < \varepsilon < 1$. A true-or-false item saying the law "applies only to blackbodies" is aiming at the strict form.

Scattering and absorption in the air

When light meets a particle in the air, it can be absorbed (its energy converted to heat or used to excite a molecule), transmitted, or scattered (redirected). Attenuation (or extinction) is the total loss from the beam: extinction = scattering + absorption. Scattering is not a loss of energy, only of direction; absorption converts the energy into heat. A sunlit surface seen from space has crossed the atmosphere twice (down to the ground and back up to the sensor), so atmospheric effects are applied twice.

Scattering regimes: Rayleigh, Mie and nonselective

The outcome depends on the size of the particle compared with the wavelength, often written as the size parameter $x = 2\pi r/\lambda$. Particles much smaller than the wavelength, such as air molecules, produce Rayleigh scattering, whose strength goes as the inverse fourth power of wavelength and is roughly equal in the forward and backward directions. Particles about as large as the wavelength, such as haze, smoke, dust and pollen, produce Mie scattering, which depends only weakly on wavelength and scatters forward more strongly. Particles much larger than the wavelength, such as cloud droplets, fog and rain, produce nonselective (geometric) scattering that treats all visible wavelengths equally, so the scattered light is white.

The scattering phase function describes how scattered light is distributed by angle. Rayleigh scattering is nearly symmetric, sending about as much light forward as backward. Mie scattering from larger particles is strongly peaked in the forward direction, which is why the sky around the Sun looks bright and white on a hazy day and why lidar and radar, which look at backscatter, see only a small part of the scattered energy.

Ranking by wavelength dependence, most to least: Rayleigh > Mie > nonselective. The same raindrop that scatters visible light nonselectively is tiny compared with a 10 cm radar wave, so for weather radar rain is a Rayleigh scatterer, which is why its echo grows so steeply with drop size.

Everyday consequences that tests ask about:

• Blue sky, not violet: violet scatters even more, but the Sun emits less violet, some is absorbed high up, and the eye is less sensitive to it.
• Red sunsets: at low Sun the path through air is many times longer, so the direct beam loses its blue; volcanic or smoke aerosols can deepen the colors, while thick haze makes sunsets milky and the sky paler.
• Deeper blue sky at altitude: less air above means less multiple scattering and less whitening by aerosol.
• Diffuse light: scattered skylight arrives from all directions, so on an overcast day objects cast no sharp shadows.
• Polluted sky: aerosols add Mie scattering, which whitens the sky, lowers its polarization and reduces visibility.

Elastic scattering keeps the photon's wavelength (Rayleigh and Mie); inelastic scattering changes it because the photon exchanges energy with the molecule. Raman scattering shifts the wavelength by a molecular vibration or rotation (an inelastic counterpart of Rayleigh, used by Raman lidar to measure water vapor and nitrogen); Brillouin scattering shifts it slightly by interacting with sound waves in a medium; Compton scattering of X-rays by electrons lengthens the wavelength and is classic evidence that light behaves as particles. Fluorescence absorbs a photon and re-emits a longer-wavelength one (chlorophyll fluoresces near 685 and 740 nm; a laser fluorosensor excites it actively).

Fog and cloud droplets (about 5 to 50 µm) are much larger than visible wavelengths, so most texts call their scattering nonselective; but nonselective scattering is the large-particle limit of Mie theory, and many keys credit fog and haze to Mie scattering. If only Rayleigh and Mie are offered, choose Mie; if nonselective is offered for fog or clouds, choose nonselective.

Scattering in images: path radiance, haze and the adjacency effect

For a satellite, scattering is mostly noise. Light scattered into the sensor's view without ever touching the target adds path radiance, a bright veil that lowers contrast and blurs the image; this haze is worst in the blue and UV because of Rayleigh scattering, which is why many sensors avoid or correct the shortest wavelengths and why NIR and SWIR bands cut through haze. Light scattered from neighboring bright areas into a dark pixel's line of sight is the adjacency effect. Aerosols scatter (and sometimes absorb) light, reduce clarity and alter spectral signatures, which can cause misclassification; optical sensors like Landsat are strongly affected while radar is not. Atmospheric correction removes these effects, turning top-of-atmosphere radiance into surface reflectance.

Three numbers describe an aerosol layer's optical effect. Aerosol optical depth (AOD, $\tau$) is the dimensionless column extinction: below about 0.1 is clean air, above 1 is thick smoke or dust. The single-scattering albedo $\omega$ is scattering divided by total extinction: near 1 for sulfate and sea salt (they only scatter, so they cool), much lower for black carbon soot (it absorbs and warms the layer). The height of the layer matters too, because an absorbing layer above bright cloud warms the climate while the same layer over dark ocean cools it. Fine aerosols act mainly on sunlight through scattering and absorption, but large dust and volcanic-ash particles also absorb and emit thermal infrared, so aerosols, like clouds and greenhouse gases, change the planet's effective emissivity. Aerosol sources, lifetimes and forcing are covered in the Division C climate chapter.

Absorption bands and atmospheric windows

Gas molecules absorb only at particular wavelengths, set by their quantized energy levels. Electronic transitions absorb in the UV and visible (ozone, NO₂, SO₂), vibrational transitions in the infrared (H₂O, CO₂, CH₄, N₂O, O₃), and rotational transitions in the far infrared and microwave (H₂O). Molecules with a changing dipole can absorb infrared; the two most abundant gases, N₂ (78.08% of dry air) and O₂ (20.95%), are symmetric diatomic molecules with no infrared vibrational bands, which is why they are not greenhouse gases and why N₂ is hard to measure spectroscopically. Argon (0.93%) is a monatomic gas and absorbs nothing useful. CO₂ (about 0.042%, about 424 ppm in 2024) and water vapor (0 to about 4%, highly variable) are minor by volume but dominate infrared absorption. Water vapor absorbs the most infrared overall and is the dominant absorber of microwaves (lines at 22.2 and 183 GHz), with oxygen absorbing near 60 and 118 GHz.

Some thermal bands are almost opaque from the surface up. The 6 to 7 µm water-vapor band and the center of the 15 µm CO₂ band absorb so strongly that a satellite sees only emission from the middle and upper troposphere there; that is why GOES water-vapor images (6.2 to 7.3 µm) show moisture aloft, not the ground, and why sounders use the edges of the CO₂ band to probe different heights. Only the window channels (about 3.7 to 4 µm, 8 to 9.5 µm and 10 to 12 µm) see the surface.

ReferenceAbsorption bands of atmospheric gases· 19 rows

Wavenumber in cm⁻¹ = 10,000 / wavelength in µm. Bands used by the listed Division C instruments are noted.

19 rows

Gas	Band (wavelength)	Wavenumber	Region	Use or significance
O₃	Hartley band, about 200 to 310 nm	About 32,000 to 50,000 cm⁻¹	UV-C, UV-B	Strong; blocks all UV-C and most UV-B (the ozone shield)
O₃	Huggins bands, about 310 to 360 nm	About 28,000 to 32,000 cm⁻¹	UV-B, UV-A	Weaker; total ozone and profiles (OMI on Aura, TROPOMI)
O₃	Chappuis band, about 400 to 650 nm (weak tail to about 850 nm), peak near 600 nm	About 12,000 to 25,000 cm⁻¹	Visible	Weak, broad; it is visible absorption, not UV
O₃	9.6 µm	About 1,040 cm⁻¹	Thermal IR	Notch inside the 8 to 14 µm window
O₂	A-band 760 nm (B-band 687 nm)	About 13,100 cm⁻¹	Far-red	OCO-2 and TROPOMI use it to measure air column and cloud height
O₂	60 GHz and 118 GHz	—	Microwave	Microwave temperature sounding
H₂O	0.94, 1.14, 1.38, 1.87, 2.7 µm	About 10,600 to 3,700 cm⁻¹	NIR, SWIR	Gaps near 1.4 and 1.9 µm; Landsat OLI cirrus band sits at 1.37 µm
H₂O	6.3 µm (about 5 to 8 µm)	About 1,595 cm⁻¹	MWIR	GOES water-vapor channels; why 5 to 8 µm is opaque
H₂O	Rotation band, beyond about 18 to 20 µm	Below about 550 cm⁻¹	Far IR	Closes the far infrared
H₂O	22.2 GHz and 183 GHz	—	Microwave	Water-vapor sounding and altimeter wet-path correction
CO₂	1.6 and 2.06 µm	About 6,250 and 4,850 cm⁻¹	SWIR	OCO-2 retrieves column CO₂ here
CO₂	2.7 and 4.3 µm	About 3,700 and 2,350 cm⁻¹	SWIR, MWIR	4.3 µm is very strong
CO₂	15 µm	About 667 cm⁻¹	Thermal IR	The main greenhouse band; used for temperature sounding (CrIS)
CH₄	1.65 and 2.3 µm	About 6,050 and 4,350 cm⁻¹	SWIR	TROPOMI methane; leak detection
CH₄	3.3 and 7.7 µm	About 3,020 and 1,306 cm⁻¹	MWIR	Thermal-IR methane
N₂O	4.5, 7.8 and 17 µm	About 2,220, 1,285 and 590 cm⁻¹	MWIR, TIR	Greenhouse absorption near the window edge
CO	2.3 and 4.6 µm	About 4,300 and 2,140 cm⁻¹	SWIR, MWIR	TROPOMI (2.3 µm); fires and pollution
NO₂	About 400 to 500 nm	—	Visible	TROPOMI and OMI map urban and power-plant NO₂
SO₂	About 300 to 320 nm (also 7.3 and 8.6 µm)	—	UV, TIR	Volcanic and power-plant SO₂ (TROPOMI, OMI, ASTER thermal)

An atmospheric window is a wavelength range the atmosphere transmits well, so a sensor can see the surface. The main windows are the visible and near infrared (about 0.4 to 1.3 µm), SWIR pockets (about 1.5 to 1.8 and 2.0 to 2.4 µm), the midwave window (3 to 5 µm), the thermal window (8 to 14 µm, interrupted by ozone at 9.6 µm) and the microwave region (wavelengths longer than about 1 cm, where clouds are transparent). In the visible, absorption by gases is small and scattering dominates. Imagers that map the surface choose windows; sounders and gas sensors deliberately choose absorption bands. Thermal satellites use 10 to 12 µm rather than 6 to 7 µm because water vapor makes 5 to 8 µm opaque, and a band at 15 µm sees CO₂ high in the atmosphere instead of the ground. If brightness temperatures from a window band and an absorbing band are nearly equal, there is little absorber in the column. Comparing two thermal bands (a split window, about 11 and 12 µm) corrects sea and land surface temperatures for water vapor.

Introductory texts often summarize the atmosphere's transparency as three broad windows: the optical window (visible and near-infrared), the infrared windows (parts of the thermal infrared, chiefly 8 to 14 µm) and the radio window (microwave and radio, roughly 1 cm to 10 m or more). Everything between is blocked by gases, and the ionosphere reflects long radio waves.

Edges matter in multiple-choice items: the 8 µm end of the 8 to 14 µm window is still affected by water vapor, so the cleanest thermal window is about 10 to 12 µm. The 15 µm region is dominated by CO₂ but also lies on the wing of water vapor's rotation band, so keys sometimes list both gases as absorbers there. Earth emits mostly between about 4 and 40 µm with a peak near 10 µm; if a thermal-mapping question offers "1 to 10 µm" or "10 µm to 1 mm", choose the longer range.

Attenuation and the Beer-Lambert law

A beam passing through an absorbing or scattering medium loses the same fraction of its intensity in each equal step, so intensity decays exponentially. The attenuation (extinction) coefficient $k$ (also written $\mu$ or $\alpha$, in m⁻¹) says how quickly; the product $\tau = kx$ is the optical depth (dimensionless), and $T = I/I_0 = e^{-\tau}$ is the transmittance. Final intensity is inversely, exponentially related to the coefficient. Along a slanted path the optical depth grows as $\tau/\cos\theta$, where $\theta$ is the angle from vertical; for an active sensor the beam goes down and back, so the two-way transmittance is $e^{-2\tau}$.

Visibility follows the same physics: the Koschmieder relation gives visual range $V \approx 3.912/k_{ext}$ for a black target against the sky. In water, attenuation is strongly wavelength dependent: NIR is absorbed within centimeters, red within a few meters, and blue-green light (about 450 to 550 nm) penetrates deepest, tens of meters in clear ocean. That is why water looks blue, why underwater photos turn blue, why bathymetric lidar uses green light, and why ocean-color sensors work in blue and green bands.

In gas retrievals the law is often written with the absorber's amount: $I = I_0 e^{-k\rho L}$, with $k$ the mass absorption coefficient (m²/kg), $\rho$ the absorber density (kg/m³) and $L$ the path length (m). Inversion means solving for $\rho$ (the concentration) from the measured ratio $I/I_0$: $\rho = \ln(I_0/I)/(kL)$.

Refraction and the inverse-square law

Refraction is the bending of light as it crosses into a medium with a different speed of light. The refractive index $n = c/v$ measures how much the medium slows light: vacuum 1 exactly, air about 1.0003, ice 1.31, water 1.33, ordinary glass about 1.5, diamond 2.42. Refraction is described by Snell's law, $n_1 \sin\theta_1 = n_2 \sin\theta_2$, where the angles are measured from the surface normal. Light entering a denser medium (higher $n$) slows down and bends toward the normal; light entering a lower-$n$ medium speeds up and bends away. In remote sensing it matters for lenses, for light entering water (the bottom of a clear lake looks shallower than it is, a correction optical bathymetry must make) and for the slight bending of light through the atmosphere near the horizon. Division B treats Snell's law qualitatively: know which way the ray bends and why.

Related facts tests use:

• Total internal reflection: light going from high to low $n$ is trapped when the incidence angle exceeds the critical angle (48.6° for water to air, about 42° for glass to air). At incidence angles below the critical angle, it cannot occur. Total internal reflection inside snow and ice grains is one reason snow is so bright.
• Dispersion: $n$ is slightly larger for shorter wavelengths, so blue bends more than red. Refraction therefore decreases as wavelength increases, which splits white light in a prism (Newton's experiment) and makes rainbows.
• Atmospheric refraction and mirages: air's refractive index depends on density, so it falls with altitude and with higher temperature. Light bends through these gradients: a hot road under cooler air bends skylight upward into your eye, producing an inferior mirage that looks like water. The same gradients bend radar beams and lengthen GPS signal paths; GNSS radio occultation turns the measured bending into refractivity profiles, which give temperature and humidity.
• Seawater: its refractive index rises with salinity and falls as temperature rises, so warmer water bends a ray slightly less (a larger refraction angle).

The inverse-square law says irradiance from a point source falls as $1/d^{2}$: double the distance and the power per unit area drops to a quarter, because the same power spreads over four times the area. It explains why the solar irradiance at Mars is less than half of Earth's and why an active sensor's echo is weak. Again, Division B asks about it qualitatively; Division C can use it in energy-balance arithmetic.

Refraction in optics and in the sky. A lens works by refraction: its refractive index and curvature set the focal length. Because the index changes with wavelength (dispersion), different colors focus at slightly different points (chromatic aberration), and each glass surface reflects a few percent of the light, so sensor optics use lens combinations and anti-reflection coatings. The green flash seen just as the Sun sets is also a refraction effect: the atmosphere refracts (bends) shorter wavelengths slightly more, so the green image of the Sun's rim sets last, while blue and violet are mostly scattered out of the long path.

Reflection and polarization

When light reaches a surface it is reflected, absorbed or transmitted, and the three fractions add to 1. How it reflects depends on roughness compared with the wavelength. Specular reflection from a smooth surface sends light off in one direction, with the angle of reflection equal to the angle of incidence (a mirror, calm water, a road for radar). Diffuse reflection from a rough surface spreads light in many directions; an ideal diffuse surface whose radiance is the same from every viewing direction is Lambertian, and for it radiance $L = M/\pi$. Real surfaces are in between, described by the bidirectional reflectance distribution function (BRDF), which is why the same field looks brighter or darker depending on sun and view angles. A surface that is rough at visible wavelengths (sand) can be smooth for 20 cm radar waves; the usual rule (the Rayleigh roughness criterion) calls a surface smooth when its height variations are less than about $\lambda/(8\cos\theta)$, with $\theta$ the incidence angle.

Remote sensing names for special reflections:

• Sun glint (sunglint): specular reflection of sunlight off water straight into the sensor; it saturates pixels and hides what lies in and under the water, so ocean sensors tilt or mask it.
• Backscatter: energy scattered back toward the source, the quantity radar measures.
• Corner reflector: two or three perpendicular surfaces (building and street, ship hull and sea) bounce energy straight back toward the source, giving very bright radar returns (double bounce).
• Retroreflector: a prism array that returns light exactly along its incoming path; satellites such as Jason carry laser retroreflectors so ground lasers can measure their orbit.
• Albedo is the fraction of incoming solar radiation a surface reflects (0 to 1). Snow is bright because light meets thousands of air-ice interfaces between grains and is scattered back out before it can be absorbed; as grains grow with age or melt, there are fewer interfaces per centimeter and albedo drops, especially in the NIR. Clear, bubble-free ice has few interfaces and is much darker than white, bubbly ice or fresh snow.

Polarization describes the orientation of a light wave's electric field; only transverse waves can be polarized. Sunlight is unpolarized because it comes from countless independent emitters with random orientations. Light becomes polarized by reflection (fully so at Brewster's angle, about 53° from the normal for water, which is why polarized sunglasses cut horizontal glare), by scattering (Rayleigh-scattered skylight is most polarized about 90° from the Sun), by transmission through a polarizing filter or absorption (dichroic materials) and by refraction in birefringent crystals, which split a beam into two polarized rays. Polarization can be linear, circular or elliptical. Because aerosol and cloud particles of different size, shape and composition polarize light differently, polarimeters are used for aerosol studies: POLDER on PARASOL, and HARP2 and SPEXone on PACE. Lidar measures the depolarization of its echo: spherical droplets keep the laser's polarization, while irregular ice crystals, dust and ash depolarize it. Radar polarization channels (HH, HV, VH, VV) are covered under radar.

Radiometric quantities and units

Two radiometric words are often confused. Irradiance is power arriving per unit area of a surface, in W/m², with no direction attached. Radiance is power per unit area per unit solid angle, in W·m⁻²·sr⁻¹, so it describes what a sensor sees looking in one direction. A satellite measures radiance; a surface energy balance uses irradiance. Radiance does not change with distance along a ray in a clear medium (a wall looks equally bright from 1 m or 10 m, it just covers a smaller angle), while the irradiance from a small source falls as $1/d^2$. A steradian (sr) is the unit of solid angle: $\Omega = A/r^2$, and a full sphere is $4\pi$ sr.

Radiometers are not limited to one band: they exist for ultraviolet, visible, near-infrared, thermal infrared (the most common for temperature) and microwave wavelengths. Atmospheric emission matters mainly in the thermal infrared and microwave, where gases and clouds radiate at their own temperature; in the visible and near-infrared the atmosphere affects the signal by scattering and absorption, not by emission.

Sensors and scanning

A sensor is active if it supplies its own energy and measures what comes back. Radar, lidar, laser altimeters and synthetic aperture radar are active. A sensor is passive if it only measures energy that already exists, either reflected sunlight or thermal emission from the surface and atmosphere. Optical cameras, multispectral imagers, infrared radiometers and most climate-monitoring radiometers are passive. The distinction is not about whether sunlight is involved: a camera photographing sunlit land is passive even though the light comes from the Sun. Mission details for every satellite named below are in the Satellites and instruments reference.

Every remote sensing system has the same chain of elements: (1) an energy source or illumination (the Sun, Earth's own thermal emission, or the sensor's own pulse); (2) radiation and the atmosphere, crossed on the way down and again on the way up; (3) interaction with the target (reflection, absorption, emission, scattering); (4) recording by the sensor; (5) transmission, reception and processing of the data, usually through a ground station; (6) interpretation and analysis; and (7) application to a real problem by the user. Questions that ask what limits a system usually point to one link in this chain.

Active and passive sensors classified

Tests love sorting lists. Use the energy source, not the wavelength or platform, to decide.

A passive visible sensor cannot collect useful data at night because there is no sunlight to reflect, and it cannot see through cloud. A passive thermal sensor works day and night because the surface emits all the time, but clouds are opaque in the thermal infrared, so it measures cloud-top temperature instead of the surface under cloud. Passive microwave sensors see through most cloud but need huge antennas for fine detail, so their pixels are tens of kilometers. Active sensors control the wavelength, polarization and angle of their illumination, but they draw more power, cost more and often have shorter lives.

Detectors

A detector turns photons into an electrical signal that is digitized into digital numbers (DN). The choice of detector material sets which wavelengths a sensor can see.

Infrared detectors are cooled because a warm detector and its housing emit their own thermal radiation and dark current, which swamps a weak signal. A detector's quantum efficiency is the fraction of photons that produce an electron (external quantum efficiency counts all incident photons; internal counts only absorbed photons). The size and number of detector elements in an array set the spatial resolution and the number of pixels per line.

Thermal-infrared detectors such as mercury cadmium telluride must be cooled, typically to about 60 to 100 K, or their own heat swamps the signal. Early missions used stored cryogens (solid or liquid coolant) that boiled away, so the coolant supply set the mission's thermal lifetime; modern instruments use mechanical cryocoolers or passive radiators, but cryocooler wear or degradation can still end or limit a thermal channel (ASTER's SWIR bands failed in 2008 when their detector temperature could no longer be held).

An analog image is continuous, like a photograph on film, where tone varies smoothly and resolution is limited by film grain. A digital image is a grid of discrete pixels, each holding a number (DN) quantized to a fixed number of levels. Scanning a photo turns analog into digital; satellite sensors record digital images directly.

Radar imaging, interferometry and elevation models

RADAR stands for RAdio Detection And Ranging. Modern radars use microwaves (centimeter to millimeter wavelengths). They work day and night and see through cloud, smoke, and all but heavy rain, because the wavelength is far longer than cloud droplets. Radar bands carry letter names from World War II; learn the order and one example for each.

ReferenceMicrowave and radar bands· 9 rows

IEEE letter bands. Wavelength $\lambda = c/f$. Longer wavelengths penetrate vegetation, dry soil and rain better; shorter ones see smaller particles and finer surface roughness but are attenuated more by rain.

9 rows

Band	Frequency	Wavelength	Example instruments and uses
P	About 0.3 to 1 GHz	About 30 to 100 cm	Biomass radar (ESA BIOMASS); deepest canopy and soil penetration
L	1 to 2 GHz	15 to 30 cm	ALOS PALSAR, NISAR; SMAP radiometer at 1.4 GHz for soil moisture; GPS signals; sees through forest canopy and dry sand
S	2 to 4 GHz	7.5 to 15 cm	NEXRAD weather radar (about 10 cm, least rain attenuation); NISAR
C	4 to 8 GHz	3.75 to 7.5 cm	Sentinel-1 (5.405 GHz, 5.55 cm), RADARSAT, SRTM main radar (5.3 GHz), Jason second altimeter frequency, ASCAT scatterometer
X	8 to 12 GHz	2.5 to 3.75 cm	TerraSAR-X, SRTM second radar (X-SAR), short-range weather radars; attenuated by heavy rain
Ku	12 to 18 GHz	1.67 to 2.5 cm	Jason and TOPEX main altimeter (13.6 GHz), GPM DPR KuPR (13.6 GHz), QuikSCAT SeaWinds
K	18 to 27 GHz	1.1 to 1.67 cm	Includes the 22.2 GHz water-vapor line; GRACE inter-satellite ranging (K and Ka)
Ka	27 to 40 GHz	0.75 to 1.1 cm	GPM DPR KaPR (35.5 GHz), SWOT KaRIn; light rain and snow
W	75 to 110 GHz	2.7 to 4 mm	CloudSat Cloud Profiling Radar (94 GHz); cloud droplets

The name letters do not follow wavelength order, so memorize "P L S C X Ku K Ka (V) W" from longest to shortest.

Side-looking geometry is a strength, not only a necessity: the oblique illumination creates shading and radar shadow that exaggerate relief, so side-looking radar images bring out faults, folds, drainage and other terrain structure for geologists, and their texture helps map forest and land use.

Radar geometry, resolution and distortions

An imaging radar looks sideways (side-looking airborne radar, SLAR), because a nadir-looking radar could not tell apart two points at equal distances on either side of the track. Terms: range is the across-track direction (perpendicular to flight); azimuth is the along-track direction (parallel to flight); slant range is the line-of-sight distance; ground range is the horizontal distance; near range is the side closest to the track and far range the farthest. The look angle is measured at the radar from vertical; the incidence angle is measured at the ground between the beam and the local vertical (or normal), and the depression angle is measured from horizontal (depression = 90° − look angle on a flat Earth).

Range resolution depends on pulse length: two targets are separated only if their echoes do not overlap, so they must be at least half a pulse length apart in slant range. Short pulses give fine resolution but carry little energy; real radars send a long pulse whose frequency sweeps across a bandwidth $B$ (a chirp) and then compress it in processing (pulse compression), giving resolution $c/2B$. Azimuth resolution of a real-aperture radar is the beam's footprint width, $\lambda R/L$, which grows with distance and shrinks with a longer antenna. A synthetic aperture radar (SAR) uses the platform's motion to combine many echoes from one spot coherently, imitating an antenna hundreds of meters to kilometers long, so it reaches meter-to-tens-of-meters detail from orbit, day or night and through cloud. Remarkably, a SAR's azimuth resolution is about half its real antenna length, independent of range and wavelength. Seasat (1978) was the first civilian Earth-observing satellite to carry an imaging SAR.

Side-looking geometry distorts mountains, and the three distortions are frequently tested:

• Foreshortening: a slope facing the radar appears compressed and bright, because its top is only a little farther from the radar than its base. Worst in mountains and at small incidence angles.
• Layover: an extreme case. When the slope facing the radar is steeper than the incidence angle, the radar pulse reaches the peak before the base, so the top is imaged closer to the radar than the bottom and the mountain appears to lean toward the sensor.
• Radar shadow: a slope facing away that is steeper than the depression angle receives no energy and appears black. Shadow grows at large incidence angles (far range); layover is worse at small incidence angles (near range).
• Slant-range scale distortion: in an image displayed in slant range, near-range features are compressed relative to far-range ones; converting to ground range corrects it.

Speckle is the grainy "salt and pepper" texture of every radar image. It comes from coherent interference of echoes from many small scatterers within one pixel, not from the ground itself. It is reduced by multi-look processing (averaging several independent looks) and by spatial filtering, both at the cost of resolution.

Slant range versus ground range. A side-looking radar measures slant range $R_s$, the straight-line distance from the antenna to the target. On flat ground the horizontal ground range is $G = \sqrt{R_s^{2} - H^{2}}$ (equivalently $G = H\tan\theta$, with $H$ the radar's height and $\theta$ the look angle from vertical). In a slant-range image, features in the near range are compressed relative to the far range (slant-range scale distortion); resampling each pixel to ground range removes the compression. Example: $H = 700$ km and $R_s = 800$ km give $G = \sqrt{800^2 - 700^2} = 387$ km.

Clutter and jamming. Clutter is unwanted echo of the radar's own pulse from things you are not trying to measure (ground, sea waves, buildings, insects in a weather radar). Jamming is interference from an outside transmitter on the radar's frequency, deliberate or accidental, that masks the real echoes.

Backscatter: roughness, moisture and polarization

An imaging radar records backscatter, the fraction of its pulse returned toward the antenna, usually expressed as the backscatter coefficient $\sigma^0$ in decibels. Smooth surfaces such as calm water and runways act like mirrors and send the pulse away, so they look dark. Rough surfaces (forest canopy, plowed fields, rough seas) scatter some energy back and look medium-bright. Buildings and steep slopes facing the radar form corner reflectors and look very bright. Wet soil returns more than dry soil because water raises the surface's dielectric constant: dry soil has a dielectric constant of about 3 to 5, wet soil 20 or more and liquid water about 80. The same physics lets passive microwave radiometers map soil moisture (wet soil has lower microwave emissivity and looks colder) and tell sea ice from open water.

Each surface sends radar energy back by a dominant mechanism, and each mechanism is brightest in particular polarization channels:

A radar transmits and receives in horizontal (H) or vertical (V) polarization. Like-polarized (co-pol) channels are HH and VV; cross-polarized channels are HV and VH. A single-pol system records one channel (HH or VV), dual-pol records one transmit with both receives (for example VV + VH, Sentinel-1's standard mode over land), and quad-pol (fully polarimetric) records all four. VV is most sensitive to surface roughness, so it maps ocean wind and soil roughness; HH penetrates canopies better and highlights double bounce (flooded forest, mangroves); VH and HV come mostly from multiple scattering in vegetation, which depolarizes the wave, so cross-pol is bright over forest and nearly black over smooth water. Longer wavelengths penetrate farther: L-band passes through forest canopy to the trunks and ground and through meters of hyper-dry sand (Shuttle radar revealed buried river channels in the Sahara), while X-band returns mostly from the top of the canopy. Clear-cuts look darker than the forest around them because they lose the canopy's volume scattering; texture is the main clue for forest type in radar images.

Interferometry and elevation models

Interferometric SAR (InSAR) compares the phase of two SAR images. Phase differences come from topography (if the two views are from slightly different positions, as in SRTM) or from ground motion between two dates. An InSAR change map shows colored fringes; each full color cycle is a fixed amount of motion toward or away from the satellite (half the radar wavelength, about 2.8 cm for Sentinel-1). Tightly packed fringes mean a lot of motion, for example a swelling volcano, a fault after an earthquake or a city sinking as groundwater is pumped. Division B tests InSAR at this interpretive level only. Radargrammetry instead uses two images of the same area from different incidence angles, like a stereo photo pair.

A digital elevation model (DEM) is a grid of ground heights. SRTM flew a radar interferometer on the Space Shuttle in February 2000 and produced a near-global DEM; ASTER stereo images produced another. Subtracting an older DEM from a newer one shows where material was removed (a landslide scar, a quarry, a retreating glacier) and where it piled up. A DEM that includes treetops and buildings is a digital surface model (DSM); one with them removed is a digital terrain model (DTM). Lidar is especially good at both because it records multiple returns.

Radar altimetry and scatterometry

A radar altimeter points straight down (nadir) and times the echo. The surface height above a reference ellipsoid is the satellite's orbital height (known to centimeters from DORIS, GPS and laser tracking) minus the measured range and minus corrections. Over the ocean the mean surface follows the geoid, the equipotential surface of gravity that approximates mean sea level, plus currents and tides; altimetry measures sea surface height, wave height and wind, and over land and ice it measures ice-sheet and lake levels and subsidence. It cannot measure sea surface temperature.

Wind speed comes from the strength of the nadir echo: wind roughens the sea with short waves, which scatter the pulse away from the vertical, so the backscatter coefficient σ⁰ falls as wind speed rises. Early altimeters (Skylab, GEOS-3 and Seasat, 1973 to 1978) first showed this, and wind is retrieved with empirical relations of the form $\sigma^0 = A\,U^{-B}$ (σ⁰ in linear units, $U$ the wind speed, $A$ and $B$ fitted constants; side-looking scatterometers use a similar power law that increases with wind).

Every correction tests ask about:

• Ionospheric delay: free electrons slow the signal by an amount proportional to $1/f^{2}$, so the range looks too long. Measuring at two frequencies (Ku 13.6 GHz and C 5.3 GHz on TOPEX and Jason) lets the delay be computed and removed.
• Wet troposphere: water vapor also slows the signal. A passive microwave radiometer on board (Jason's AMR near 18.7, 23.8 and 34 GHz; TOPEX used 18, 21 and 37 GHz, with 21 GHz as the water-vapor channel) measures it.
• Dry troposphere (from surface pressure models), sea-state bias, tides and the inverse barometer effect are removed with models.

The returned waveform carries three signals: the timing of its leading edge gives the range to mean sea level, the slope of the leading edge gives significant wave height (rough seas smear the echo), and the strength of the return gives wind speed (a smooth sea is brighter at nadir). Altimeters have a small field of view and see only the area directly beneath them; over land, steep slopes and rough terrain degrade accuracy. Jason-class altimeters measure global sea level to a few centimeters per measurement; averaging many measurements gives rise rates of millimeters per year. Laser altimeters (ICESat-2) do the same with light, which suits ice and vegetation but not cloudy skies.

A scatterometer is a radar that measures backscatter (not range) from the sea surface at several angles. Because wind-driven ripples set the backscatter, it retrieves ocean surface wind speed and direction (QuikSCAT's SeaWinds, ASCAT). Long-pulse scatterometers view large areas at once, trading resolution for coverage.

The altimeter's footprint is the patch of surface its pulse illuminates. For pulse-limited ocean altimeters such as Jason it is a few kilometers across (about 2 to 10 km, larger in high seas), much smaller than the beam-limited footprint the antenna alone would give; SAR-mode altimeters (CryoSat-2, Sentinel-6) sharpen it to about 300 m along track.

Altimeters save transmitter power with pulse compression: instead of a very short, high-power pulse they send a longer, lower-power frequency-swept pulse (a chirp) and compress the echo in processing, which gives the range resolution of a short pulse; on-board waveform tracking and averaging of many echoes further reduce the power needed. QuikSCAT (1999 to 2009), a Ku-band scatterometer that mapped ocean wind speed and direction, flew in a polar sun-synchronous orbit (about 800 km) with an 1,800 km swath, because global daily coverage of the oceans needs a near-polar orbit.

Weather, precipitation and Doppler radar

A precipitation radar sends pulses into clouds and measures the echo power from raindrops, snow and hail at each range, giving a 3D picture of precipitation. Echo strength is expressed as the reflectivity factor $Z$, which for small drops grows as the sixth power of drop diameter, so a few large drops dominate. It is reported in dBZ ($10\log_{10}Z$): roughly 20 dBZ is light rain, about 40 dBZ heavy rain, and above 50 to 55 dBZ suggests hail. Rain rate is estimated from a Z-R relationship (a common one is $Z = 200R^{1.6}$, with $R$ in mm/h). Liquid water returns about five times as much power as ice of the same size, so melting snow produces a bright band; hail's large echoes can make rain look heavier than it is. Radar cannot measure rain temperature.

NEXRAD is the U.S. network of about 160 WSR-88D ground Doppler radars (S band, about 10 cm wavelength, so heavy rain barely attenuates the beam). It scans a series of elevation angles every few minutes. Base reflectivity is the reflectivity from a single tilt, usually the lowest (0.5°), shown out to about 230 km (124 nautical miles), with a long-range version out to about 460 km (248 nautical miles). Composite reflectivity is the maximum reflectivity found at any elevation angle above each point, out to about 230 km; it shows the strongest cores in a storm, including hail aloft that base reflectivity can miss. Base velocity gives the radial (toward or away) speed of targets, out to about 230 km. Because the beam rises with distance and Earth curves away, distant radar echoes come from high in the storm and low clouds far away are missed.

Wavelength choice is a trade-off. Short wavelengths (Ka, W) detect small cloud and drizzle particles (CloudSat's 94 GHz radar, nadir-pointing) but are strongly attenuated by heavy rain; long wavelengths (S band, as in the US NEXRAD network) pass through heavy rain with little attenuation. Attenuation (loss of signal through the rain itself) is worst for X, Ku and Ka bands. GPM's Dual-frequency Precipitation Radar (Ku 13.6 GHz and Ka 35.5 GHz) compares the two frequencies, which respond differently to large and small drops, to estimate drop size distribution and to separate rain from snow.

Doppler radar also measures the frequency shift of the echo, which gives the speed of raindrops toward or away from the radar (the radial velocity); side-by-side inbound and outbound velocities (a velocity couplet) reveal rotation in a storm, while the hook echo is a reflectivity signature. A shift to higher frequency means motion toward the radar; lower frequency means motion away. One Doppler radar measures only the radial component, so two or more radars are needed to map the full horizontal wind. Dual-polarization radar transmits and receives both horizontal and vertical pulses, so it can tell drop shape and precipitation type (rain, hail, snow, even insects and smoke). Airborne Doppler radars on hurricane hunter aircraft map storm winds, and satellite radars are not blocked by terrain.

Received power falls very steeply with distance because the pulse spreads out on the way to the target and the echo spreads out again on the way back: the radar equation gives $P_r = P_t G^2\lambda^2\sigma/[(4\pi)^3R^4]$, where $G$ is antenna gain and $\sigma$ the target's radar cross section. Doubling the range cuts the echo to 1/16. The wave's speed does not appear.

Lidar

Lidar (LIght Detection And Ranging) times laser pulses instead of radio pulses, so it maps heights and cloud or aerosol layers with fine vertical detail but cannot see through thick cloud. It is the laser counterpart of radar. A system has a laser, a scanner, a receiver telescope and detector, and, on aircraft, a GPS receiver and an inertial measurement unit (IMU, which measures roll, pitch and yaw) to locate every return. Most lidars fire many thousands of pulses per second (150 kHz = 150,000 pulses per second), building a point cloud of x, y, z coordinates, each with an intensity (return strength, which depends on surface reflectance but also on range, angle and roughness, so it is hard to calibrate).

A single pulse over forest returns several echoes: the first return from the canopy top, intermediate returns from branches and the last return from the ground, so tree height is the first-return elevation minus the last-return elevation. Bathymetric lidar compares the green echo from the water surface with the echo from the bottom; it works best in shallow, clear water and fails in deep or turbid water. Depolarization of the echo tells particle shape: spherical liquid droplets do not depolarize, irregular ice crystals, dust and volcanic ash do. Lidar often flies at night because sunlight adds background noise to the faint returns. Lidar's short wavelength gives much finer resolution and less noise than radar and maps canopy height and terrain under trees, but clouds, fog and heavy rain block it and it needs a clear line of sight; radar is better through weather and for large areas and storms. Photogrammetry (stereo photos) is cheaper but cannot see the ground under a canopy.

More lidar details tests ask. Differential absorption lidar (DIAL) fires two wavelengths, one on a gas's absorption line and one just off it, and uses the ratio of their returns to get that gas's concentration along the path; because both pulses share the same optics and atmosphere, most instrument and aerosol effects cancel. Raman lidar detects light scattered at a shifted wavelength by a specific molecule (water vapor, nitrogen); the Raman return is about a thousand times weaker than the elastic return, so it needs narrowband filters or a spectrometer to block the elastic line and works best at night. Lidar lasers include solid-state Nd:YAG (1,064 nm and doubled 532 nm) and gas excimer lasers for the ultraviolet (XeCl at 308 nm, XeF at 351 nm). For airborne mapping, the number of points on a target is its area times the point density (times the returns per pulse if density is quoted per pulse): a tree crown 3 m in radius at 30 points per m² gets $\pi(3)^2 \times 30 \approx 850$ points. Beyond elevation and canopy height, lidar is used for obstacle detection and collision avoidance on self-driving cars and drones, and for wind (Doppler lidar).

Micropulse lidar (MPL, used in NASA's MPLNET ground network) fires very low-energy pulses (microjoules) thousands of times per second, so it is eye-safe and can run unattended around the clock to profile clouds and aerosols.

Space gravimetry

Space-based gravimetry (Division C, GRACE and GRACE-FO) does not sense light at all. Two identical satellites fly in the same low polar orbit about 220 km apart and measure changes in their separation to within about 10 micrometers (micron level) with a K- and Ka-band microwave ranging link (GRACE-FO adds a laser ranging interferometer); accelerometers remove non-gravitational forces such as drag. As the pair approaches a region of extra mass (an ice sheet, a full aquifer), the leading satellite is pulled forward first and speeds up, so the gap widens; then the trailing one is pulled and catches up; past the mass, the leader is slowed. Over a region that has lost mass (a depleted aquifer), the pull is weaker and the leader accelerates less. Monthly maps of these changes show mass moving: ice-sheet loss, groundwater depletion (California's Central Valley, northern India), ocean mass gain and seasonal water storage. GRACE measures mass, so it misses thermal expansion of seawater; how this feeds sea-level budgets is covered in the climate chapter. The underlying principle is Newton's law of gravitation, $F = Gm_1m_2/r^2$.

Gravity is measured as an acceleration, in m/s² (standard surface value about 9.81 m/s²). Geophysicists also use the Gal (after Galileo): 1 Gal = 1 cm/s² = 0.01 m/s², so 1 milligal (mGal) = 10⁻⁵ m/s², about one-millionth of Earth's surface gravity. GRACE's month-to-month gravity changes are tiny (on the order of microgals, µGal), so monthly maps are usually converted to equivalent water thickness in centimeters.

Three tracking modes exist. High-low satellite-to-satellite tracking: high GPS satellites track a single low satellite (CHAMP, 2000). Low-low tracking: two satellites in the same low orbit range to each other (GRACE, GRACE-FO). Gravity gradiometry: accelerometers inside one satellite measure the difference in gravity across the spacecraft (ESA's GOCE, 2009 to 2013), which gives the finest static gravity field.

Scanning systems

A satellite builds an image by moving and sampling. A whiskbroom (across-track) scanner uses a rotating or oscillating mirror to sweep detectors across the track, one ground spot at a time, while the platform moves forward. A pushbroom (along-track) scanner uses a line of detectors, one per cross-track pixel, that images an entire strip at once; the platform's motion builds the second dimension. Pushbroom designs have no moving mirror, so each pixel is viewed longer, but they need many detectors calibrated against each other. A framing camera captures a whole two-dimensional scene at once, as eyes and ordinary cameras do.

Through Landsat 7, every Landsat scanning imager (MSS, TM, ETM+) was a whiskbroom (Landsats 1 to 3 also carried RBV framing cameras); Landsat 8 introduced pushbroom sensors. Modern missions prefer pushbroom designs because there is no mirror to wear out, they are smaller and lighter, and the longer dwell time raises the signal-to-noise ratio. Dwell time is calculated in the coverage section below.

Geometric distortions of scanned and photographed images

• Panoramic (tangential scale) distortion: a whiskbroom mirror sweeps at a constant angle per step, but the ground distance per step grows toward the swath edges, because the sensor is farther away and looks obliquely. Pixels near the edge cover more ground, so features there are compressed in the image and straight lines can bend into S shapes. Off-nadir resolution is therefore degraded.
• Relief displacement: tall objects lean away from nadir. In a vertical aerial photograph they lean radially away from the photo center (at the exact center you see only rooftops); in a scanner image they lean away from the nadir line, perpendicular to the track. An orthophoto has relief displacement removed, so it has uniform scale.
• Skew distortion: Earth rotates eastward under the satellite while a scene is built line by line, so each line is offset slightly west of the previous one and the image is skewed; it is corrected by deskewing.
• Platform motion: roll, pitch, yaw and altitude changes shift lines; aircraft carry IMUs to correct them.
• Scan-line gaps and striping: Landsat 7's scan line corrector failed on May 31, 2003, leaving wedge-shaped gaps (about 22% of each scene) that widen toward the scene edges. Uneven detectors cause striping (every 16th line in TM, along-track stripes in pushbroom images), removed by destriping.
• Radar distortions (foreshortening, layover, shadow, slant-range compression) are listed in the radar section above.

Imagers, radiometers and spectrometers

An imager forms an image in several bands. A multispectral imager records a handful of broad bands. A radiometer measures radiance or irradiance in a few broad channels without necessarily forming a detailed image, as climate instruments do. A spectrometer records many narrow bands and identifies composition from the shape of the spectrum. A spectroradiometer (such as MODIS, the Moderate Resolution Imaging Spectroradiometer) combines both: calibrated radiance in many bands. An optical imager records reflected visible and near-infrared light, so it needs daylight and gives fine detail; an infrared (thermal) imager records emitted heat, so it works at night and shows temperature but has coarser pixels (longer wavelengths, fewer photons). A sounder measures radiance in many channels spread across an absorption band; each channel "sees" a different altitude, so together they retrieve vertical profiles of temperature, humidity or gases (CrIS on Suomi NPP, AIRS on Aqua).

Microwave instruments come in the same classes. Passive microwave radiometers (imagers) measure the brightness temperature of naturally emitted microwaves: AMSR-E and AMSR2, GMI on GPM and the SMAP radiometer map rain, sea ice, soil moisture and sea-surface temperature through cloud. Passive microwave sounders use many channels on the edges of the oxygen (near 50 to 60 GHz and 118 GHz) and water-vapor (22 and 183 GHz) absorption lines to retrieve temperature and humidity profiles (AMSU and ATMS looking down; MLS on Aura looking at the limb). The altimeter's on-board water-vapor radiometer (Jason's AMR) is another passive microwave radiometer. Microwave spectrometers in the strict sense are rare on these lists; limb sounders such as MLS resolve individual emission lines. Active microwave instruments are radars, scatterometers and altimeters.

A Crookes radiometer (the 'light mill' sold as a toy) is a partial-vacuum bulb with vanes black on one side and silver on the other. Light or infrared heats the black faces more, and gas molecules rebounding from the warmer faces (mostly near the vane edges, an effect called thermal transpiration) push them, so the vanes spin with the black sides trailing (moving away from the light). It spins faster in brighter light, but it is a demonstration of thermal gas effects, not a measuring instrument and not radiation pressure (radiation pressure alone would push the shiny sides harder).

Rule of thumb for classification questions: a sensor that outputs a number (radiance or a temperature reading) rather than a picture is a radiometer, such as a handheld infrared thermometer or CERES; one that builds a picture is an imager (or imaging radiometer); one that splits the light into many narrow bands is a spectrometer. Measuring snow depth or snow water equivalent needs an active technique (lidar or radar altimetry from two dates, snow-free minus snow-covered) or passive microwave scattering; measuring snow albedo is passive, so a mission doing both needs both kinds of sensor.

Spectral classes of imagers

Imagers are classed by how many bands they record and how narrow those bands are; more bands identify materials better but leave less light per band.

Higher spectral resolution means more, narrower bands; it identifies materials by narrow absorption features (minerals, gases), but each band collects less light, lowering SNR and often forcing larger pixels. Normal color film and sensors have blue, green and red layers, so there is no separate yellow layer; color in images of invisible wavelengths is assigned in processing (false color). Elements of image interpretation, band combinations and spectral indices are covered in the Division B map and image chapter.

Label questions: a sensor with a handful of broad bands (Landsat OLI, Sentinel-2 MSI) is multispectral; tens of bands (MODIS's 36) are sometimes called superspectral; hundreds of narrow contiguous bands (PACE OCI, Hyperion) are hyperspectral. If both "multispectral" and "superspectral" are offered for MODIS, "superspectral" is the more specific answer; if only multispectral and hyperspectral are offered, MODIS is multispectral.

Spectroscopy and gas retrieval

Spectra come in three basic types, described by Kirchhoff's rules of spectroscopy: a continuous spectrum from a hot dense source (a blackbody curve, the Sun's photosphere, a lamp filament), an emission spectrum where a hot, low-density gas emits only at particular wavelengths (bright lines on a dark background, like a neon sign), and an absorption spectrum where a cooler gas removes particular wavelengths from a continuous background (dark lines, like the solar spectrum's Fraunhofer lines or sunlight that has crossed Earth's atmosphere). Gas identification relies on absorption features. Each line exists because an electron, vibration or rotation jumps between quantized energy levels, and the photon's frequency equals the energy difference divided by $h$. For hydrogen's visible Balmer series, $1/\lambda = R_H(1/2^2 - 1/n^2)$ with $R_H = 1.097\times10^{7}$ m⁻¹: the drop from $n = 6$ to $n = 2$ gives 410 nm (violet). Heavier hydrogen-like ions (He⁺, Li²⁺) have energy levels scaled by $Z^2$, so the same jump emits shorter wavelengths.

A band ratio divides one band by another (NDVI is a normalized version). Ratios cancel brightness differences caused by slope and shadow, and they isolate features: a ratio of a band inside a gas's absorption line to a nearby band outside it measures how much of that gas lies along the path. OCO-2 uses this idea in the 1.6 and 2.06 µm CO₂ bands, with an oxygen band as reference for the air column. A deeper absorption dip means more gas along the path, but clouds, aerosols, surface reflectance and viewing geometry also change the path length, so retrievals must account for them. The Division C list names atmospheric interactions with the terrestrial biosphere: plants draw down CO₂ each Northern Hemisphere summer and release it in winter (seen in OCO-2 maps), wetlands emit methane (seen by TROPOMI), fires release CO and aerosol, and healthy leaves emit a faint solar-induced fluorescence in the far red, which OCO-2's spectrometer retrieves near 760 nm as a measure of photosynthesis.

Viewing geometry: nadir, limb, occultation and multi-angle

Where a sensor points decides what it can measure: straight down for the surface and total columns, through the limb for thin layers and vertical profiles.

Image processing pipeline

Raw data pass through standard steps before analysis. Tests ask what each step does.

1. Radiometric correction and calibration: convert DN to radiance using the sensor's gain and offset, then to top-of-atmosphere reflectance using sun angle and Earth-Sun distance; remove striping and detector differences. Sensors are calibrated before launch, by onboard sources (lamps, solar diffusers, blackbodies) and by vicarious calibration over stable, bright, uniform desert sites (Libya-4, Railroad Valley Playa), which also tracks slow drift over years.
2. Atmospheric correction: remove scattering and absorption to get surface reflectance, using models (radiative transfer codes such as MODTRAN) or simple image-based methods such as dark object subtraction.
3. Geometric correction: remove distortions and register the image to map coordinates with ground control points; orthorectification also uses a DEM to remove relief displacement.
4. Enhancement: a histogram shows how many pixels have each DN; a contrast stretch spreads the used DN range over the full display range; density slicing groups ranges of DN into one color; low-pass filters smooth noise and speckle, high-pass filters sharpen edges; pansharpening merges a sharp panchromatic band with coarser color bands.
5. Transformation: band ratios and indices, principal component analysis (compresses correlated bands into a few new ones), and composites.
6. Classification and change detection: covered in the Division B chapter.

Ground-based measurements taken at the same time as the overpass (ground truth) are used to calibrate and validate satellite products.

Film versus digital, and long records. Photographic film records only about 0.3 to 0.9 µm in at most three broad emulsion layers, cannot sense SWIR or thermal infrared and is hard to calibrate; digital detectors cover the SWIR and thermal ranges in many narrow, calibrated bands. A climate data record spanning decades has to be stitched from several satellites, so the missions must overlap and be cross-calibrated against each other (and against stable references such as desert calibration sites) so that a change of instrument is not mistaken for a climate trend.

A mosaic is a single seamless image stitched from several overlapping scenes or flight lines, after each is georeferenced and its brightness balanced so the seams do not show; national and global base maps (and Google Earth) are mosaics of many dates.

Orbits and satellites

Satellites stay in orbit because gravity supplies the centripetal force that bends their path around Earth. The orbit's altitude sets its period, how much of the planet it sees at once and how fine its detail can be. Both divisions list orbit types by name, but the lists differ: Division B names low, medium and high Earth orbit, polar and non-polar, geostationary and geosynchronous; Division C adds sun-synchronous and uses the abbreviation HEO. Which orbit each named mission flies is in the Satellites and instruments reference; this chapter is the physics and the orbit types.

Kepler's laws and orbital calculations

Kepler's laws describe every such orbit. (1) Orbits are ellipses with the central body at one focus (a circle is the special case of zero eccentricity; parabolas and hyperbolas are unbound paths). (2) A line from the central body to the satellite sweeps equal areas in equal times, so the satellite moves fastest at perigee and slowest at apogee. (3) The square of the period is proportional to the cube of the semi-major axis. The third law depends only on the central body's mass, not the satellite's, so every satellite at the same altitude has the same period.

Eccentricity $e$ runs from 0 (circle) toward 1 (very elongated). Earth's orbit around the Sun has $e \approx 0.017$, Venus about 0.007 (the most circular planetary orbit), Mars about 0.093 and Mercury about 0.21; Neptune's large moon Triton has an almost perfectly circular (but retrograde) orbit. A Molniya orbit has $e \approx 0.74$. For an ellipse, perigee distance $= a(1-e)$, apogee distance $= a(1+e)$, and the semi-minor axis $b = a\sqrt{1-e^{2}}$.

Ellipse vocabulary: the semi-major axis $a$ is half the longest diameter; the semi-minor axis $b$ is half the shortest; the foci are the two points inside the ellipse (Earth sits at one); eccentricity $e$ measures elongation, from 0 (circle) toward 1. The closest point to Earth is perigee and the farthest apogee (around the Sun, perihelion and aphelion; in general, periapsis and apoapsis). Inclination is the angle between the orbit plane and the equator: 0° is equatorial, 90° polar, below 90° prograde (moving eastward with Earth's rotation) and above 90° retrograde.

Division C may calculate with Kepler's third law; Division B treats it qualitatively (higher orbit, longer period). The law also gives a ratio method that skips the constants. If one orbit's radius is 4 times another's, its period is $4^{3/2} = 8$ times longer. A planet at 4 AU has an 8-year period if one at 1 AU has 1 year.

Sun-synchronous orbits and equator crossing times

A sun-synchronous orbit is slightly retrograde, which makes Earth's equatorial bulge swing the orbit plane eastward by about 1° per day (360° in a year, 0.986° per day), matching the Sun's apparent motion. Consistent local time matters because comparing two dates only means something if shadows and sun angle are similar. Sun-synchronous satellites fly at about 600 to 800 km with inclinations near 98°, so they are near-polar but never pass exactly over the poles. They are not always in sunlight (except special dawn-dusk orbits such as Sentinel-1's, which ride the day-night line).

A sun-synchronous orbit is described by its equator crossing time. A descending node is a north-to-south crossing; an ascending node is south-to-north. Morning satellites (Terra about 10:30, Landsat about 10:00, Sentinel-2 about 10:30) cross southbound on the day side, before afternoon clouds build over land. Afternoon satellites (Aqua, the A-Train and Suomi NPP, about 1:30 p.m.) cross northbound, which helps study clouds and convection that develop in the afternoon. Pairing a morning and an afternoon platform (Terra and Aqua both carry MODIS) doubles daily views, shows diurnal change and reduces gaps from clouds, shadows and glare. Most remote sensing satellites use sun-synchronous orbits because passive sensors need consistent illumination.

Geostationary versus geosynchronous

A geosynchronous orbit is any orbit whose period equals one sidereal day (23 h 56 min 4 s = 86,164 s, Earth's rotation relative to the stars, not the 24 h solar day). A geostationary orbit is the special geosynchronous orbit that is circular, at 35,786 km altitude, with 0° inclination, so the satellite appears fixed in the sky above one point on the equator: every geostationary orbit is geosynchronous, but not the reverse. A geosynchronous satellite with some inclination drifts north and south of its spot each day, and with some eccentricity also east and west, tracing a figure eight (an analemma); the north-south swing can give better views of higher latitudes.

Geostationary satellites give the best temporal resolution (images every 1 to 15 minutes; GOES scans a full disk every 10 minutes and smaller sectors every 30 seconds to 1 minute), which is why they are used for weather, hurricanes, fires and communications. Their weaknesses come from distance: at about 50 times the height of a 705 km imager, the same instrument gives about 50 times larger pixels (GOES ABI pixels are 0.5 to 2 km at the sub-satellite point and grow toward the edge of the disk), and they see the poles poorly or not at all. The theoretical edge of view is about 81° latitude, and useful data stop near 70°. A geostationary satellite sees about 42% of Earth's surface (archive keys usually round this to "about a third"; pick that if both a third and a half are offered), so at least three are needed to cover all longitudes, and polar orbiters must fill in the poles. Infrared and visible sensors work from GEO, but passive microwave sensors would need impractically large antennas there, so they fly in LEO.

Distance also costs time. A signal relayed through a geostationary satellite travels about 72,000 km up and down, adding roughly 0.24 s one way (about 0.5 s for a question and its reply), while low-orbit links add only tens of milliseconds. When comparing orbits, list altitude, period, speed, coverage per satellite, number of satellites needed, resolution and signal delay.

Because a geostationary satellite sits over the equator, it views mid- and high-latitude places obliquely: the continental United States is seen at view (zenith) angles of roughly 30 to 60°, so pixels are stretched, cloud tops appear displaced poleward of their true position (parallax), and tall buildings or narrow urban canyons are seen from the side rather than straight down. A polar orbiter viewing near nadir does better for such targets.

Other orbits: Molniya, Lagrange points, graveyards and decay

• Molniya orbits are highly elliptical, inclined at 63.4° (the critical inclination at which the perigee does not drift), with a 12-hour period, perigee of a few hundred to about 1,000 km and apogee near 40,000 km over the Northern Hemisphere. By Kepler's second law the satellite spends about two thirds of each orbit near apogee, giving Russia and Canada long views that GEO cannot. A Tundra orbit is the 24-hour (geosynchronous) version.
• Lagrange points are where the gravity of the Sun and Earth plus orbital motion balance. L1, about 1.5 million km toward the Sun, gives a constant view of Earth's sunlit face (DSCOVR and its EPIC camera); L2 lies the same distance on the night side (JWST). L1, L2 and L3 are unstable (spacecraft need station-keeping); L4 and L5 are stable.
• Graveyard orbit: dead geostationary satellites are boosted a few hundred kilometers above the GEO ring to clear the valuable slots.
• Orbital decay: in LEO, thin upper-atmosphere drag lowers orbits, especially below about 600 km and when solar activity heats and expands the upper atmosphere. LEO satellites have the shortest lifespans, need reboosts, and crowd the region where most satellites fly (debris collisions could cascade, the Kessler syndrome). MEO satellites spend time in the Van Allen radiation belts, so they need radiation-hardened electronics.
• Constellations: GPS uses 24 nominal satellites (more are operational) in six MEO planes inclined 55°, so at least four are visible from anywhere for positioning. Covering Earth continuously takes about 3 GEO satellites, about 24 in MEO, and hundreds to thousands in LEO.

Coverage, swath and revisit

Low orbits give finer spatial detail for the same instrument, because each detector's ground footprint shrinks with altitude. The price is coverage: a low satellite sees a narrow strip, the swath (the width of ground imaged perpendicular to the flight direction). A geostationary satellite sees a hemisphere and images a storm every few minutes, but with coarser pixels. The point directly below the satellite is the nadir; the line of nadir points is the ground track.

Do not confuse three periods. The orbital period is the time for one orbit. The revisit period (repeat cycle, temporal resolution) is the time between images of one location, set by the orbit, the swath and any off-nadir pointing. For a camera that takes frames, the frame rate must be high enough that successive frames just touch along the track: $f = v_{g}/L$, where $v_{g}$ is the speed of the ground track and $L$ the along-track footprint length. A shorter orbital period does not by itself mean better temporal resolution.

Because Earth rotates eastward under the orbit, each successive ground track lands west of the last. The shift per orbit is the fraction of a day the orbit takes, times 360°. Polar swaths converge toward the poles, so high latitudes are revisited much more often than the equator, and adjacent swaths overlap there (sidelap, or lateral overlap).

Revisit can be shorter than the repeat cycle: pointing the sensor off-nadir (ASTER, commercial imagers) or flying two or more satellites in the same orbit shortens it. Two satellites phased 180° apart halve the revisit, which is how Sentinel-1A and 1B gave 6 days instead of 12, Sentinel-2A and 2B gave 5 days instead of 10, and Landsat 8 and 9 give 8 days instead of 16. A constellation's temporal resolution is therefore better than any single member's.

Dwell time is how long a detector stares at one ground cell while collecting its signal. It is set by platform speed, pixel size and how many pixels share the scan. A pushbroom detector stares at its cell for a whole line time; a whiskbroom shares that line time among every pixel across the swath, so its dwell time is shorter by a factor equal to the number of pixels it sweeps per line (divided by the number of detectors sweeping together). Longer dwell time collects more photons and raises the signal-to-noise ratio. Time-delay integration (TDI) lets several detector rows look at the same cell in turn to lengthen effective dwell time.

Adding satellites to a constellation shortens the revisit interval (better temporal resolution) below that of any single member, but each satellite's orbital period is unchanged.

Weather satellites (GOES, VIIRS, MODIS) trade spatial detail for frequent, wide coverage: pixels of 0.25 to 2 km and 16 to 36 bands chosen largely for clouds, water vapor and temperature. Land-use satellites (Landsat, Sentinel-2) have 10 to 30 m pixels and bands chosen for vegetation, soil and water, but revisit only every few days to weeks. Among the Division C satellites, CloudSat, CALIPSO and GPM are dedicated to clouds, aerosols and precipitation; Jason, GRACE and OCO-2 are not.

Image quality: resolution and signal

Image quality has four independent parts, and both divisions test each one. Spatial resolution is the smallest ground feature a sensor can distinguish. Spectral resolution is the number and width of the bands it measures. Radiometric resolution is how many brightness levels it can distinguish, limited in practice by noise. Temporal resolution is how often it revisits a location. A sensor can be excellent in one and poor in another, so always name the type of resolution a question is asking about.

The four resolutions and their trade-offs

The resolutions compete. Finer pixels need more detectors for the same swath, so with fixed detectors the swath narrows and revisit slows; that is why Landsat (30 m, 185 km swath, 16 days) and MODIS (250 to 1,000 m, 2,330 km swath, 1 to 2 days) are complementary, and why weather satellites trade detail for coverage. Narrower bands and smaller pixels each cut the light per pixel, lowering SNR, which is why panchromatic bands (one broad band, lots of light) get the finest pixels and why MODIS uses 250 m for a few bright bands but 1 km for many narrow ones. More bits, bands and pixels also mean more data to store and transmit. Choose by the job: a continental heat wave needs daily kilometer-scale data; a crop field or a car park needs meters; mineral mapping needs spectral resolution; a fast storm needs temporal resolution.

IFOV, GSD, FOV and swath

Spatial resolution has two closely related quantities that students confuse. The instantaneous field of view (IFOV) is the angle one detector element sees, set by the detector size divided by the focal length. The ground sample distance (GSD) is the ground size of one pixel (formally the distance between adjacent pixel centers on the ground). The field of view (FOV) is the whole angular width the sensor can see; on the ground it becomes the swath. The ground area seen by one IFOV is the resolution cell. Only the GSD is measured on the ground, and it is the pixel size, not the resolution limit. The calculations below belong to the Division C imaging-quality item; Division B needs the ideas (lower altitude or a narrower IFOV gives smaller pixels) but not the arithmetic. A larger GSD means coarser (lower) spatial resolution; the orbital period does not affect GSD.

The diffraction limit

Even perfect optics blur a point into a small spot, because light diffracts (spreads) through the aperture. The smallest angle a circular aperture can resolve is $1.22\lambda/D$ (the Rayleigh criterion). Resolution improves with a bigger aperture or a shorter wavelength; that is why thermal and microwave sensors, with long wavelengths, have coarse pixels unless their apertures are huge, and why SAR synthesizes a long antenna. A sensor's real resolution is set by whichever is larger: the diffraction angle or the IFOV.

Pixels, mixed pixels and off-nadir growth

A pixel (picture element) is the smallest unit of a digital image: the single value one detector element records for one ground cell during one sampling interval, stored as a digital number. A pixel that contains more than one surface type is a mixed pixel ("mixel"); its value is a blend, so it is common at boundaries (urban-rural edges, shorelines, field edges) and in coarse images, and it causes classification errors. Uniform areas such as a desert dune field or snowfield produce few mixels. Spatial resolution and pixel size are not identical: images can be resampled to smaller pixels without gaining detail, and the ground resolved distance (the smallest separation at which two objects can be told apart) is typically about twice the GSD, because at least two pixels are needed to show a gap between objects.

Pixels grow away from nadir. For a flat Earth, a pixel viewed at angle $\theta$ from nadir is about $H \cdot \text{IFOV}/\cos\theta$ long in the along-track direction and $H \cdot \text{IFOV}/\cos^2\theta$ wide across track, which is why wide-swath sensors (MODIS, GOES) have much coarser pixels at their edges.

Photographic scale

For a vertical aerial photograph over flat ground, the image scale is the ratio of the camera's focal length to its height above the ground: $\text{scale} = f/H$. With units made the same, it is written as a representative fraction.

Signal-to-noise ratio and sensitivity

Division B limits radiometric resolution to signal-to-noise ratio and sensitivity. The signal-to-noise ratio (SNR) is how strong the measured signal is compared with the random noise of the detector and electronics. A high SNR means two surfaces with slightly different brightness can be told apart; a low SNR buries small differences in speckle and grain. Sensitivity is the smallest change in radiance (or, for a thermal sensor, temperature) the instrument can reliably detect. Dark targets such as water and dim bands such as narrow thermal or SWIR channels have the lowest signal, so they suffer most from noise. Larger pixels, longer viewing time per pixel (one reason pushbroom sensors outperform whiskbroom ones) and cooled detectors all raise SNR.

A common detection rule is that a signal must be at least about 3 times the noise (SNR ≥ 3) to be reliably detected. Noise has several sources: photon (shot) noise, which grows as the square root of the signal; thermal noise and dark current in the detector (reduced by cooling); read and electronic noise; quantization noise from rounding to whole DN; and striping from mismatched detectors. SNR goes up with a larger aperture, bigger pixels, wider bands, longer dwell time or TDI, cooler detectors, brighter targets, and by averaging pixels or radar looks (which trades away resolution). It goes down when bands are narrowed (hyperspectral) or pixels shrunk. For thermal sensors sensitivity is quoted as the noise-equivalent temperature difference (NEΔT or NEdT): the temperature change that produces a signal equal to the noise, typically a few hundredths to about a tenth of a kelvin for modern thermal imagers. Sensor degradation over years lowers sensitivity, which is one reason long climate records need careful recalibration.

SNR for active sensors. For radar and lidar the signal is the echo, so SNR rises with anything that returns more energy: more transmitted power, a larger receiving aperture, a shorter range (the echo falls as $1/R^2$ for an extended target, $1/R^4$ for a point target in radar), and a target that scatters more. A target's backscatter grows with the number density of scatterers and their scattering cross-section, which depends on wavelength: molecules scatter lidar light as $1/\lambda^4$ (so ultraviolet and visible lidars get strong air returns), and a weather radar's echo from small drops grows with drop diameter to the sixth power and falls with radar wavelength to the fourth power (short-wavelength cloud radars such as CloudSat's 3 mm see small cloud drops that a 10 cm weather radar misses). Dense, strongly scattering targets give high SNR; thin, clear air gives low SNR.

A common detection rule: an SNR of about 3 is only the threshold of detection (a barely detectable signal), an SNR of about 5 or more is a confident detection, and quantitative measurements usually need SNRs of tens to hundreds.

Bit depth, dynamic range and saturation

Radiometric resolution is also described by bit depth: an 8-bit sensor records $2^{8} = 256$ levels per band and a 12-bit sensor records $2^{12} = 4{,}096$. Finer quantization does not help if the noise is larger than the level spacing, so a high bit depth with a low SNR gives little extra information: a 12-bit sensor still records 4,096 levels, but fewer of them are usefully separable. Analog-to-digital conversion turns the detector's voltage into these digital numbers (DN). On an 8-bit grayscale display, 0 is black, 255 is white and a value near 127 or 128 is mid-gray.

Dynamic range is the span between the faintest and brightest signals a sensor records without error. Very bright targets (fresh snow, cloud tops, sun glint, white sand) can exceed the top level and saturate: every bright pixel reads 255 (for 8 bits) and the detail is lost. Older 8-bit sensors such as Landsat 7 ETM+ saturated over snow and cloud and used high and low gain settings; 12- and 14-bit OLI rarely saturate. Radar, which does not measure sunlight, is unaffected by surface albedo. For detecting small infrared differences from a fine-resolution, daily-revisit sensor, 12 to 16 bits is appropriate.

Satellites and instruments reference

Tests name satellites and instruments constantly. A question may ask what an acronym stands for, which platform carries an instrument, when a mission launched, what orbit it flies, or which sensor you would choose for a job. This chapter is the binder lookup for all of that. The first table covers platforms (the spacecraft), the second covers instruments (what is on board), the third is a quick-pick table for "which satellite would you use", and the last expands every acronym that tests ask you to write out.

Platforms

Use this table for "which satellite", "when", "what orbit" and "what does it carry" questions. Altitudes are nominal; the equator-crossing time is local solar time. "Desc." means the satellite crosses the equator heading south (descending node) on the daylight side, "asc." heading north.

ReferenceSatellite platforms· 42 rows

SSO = sun-synchronous orbit. GEO = geostationary. ISS = International Space Station. Div = which 2027 list names it (B, C, both, or "—" for background missions that tests still ask about).

42 rows

Platform	Full name, agency	Launch (end)	Orbit	Instruments	Known for	Div
Landsat 1 to 4	Landsat 1 was first called ERTS-1 (Earth Resources Technology Satellite); NASA	L1 July 23, 1972; L4 July 16, 1982; Landsat 6 (October 5, 1993) failed to reach orbit	SSO, about 900 km (L1 to 3), 705 km (L4)	MSS (all); TM first flown on Landsat 4	Start of the continuous civilian land record (1972 to today); MSS had 4 bands at about 80 m	—
Landsat 5	NASA, USGS	March 1, 1984 (retired June 2013)	SSO, 705 km, 16-day repeat	TM, MSS	Guinness record: longest-operating Earth-observing satellite (about 29 years); the backbone of long change studies from the 1980s	B
Landsat 7	NASA, USGS	April 15, 1999 (science imaging ended January 19, 2024; decommissioned June 4, 2025)	SSO, 705 km (lowered to 697 km in 2022), 98.9 min period, 16 days, about 10:00 a.m. desc. (drifted earlier after 2017)	ETM+	Added a 15 m panchromatic band; its scan line corrector (SLC) failed on May 31, 2003, leaving wedge-shaped data gaps in every later image	B
Landsat 8	NASA, USGS	February 11, 2013	SSO, 705 km, 98.2°, 99 min, 16 days, about 10:00 a.m. desc.	OLI, TIRS	Pushbroom imager (no scanning mirror); 12-bit data; 9 OLI bands plus 2 thermal bands	B
Landsat 9	NASA, USGS	September 27, 2021	Same orbit as Landsat 8, offset 8 days	OLI-2, TIRS-2	Near-copy of Landsat 8 with 14-bit data; Landsat 8 and 9 together image each place every 8 days	B
Sentinel-1A, 1B (1C)	ESA, Copernicus	1A April 3, 2014 (the first Copernicus satellite); 1B April 25, 2016 (failed December 2021); 1C December 2024	SSO, 693 km, dawn-dusk (about 6 a.m. / 6 p.m. crossings), 12-day repeat each	C-band SAR (5.405 GHz)	All-weather, day-night radar; two satellites in the same orbit phased 180° apart give a 6-day repeat; floods, InSAR subsidence, sea ice, oil slicks	B
Sentinel-2A, 2B (2C)	ESA, Copernicus	2A June 23, 2015; 2B March 7, 2017; 2C September 2024	SSO, 786 km, 10:30 a.m. desc., 10-day repeat each	MSI	13 bands from 443 to 2,190 nm at 10, 20 or 60 m, including three red-edge bands; 290 km swath; 5-day repeat with two satellites	B
Sentinel-5P	ESA, Copernicus (P = Precursor)	October 13, 2017	SSO, 824 km, 1:30 p.m. asc.	TROPOMI	Daily global maps of NO₂, ozone, SO₂, CO, CH₄, formaldehyde and aerosol	C
Sentinel-6 Michael Freilich	ESA, EUMETSAT, NASA, NOAA, CNES	November 21, 2020	1,336 km, 66°, non-sun-synchronous, 10-day repeat (Jason orbit)	Poseidon-4 radar altimeter, AMR-C radiometer	Continues the Jason sea-level record	C (as Jason successor)
Terra (EOS AM-1)	NASA flagship, with Japan and Canada; Latin for "Earth"	December 18, 1999	SSO, 705 km, 98.8 min period, about 10:30 a.m. desc.	ASTER, CERES, MISR, MODIS, MOPITT	"Morning" Earth Observing System satellite; land, clouds, aerosols, radiation; drifting toward end of mission as its fuel runs out	both
Aqua (EOS PM-1)	NASA, with Japan and Brazil; Latin for "water"	May 4, 2002	SSO, 705 km, about 1:30 p.m. asc.	MODIS, CERES, AIRS, AMSU-A, HSB, AMSR-E	Water cycle: evaporation, clouds, precipitation, sea ice, SST; first A-Train member; nearing end of mission as fuel runs out	C
Aura	NASA, with the Netherlands, Finland and the UK; Latin for "breeze"	July 15, 2004	SSO, 705 km, about 1:45 p.m. asc.	OMI, MLS, TES, HIRDLS	Atmospheric chemistry: the ozone layer, air quality, trace gases	C
PARASOL	CNES (France)	December 18, 2004 (left A-Train 2009, ended 2013)	SSO, A-Train	POLDER	Polarization and multi-angle views of aerosols and clouds	—
CALIPSO	NASA and CNES	April 28, 2006, on one rocket with CloudSat (science ended August 1, 2023)	SSO, 705 km A-Train, lowered to 688 km in 2018	CALIOP lidar, IIR, WFC	First polarization lidar in space, optimized for aerosols and clouds; vertical profiles of thin clouds and aerosol layers	C
CloudSat	NASA (with the Canadian Space Agency)	April 28, 2006, with CALIPSO (radar science ended December 20, 2023; decommissioned 2024)	SSO, 705 km A-Train, lowered to 688 km in 2018	CPR (94 GHz radar)	First satellite millimeter-wave cloud radar; vertical structure of thick clouds and light precipitation	C
GCOM-W1 "Shizuku"	JAXA (Global Change Observation Mission – Water)	May 2012	SSO, A-Train	AMSR2	Passive microwave water cycle: SST, sea ice, soil moisture, rain	—
OCO	NASA (Orbiting Carbon Observatory)	February 24, 2009: launch failure (the rocket's fairing did not separate)	—	—	Never reached orbit; replaced by OCO-2	—
Glory	NASA	March 4, 2011: launch failure (same fairing problem)	—	—	Planned A-Train aerosol and solar-irradiance mission	—
OCO-2	NASA	July 2, 2014	SSO, 705 km, leads the A-Train, about 1:36 p.m. asc., 16-day repeat	Three-grating spectrometer	Column CO₂ (XCO₂) at about 1.3 × 2.3 km footprints; solar-induced fluorescence	C
OCO-3	NASA, on the ISS	May 4, 2019, from Cape Canaveral	ISS orbit (about 400 km, 51.6°)	Spare OCO-2 spectrometer with a pointing mirror	CO₂ at different times of day; snapshot maps of cities	—
Suomi NPP	NASA and NOAA (National Polar-orbiting Partnership; named for Verner Suomi)	October 28, 2011	SSO, 824 km, 1:30 p.m. asc.	VIIRS, CrIS, ATMS, OMPS, CERES	Bridge between NASA's EOS satellites and NOAA's operational JPSS series	C
NOAA-20, NOAA-21 (JPSS-1, -2)	NOAA and NASA (Joint Polar Satellite System)	November 18, 2017; November 10, 2022	Same orbit as Suomi NPP	VIIRS, CrIS, ATMS, OMPS (CERES on NOAA-20)	Operational polar weather satellites	—
GOES (GOES-R series: 16, 17, 18, 19)	NOAA, with NASA (Geostationary Operational Environmental Satellite)	GOES-1 1975; GOES-16 November 19, 2016; GOES-17 March 1, 2018; GOES-18 March 1, 2022; GOES-19 June 25, 2024	GEO, 35,786 km; GOES-East near 75.2° W, GOES-West near 137.2° W	ABI, GLM, space-weather sensors	Continuous weather watching of the Americas: storms, hurricanes, fires, smoke, ash, fog; GOES-19 became GOES-East in 2025 and GOES-18 is GOES-West	both
TRMM	NASA and JAXA (Tropical Rainfall Measuring Mission)	November 27, 1997 (ended 2015)	35° inclination, about 350 to 400 km	Precipitation Radar, TMI, VIRS, LIS, CERES	First space-borne rain radar; tropical rainfall	—
GPM Core Observatory	NASA and JAXA (Global Precipitation Measurement)	February 27, 2014 (UTC), from Japan	407 km, 65° inclination, non-sun-synchronous	DPR, GMI	Rain and snow in 3D from the tropics to high latitudes; calibrates a constellation of partner satellites for global rain maps every 30 minutes (IMERG)	C
TOPEX/Poseidon	NASA and CNES	August 10, 1992 (ended January 2006)	1,336 km, 66°, 9.9-day repeat	Dual-frequency altimeter, microwave radiometer	Began the precise satellite sea-level record; mapped El Niño from sea-surface height	C (Jason family)
Jason-1, Jason-2 (OSTM), Jason-3	NASA and CNES; NOAA and EUMETSAT from Jason-2	Jason-1 December 7, 2001; Jason-2 June 20, 2008; Jason-3 January 17, 2016	1,336 km, 66°, 9.9-day repeat (same ground track as TOPEX)	Poseidon altimeter (C and Ku band), AMR (JMR on Jason-1), DORIS, GPS, laser retroreflector	Sea-surface height to about 3 cm per pass; global mean sea level, ENSO, wave height, wind speed	C
GRACE	NASA and DLR (Gravity Recovery and Climate Experiment); nicknamed "Tom and Jerry"	March 17, 2002 (ended October 2017)	Near-polar (89°), about 500 km, two satellites about 220 km apart	K-band microwave ranging, accelerometers, GPS	Monthly gravity maps: ice-sheet loss, groundwater depletion, ocean mass	C
GRACE-FO (Follow-On)	NASA and GFZ	May 22, 2018	Same as GRACE	K-band ranging plus a laser ranging interferometer	Continues GRACE; the laser system is about 20 times more precise	C
PACE	NASA (Plankton, Aerosol, Cloud, ocean Ecosystem)	February 8, 2024	SSO, 676 km, about 1 p.m. asc.	OCI, HARP2, SPEXone	Hyperspectral ocean color (phytoplankton types), aerosols and clouds with polarimeters	C
ECOSTRESS	NASA JPL, mounted on the ISS (ECOsystem Spaceborne Thermal Radiometer Experiment on Space Station)	June 29, 2018	ISS orbit, about 400 km, 51.6°, not sun-synchronous	PHyTIR thermal radiometer	Land surface temperature at about 70 m at different times of day: plant water stress, urban heat islands	both
SRTM	NASA, NGA, with Germany and Italy (Shuttle Radar Topography Mission)	Space Shuttle Endeavour STS-99, February 11 to 22, 2000	Shuttle orbit about 233 km, 57°	C-band and X-band radar with a 60 m mast for the second antenna	Single-pass InSAR elevation of land from 60° N to 56° S (about 80% of land); 1 arc-second (about 30 m) and 3 arc-second (about 90 m) DEMs	B
Seasat	NASA JPL	June 27, 1978 (failed after 106 days)	800 km, 108°	L-band SAR, radar altimeter, scatterometer	First civilian satellite SAR and first ocean-dedicated satellite	—
ICESat, ICESat-2	NASA (Ice, Cloud and land Elevation Satellite)	ICESat 2003 (ended 2010); ICESat-2 September 15, 2018	Polar, about 500 to 600 km	GLAS lidar (1064 and 532 nm); ATLAS photon-counting lidar (532 nm, 6 beams)	Ice-sheet and sea-ice height change, canopy height	—
CryoSat-2	ESA	April 8, 2010	717 km, 92°	SIRAL radar altimeter (Ku band, SAR and interferometric modes)	Sea-ice thickness and ice-sheet elevation	—
SWOT	NASA and CNES (Surface Water and Ocean Topography)	December 16, 2022	890 km, 77.6°	KaRIn Ka-band radar interferometer	Wide-swath heights of oceans, lakes and rivers	—
SMAP	NASA (Soil Moisture Active Passive)	January 31, 2015	SSO, 685 km, 6 a.m. / 6 p.m.	L-band radiometer (its radar failed in July 2015)	Global soil moisture and freeze-thaw	—
SPOT 1 to 7	CNES, later Airbus (Satellite Pour l'Observation de la Terre)	SPOT 1 February 22, 1986	SSO, about 822 km (SPOT 1 to 5); 694 km (SPOT 6, 7)	HRV pushbroom imagers (later HRG, NAOMI)	First satellite pushbroom imager; off-nadir pointing for stereo	—
DSCOVR	NOAA, NASA, US Air Force	February 11, 2015	Sun-Earth L1 point, about 1.5 million km sunward	EPIC camera, NISTAR radiometer, solar-wind sensors	Whole sunlit Earth in one frame; space-weather warnings	—
TIROS-1	NASA (Television Infrared Observation Satellite)	April 1, 1960	LEO, about 700 km	TV cameras	First successful weather satellite	—
Syncom 2, Syncom 3	NASA and Hughes	July 26, 1963; August 19, 1964	Geosynchronous (inclined); Syncom 3 the first geostationary	Communications	Designed by Harold Rosen's team at Hughes; first geosynchronous and first geostationary satellites	—
Sputnik 1	Soviet Union	October 4, 1957	Elliptical LEO	Radio transmitter	First artificial satellite	—

How to use it. A question that names a phenomenon usually expects one platform from your division's list: El Niño from sea-surface height means Jason; groundwater loss means GRACE; CO₂ means OCO-2; ozone means Aura; aerosol layers at height mean CALIPSO; cloud vertical structure means CloudSat; rain in 3D means GPM; continuous storm tracking means GOES. If a question asks for a launch year, the platform's row has it; if it asks for an instrument, check the instruments table next.

Every Landsat (1 to 9) was launched from Vandenberg Air Force Base (now Space Force Base) in California, the usual site for polar and sun-synchronous launches. Aqua's mission statement is to study Earth's water cycle: water in all its forms, liquid, solid and vapor (evaporation, water vapor, clouds, precipitation, soil moisture, sea and land ice, snow). NASA's Earth Observing System (EOS) includes Terra, Aqua, Aura, Landsat 7, ICESat and SORCE among others; the A-Train members are a subset. Other radar (microwave) imaging missions tests mention: ERS-1 and ERS-2 (ESA, 1991 and 1995, C-band SAR), Envisat (ESA, 2002 to 2012, ASAR C-band), RADARSAT (Canada, C band) and RISAT (India, C- and X-band). Before ABI, GOES imagers were the VISSR (Visible and Infrared Spin Scan Radiometer) on the spinning GOES-1 to 7 (with the VAS sounder added on GOES-4 to 7) and the GOES Imager and Sounder on three-axis-stabilized GOES-8 to 15; ABI began with GOES-16.

The A-Train

The A-Train ("Afternoon Train", also called the Afternoon Constellation) is a group of sun-synchronous satellites flying one behind another, a few seconds to minutes apart, at about 705 km. They cross the equator northbound at about 1:30 p.m. local time, so their instruments see the same clouds and air within minutes. That lets scientists combine measurements of one scene: MODIS gives the horizontal picture, CloudSat's radar and CALIPSO's lidar give the vertical slice, CERES gives the energy budget and OCO-2 the CO₂.

Each member had its own lead: OCO-2 (NASA JPL), GCOM-W1 (JAXA), Aqua and Aura (NASA Goddard Space Flight Center), CALIPSO (NASA Langley Research Center with France's CNES), CloudSat (NASA JPL with the Canadian Space Agency and the U.S. Air Force) and PARASOL (CNES).

• Order (from the front) at the constellation's peak, around 2014 to 2018: OCO-2, GCOM-W1, Aqua, CALIPSO, CloudSat, Aura. Before 2011, CloudSat flew about 15 seconds ahead of CALIPSO, so older sources list CloudSat first; after a 2011 battery anomaly CloudSat rejoined in 2012 about 108 seconds behind CALIPSO. PARASOL flew in it from 2004 to 2009.
• Launch order of members: Aqua (2002, the first), Aura (2004), PARASOL (2004), CloudSat and CALIPSO together (2006), GCOM-W1 (2012), OCO-2 (2014).
• Failed members: OCO (2009) and Glory (2011) were lost at launch.
• Changes: CloudSat and CALIPSO lowered their orbits to 688 km in 2018, forming the "C-Train"; CALIPSO ended science on August 1, 2023 and CloudSat on December 20, 2023. Aqua and Aura, low on fuel, have drifted out of formation, so by 2024 to 2025 only OCO-2 and GCOM-W1 remain in the A-Train.
• Not members: Terra (it flies in the morning), Landsat, Suomi NPP (same crossing time but a higher 824 km orbit), GOES (geostationary), GRACE.

Instruments

Use this table when a question names an instrument or asks what an instrument measures. "Bands" counts spectral channels.

ReferenceInstruments on the listed platforms· 46 rows

VNIR = visible and near-infrared. SWIR = shortwave infrared (about 1 to 3 µm). TIR = thermal infrared (about 8 to 14 µm).

46 rows

Instrument	Full name	Platform(s)	Technique	Bands, resolution, swath	Measures
MSS	Multispectral Scanner	Landsat 1 to 5	Passive whiskbroom scanner	4 bands (green, red, two NIR), about 80 m, 185 km	First Landsat land imagery
TM	Thematic Mapper	Landsat 4, 5	Passive whiskbroom (oscillating mirror)	7 bands: 3 visible, NIR, 2 SWIR at 30 m; thermal at 120 m; 185 km; 8-bit	Land cover, vegetation, water
ETM+	Enhanced Thematic Mapper Plus	Landsat 7	Passive whiskbroom	TM's bands plus band 8 panchromatic (0.52 to 0.90 µm) at 15 m; thermal improved to 60 m; 185 km; 8-bit	As TM, with sharper pan detail; SLC failure gaps after 2003
OLI, OLI-2	Operational Land Imager	Landsat 8, 9	Passive pushbroom	9 bands: coastal/aerosol, blue, green, red, NIR, SWIR1, SWIR2 at 30 m; pan at 15 m; cirrus; 185 km; 12-bit (OLI-2 14-bit)	Land, coast, water, vegetation
TIRS, TIRS-2	Thermal Infrared Sensor	Landsat 8, 9	Passive pushbroom thermal	2 bands (10.6 to 11.2 and 11.5 to 12.5 µm) at 100 m	Surface temperature, irrigation water use, heat islands
MSI	Multispectral Instrument	Sentinel-2	Passive pushbroom	13 bands, 443 to 2,190 nm, 10/20/60 m, 290 km, 12-bit	Vegetation (red-edge), land cover, coasts
C-SAR	C-band Synthetic Aperture Radar	Sentinel-1	Active SAR, 5.405 GHz (5.5 cm)	Interferometric Wide mode 250 km at about 5 × 20 m (the default over land); Extra Wide 400 km; Stripmap 80 km at 5 × 5 m; Wave mode; single or dual polarization (VV+VH or HH+HV)	Floods, deformation (InSAR), sea ice, ships, oil spills, soil moisture
MODIS	Moderate Resolution Imaging Spectroradiometer	Terra, Aqua	Passive whiskbroom (rotating double-sided mirror)	36 bands, 0.405 to 14.385 µm; bands 1 and 2 at 250 m, bands 3 to 7 at 500 m, bands 8 to 36 at 1 km; 2,330 km swath; whole Earth every 1 to 2 days; 12-bit	Clouds, aerosols, fires, SST, land temperature, snow, ocean color, vegetation (NDVI from band 1 red and band 2 NIR); bands 20 to 25 and 27 to 36 are thermal-emissive
ASTER	Advanced Spaceborne Thermal Emission and Reflection Radiometer	Terra	Passive, three telescopes; Japanese (METI)	14 bands: VNIR 3 bands at 15 m plus a backward-looking band 3B for stereo; SWIR 6 bands at 30 m (failed 2008); TIR 5 bands at 90 m; 60 km swath, pointable	Surface temperature and emissivity, minerals, volcanoes, glaciers; stereo DEM (ASTER GDEM)
MISR	Multi-angle Imaging SpectroRadiometer	Terra	Passive, 9 cameras at 9 view angles	4 bands, 275 m to 1.1 km, 360 km	Aerosols, cloud heights, surface roughness from multiple angles
MOPITT	Measurements Of Pollution In The Troposphere	Terra	Passive gas-correlation radiometer; Canadian	22 km pixels, 640 km	Carbon monoxide
CERES	Clouds and the Earth's Radiant Energy System	TRMM, Terra, Aqua, Suomi NPP, NOAA-20	Passive broadband scanning radiometer	3 channels: shortwave (0.3 to 5 µm), total (0.3 to beyond 100 µm), window (8 to 12 µm); about 20 km (Terra, Aqua)	Reflected sunlight and emitted longwave at the top of the atmosphere: Earth's energy budget, cloud radiative effect
AIRS	Atmospheric Infrared Sounder	Aqua	Passive grating spectrometer	2,378 IR channels, 3.7 to 15.4 µm	Temperature and humidity profiles, CO₂, CH₄
AMSU-A, HSB	Advanced Microwave Sounding Unit; Humidity Sounder for Brazil	Aqua	Passive microwave sounders	AMSU 15 channels; HSB failed in 2003	Temperature and humidity through clouds
AMSR-E, AMSR2	Advanced Microwave Scanning Radiometer (for EOS; 2)	Aqua (Japanese instrument); GCOM-W1	Passive microwave, conical scan	6.9 to 89 GHz	SST through clouds, sea ice, soil moisture, rain, snow
OMI	Ozone Monitoring Instrument	Aura	Passive UV-visible imaging spectrometer; Dutch and Finnish	270 to 500 nm, 13 × 24 km, 2,600 km swath, daily	Ozone column, NO₂, SO₂, aerosols
MLS	Microwave Limb Sounder	Aura (also flew on UARS)	Passive microwave, views the limb	118 GHz to 2.5 THz	Stratospheric ozone, ClO, HCl, water vapor, temperature: ozone-hole chemistry
TES	Tropospheric Emission Spectrometer	Aura (ended 2018)	Passive Fourier-transform IR spectrometer	3.3 to 15.4 µm	Tropospheric ozone, CO, water vapor
HIRDLS	High Resolution Dynamics Limb Sounder	Aura (ended 2008)	Passive IR limb sounder; UK and US	21 channels	Upper troposphere and stratosphere temperature, ozone, aerosols
CALIOP	Cloud-Aerosol Lidar with Orthogonal Polarization	CALIPSO	Active lidar	532 and 1,064 nm, with polarization at 532 nm; 30 m vertical, 333 m horizontal	Height and type of aerosol and thin-cloud layers; polarization separates ice from water and dust from smoke
IIR, WFC	Imaging Infrared Radiometer; Wide Field Camera	CALIPSO	Passive	IIR 3 thermal channels; WFC one visible band	Context for the lidar curtain; cirrus properties
CPR	Cloud Profiling Radar	CloudSat	Active nadir radar, 94 GHz (about 3 mm, W band)	About 500 m vertical resolution (sampled every 240 m), 1.4 × 1.7 km footprint	Vertical structure of clouds, drizzle and light rain
ABI	Advanced Baseline Imager	GOES-16 to 19	Passive imager	16 bands: 2 visible, 4 near-IR, 10 IR; 0.5 km (red band), 1 km, 2 km; full disk every 10 min, continental US every 5 min, mesoscale boxes every 1 min	Clouds, storms, fires, smoke, dust, fog, SST, water vapor
GLM	Geostationary Lightning Mapper	GOES-16 to 19	Passive optical detector at 777.4 nm	8 to 14 km pixels	Total lightning, day and night
VIIRS	Visible Infrared Imaging Radiometer Suite	Suomi NPP, NOAA-20, NOAA-21	Passive whiskbroom	22 bands (0.41 to 12.5 µm): 5 imagery bands at 375 m, 16 at 750 m, and a Day/Night Band; 3,040 km swath	Successor to MODIS and AVHRR: fires, night lights, SST, clouds, ocean color
CrIS	Cross-track Infrared Sounder	Suomi NPP, NOAA-20, -21	Passive Fourier-transform IR spectrometer	1,305 channels (2,211 in full-resolution mode), 3.9 to 15.4 µm; 14 km footprints	Temperature and humidity profiles for forecasts; trace gases
ATMS	Advanced Technology Microwave Sounder	Suomi NPP, NOAA-20, -21	Passive microwave sounder	22 channels, 23 to 183 GHz	Temperature and humidity through clouds
OMPS	Ozone Mapping and Profiler Suite	Suomi NPP, NOAA-20, -21	Passive UV spectrometers	Nadir mapper and limb profiler	Total ozone and ozone profiles
DPR	Dual-frequency Precipitation Radar	GPM Core	Active radar at Ku (13.6 GHz) and Ka (35.5 GHz)	5 km footprint, 250 m vertical, about 245 km swath	3D rain and snow; two frequencies attenuate differently, which gives drop size
GMI	GPM Microwave Imager	GPM Core	Passive microwave, conical scan	13 channels, 10 to 183 GHz, 885 km swath	Rain rate, from heavy to light rain and snow
OCI	Ocean Color Instrument	PACE	Passive hyperspectral spectrometer	340 to 890 nm at 5 nm resolution (2.5 nm sampling) plus SWIR bands to 2,260 nm; about 1 km; global every 1 to 2 days	Phytoplankton community types, chlorophyll, aerosols, clouds
HARP2, SPEXone	Hyper-Angular Rainbow Polarimeter 2; Spectro-polarimeter	PACE	Passive multi-angle polarimeters (HARP2 from the US, SPEXone from the Netherlands)	HARP2 4 bands at up to 60 angles, wide swath; SPEXone 385 to 770 nm at 5 angles, narrow swath	Aerosol and cloud-droplet size, shape and type
Poseidon-3 and 3B; Poseidon-4	Radar altimeters	Jason-2, Jason-3; Sentinel-6	Active nadir radar at Ku (13.6 GHz) and C (5.3 GHz) band	Two frequencies so the ionosphere's delay (which depends on frequency) can be removed	Sea-surface height, significant wave height (echo shape), wind speed (echo strength)
AMR	Advanced Microwave Radiometer	Jason-2, Jason-3 (AMR-C on Sentinel-6)	Passive microwave, 18.7, 23.8 and 34 GHz	Nadir	Water vapor along the path, to correct the altimeter's wet-troposphere delay
DORIS	Doppler Orbitography and Radiopositioning Integrated by Satellite	Jason series, Sentinel-3, SPOT	Ground beacons and Doppler shift	—	Precise orbit position so the altimeter's height is accurate
KBR, LRI	K-Band Ranging system; Laser Ranging Interferometer	GRACE; GRACE-FO (both)	Microwave and laser ranging between the two satellites	Measures changes in separation to within about 10 micrometers (KBR) or far better, nanometer level (LRI)	Gravity anomalies, hence mass change
OCO-2 spectrometer	Three grating spectrometers	OCO-2, OCO-3	Passive, reflected sunlight	O₂ A-band (0.765 µm), weak CO₂ band (1.61 µm), strong CO₂ band (2.06 µm); 10.6 km swath	XCO₂ (column-average CO₂ in ppm); solar-induced fluorescence from the O₂ A-band
TROPOMI	TROPOspheric Monitoring Instrument	Sentinel-5P	Passive pushbroom imaging spectrometer; Dutch and ESA	UV, visible, NIR (O₂ A-band) and SWIR (2.3 µm) spectrometers; 5.5 × 3.5 km pixels (7 × 3.5 km before August 2019); 2,600 km swath; daily	NO₂, O₃, SO₂, CO, CH₄, formaldehyde, aerosol index, clouds
PHyTIR	Prototype HyspIRI Thermal Infrared Radiometer	ECOSTRESS	Passive thermal whiskbroom (rotating mirror)	5 TIR bands (8 to 12.5 µm) plus 1.6 µm; 38 × 69 m pixels (usually quoted as about 70 m); 384 km swath	Land surface temperature, evapotranspiration, plant water stress
SRTM radars	C-band and X-band SAR	Space Shuttle Endeavour	Active single-pass interferometry with two antennas 60 m apart	C band 5.3 GHz (5.6 cm, 225 km swath); X band 9.6 GHz (3.1 cm, 50 km swath)	Near-global DEM
AVHRR	Advanced Very High Resolution Radiometer	NOAA polar satellites since 1978; MetOp	Passive whiskbroom	5 or 6 channels, 1.1 km, 2,900 km swath	The longest daily global record: NDVI since 1981, SST, snow, clouds
SeaWiFS	Sea-viewing Wide Field-of-view Sensor	OrbView-2 (1997 to 2010)	Passive	8 bands, 402 to 885 nm, 1.1 km	Ocean chlorophyll
POLDER	POLarization and Directionality of the Earth's Reflectances	PARASOL (and ADEOS)	Passive wide-field polarimeter	Multiple angles and polarizations	Aerosol and cloud properties
GLAS, ATLAS	Geoscience Laser Altimeter System; Advanced Topographic Laser Altimeter System	ICESat; ICESat-2	Active lidar	GLAS 1,064 and 532 nm; ATLAS 532 nm, 6 beams, 10,000 pulses per second, photon counting	Ice, sea-ice and canopy height
SIRAL	SAR Interferometric Radar Altimeter	CryoSat-2	Active Ku-band altimeter	SAR mode sharpens the along-track footprint to about 300 m	Sea-ice freeboard and thickness; ice-sheet edges
KaRIn	Ka-band Radar Interferometer	SWOT	Active, two antennas on a 10 m boom	35.75 GHz, two 50 km swaths	Water surface height in 2D

Which satellite for which job

This is the pick-one table. When a question asks which mission is "best suited", match the quantity first, then check the orbit (continuous watching needs GEO; global detail needs a polar LEO).

Acronyms

Tests often ask you to write an acronym out in full. The table is alphabetical.

ReferenceRemote sensing acronyms· 101 rows

101 rows

Acronym	Stands for
ABI	Advanced Baseline Imager
AIRS	Atmospheric Infrared Sounder
AMR	Advanced Microwave Radiometer
AMSR-E	Advanced Microwave Scanning Radiometer for EOS
AMSU	Advanced Microwave Sounding Unit
AOD	Aerosol optical depth
ARVI	Atmospherically Resistant Vegetation Index
ASTER	Advanced Spaceborne Thermal Emission and Reflection Radiometer
ATLAS	Advanced Topographic Laser Altimeter System
ATMS	Advanced Technology Microwave Sounder
AVHRR	Advanced Very High Resolution Radiometer
A-Train	Afternoon Train (the Afternoon Constellation)
CALIOP	Cloud-Aerosol Lidar with Orthogonal Polarization
CALIPSO	Cloud-Aerosol Lidar and Infrared Pathfinder Satellite Observations
CCD	Charge-coupled device
CERES	Clouds and the Earth's Radiant Energy System
CMOS	Complementary metal-oxide-semiconductor
CONUS	Conterminous (contiguous) United States
CPR	Cloud Profiling Radar
CrIS	Cross-track Infrared Sounder
DEM	Digital elevation model
DN	Digital number
DORIS	Doppler Orbitography and Radiopositioning Integrated by Satellite
DPR	Dual-frequency Precipitation Radar
ECOSTRESS	ECOsystem Spaceborne Thermal Radiometer Experiment on Space Station
EM, EMR	Electromagnetic, electromagnetic radiation
EOS	Earth Observing System
ERTS	Earth Resources Technology Satellite (Landsat 1's original name)
ESA	European Space Agency
ETM+	Enhanced Thematic Mapper Plus
EVI	Enhanced Vegetation Index
FOV	Field of view
GCOM-W	Global Change Observation Mission – Water
GEO	Geostationary Earth orbit
GIS	Geographic information system
GLM	Geostationary Lightning Mapper
GMI	GPM Microwave Imager
GOES	Geostationary Operational Environmental Satellite
GPM	Global Precipitation Measurement
GPS	Global Positioning System
GRACE (-FO)	Gravity Recovery and Climate Experiment (Follow-On)
GSD	Ground sample distance
HEO	High Earth orbit (also used for highly elliptical orbit)
HIRDLS	High Resolution Dynamics Limb Sounder
HSB	Humidity Sounder for Brazil
ICESat	Ice, Cloud and land Elevation Satellite
IFOV	Instantaneous field of view
InSAR	Interferometric synthetic aperture radar
ISS	International Space Station
JAXA	Japan Aerospace Exploration Agency
JASON	Named after Jason of the Argonauts (echoing the French JASO altimetry workshops); sometimes expanded as Joint Altimetry Satellite Oceanography Network
JPSS	Joint Polar Satellite System
LEO, MEO	Low Earth orbit, medium Earth orbit
LIDAR	Light detection and ranging
LST	Land surface temperature
MISR	Multi-angle Imaging SpectroRadiometer
MLS	Microwave Limb Sounder
MODIS	Moderate Resolution Imaging Spectroradiometer
MOPITT	Measurements Of Pollution In The Troposphere
MSI	Multispectral Instrument
MSS	Multispectral Scanner
NBR	Normalized Burn Ratio
NDVI	Normalized Difference Vegetation Index
NDWI	Normalized Difference Water Index
NIR, SWIR, MWIR, TIR	Near-, shortwave-, midwave-, thermal-infrared
NLCD	National Land Cover Database
NOAA	National Oceanic and Atmospheric Administration
NPP	National Polar-orbiting Partnership (Suomi NPP)
OCI	Ocean Color Instrument
OCO	Orbiting Carbon Observatory
OLI	Operational Land Imager
OLR	Outgoing longwave radiation
OMI	Ozone Monitoring Instrument
OMPS	Ozone Mapping and Profiler Suite
PACE	Plankton, Aerosol, Cloud, ocean Ecosystem
PARASOL	Polarization and Anisotropy of Reflectances for Atmospheric Sciences coupled with Observations from a Lidar
POES	Polar-orbiting Operational Environmental Satellite
POLDER	POLarization and Directionality of the Earth's Reflectances
RADAR	Radio detection and ranging
RGB	Red, green, blue
SAR	Synthetic aperture radar
SAVI	Soil-Adjusted Vegetation Index
SeaWiFS	Sea-viewing Wide Field-of-view Sensor
SLC	Scan line corrector (Landsat 7)
SMAP	Soil Moisture Active Passive
SNR	Signal-to-noise ratio
SONAR	Sound navigation and ranging (acoustic, not electromagnetic)
SRTM	Shuttle Radar Topography Mission
SST	Sea-surface temperature
SWOT	Surface Water and Ocean Topography
TES	Tropospheric Emission Spectrometer
TIROS	Television Infrared Observation Satellite
TIRS	Thermal Infrared Sensor
TM	Thematic Mapper
TOA	Top of atmosphere
TOPEX	Topography Experiment (TOPEX/Poseidon)
TRMM	Tropical Rainfall Measuring Mission
TROPOMI	TROPOspheric Monitoring Instrument
UTM	Universal Transverse Mercator
VIIRS	Visible Infrared Imaging Radiometer Suite
VNIR	Visible and near-infrared

Reading maps and images (Division B)

Interpretation is 65 to 75% of Division B points, so this chapter deserves the most practice. Read the map before you read the feature. A map's marginal information (title, location, date, scale, orientation or north arrow, legend and labels) controls every measurement you make on it. Check the scale, the units, the date and which way is up before you measure; a "change" question is meaningless until you know the two dates.

Marginal information and symbology

Marginal information is everything printed around the map body that tells you how to read it. On a USGS 7.5-minute topographic quadrangle you will find:

Symbology is the visual language of the map: point symbols (benchmarks, buildings), line symbols (roads, streams, contours, boundaries) and area symbols (woodland tint, water, urban tint). USGS topographic maps use colors consistently:

On satellite-derived maps the legend does the same job: it tells you whether bright means hot, wet, high, rough or dense vegetation, and it gives the classes of a classified map. Never interpret a colored map without its legend.

Elements of image interpretation

Analysts identify features on an image with a standard set of clues. Tests ask you to name them or to say which one a description uses.

Tone is usually the first clue and texture separates targets of similar tone. Nadir is the point on the ground directly below the sensor; features at nadir are seen straight down and suffer the least distortion.

Related display and map vocabulary: pseudocolor (a single band or index displayed with a color ramp), density slicing (ranges of brightness values assigned to discrete colors or classes), choropleth map (areas such as counties shaded by a data value) and geometric accuracy (how faithfully an image or map shows the true position and shape of features; geometric resolution usually just means spatial resolution). Analysts also separate spatial variation (arrangement across the scene, which gives shape, size and texture), spectral variation (differences in tone or color between bands) and temporal variation (change between dates).

Display types

The 2027 list names the displays you must read. Each answers a different question.

A composite image combines several bands into one picture by assigning one band to each of the red, green and blue display channels. A true-color (natural-color) composite uses the red, green and blue bands; it does not use a "yellow" band, a common false statement on tests.

Scale, distance, perimeter and area

A representative fraction scale such as 1:24,000 means one unit on the map equals 24,000 of the same unit on the ground. A verbal scale says it in words ("1 inch equals 2,000 feet"). A bar scale gives the same information graphically and stays correct if the map is enlarged or shrunk. Ground distance is map distance times the scale denominator, and a perimeter is the sum of such distances around a feature. Area is the number of pixels or grid squares times the area each one covers; for a satellite image, one pixel covers the square of its ground size.

Large versus small scale. A larger fraction means a larger scale. 1:1,000 (one thousandth) is larger than 1:100,000 (one hundred-thousandth), so it shows a smaller area in more detail. Ranking scales from smallest to largest therefore means ranking denominators from largest to smallest. A scale of 1:1 is life size.

Useful conversions: 1 km = 1,000 m = 100,000 cm; 1 mi = 5,280 ft = 63,360 in = 1.609 km; 1 hectare (ha) = 10,000 m² (a 100 m square); 1 km² = 100 ha; 1 acre ≈ 4,047 m² ≈ 0.405 ha; 1 mi² = 640 acres ≈ 2.59 km².

Air photos. Photo scale ($f/H$) and relief displacement are worked in Photographic scale; an orthophoto has that displacement and tilt removed, so it has uniform scale and can be measured like a map. Overlapping photos (about 60% overlap along the flight line) viewed through a stereoscope give a three-dimensional view, because each eye sees the scene from a slightly different position (parallax).

Elevation, altitude and slope

Elevation is the height of the ground above sea level; altitude usually refers to the height of something above the ground or sea level, such as an aircraft or satellite. Contour lines connect points of equal elevation, and the contour interval is the elevation step between adjacent lines. It is constant on one map so that spacing can be read as steepness everywhere. Closely spaced contours mean a steep slope; widely spaced ones mean gentle ground. Elevation between two contours is estimated by interpolation.

Rules for reading contours:

• Contours never cross or split (except at an overhanging cliff) because a point has only one elevation.
• Where a contour crosses a stream it bends into a V (or U) that points upstream, toward higher ground.
• A set of closed loops marks a hill, with the highest loop at the top. Closed loops with short tick marks (hachures) pointing inward mark a depression.
• Every fifth contour is a thicker, labeled index contour; the lines between are intermediate contours.
• Evenly spaced contours mean a uniform slope; contours that crowd together near the bottom of a hill mean a convex profile (slope steepening downhill); contours crowded near the top mean a concave profile.
• Spot elevations and benchmarks (BM, surveyed and marked on the ground) give exact heights; relief is the difference between the highest and lowest points.

A topographic profile is a side view along a line. Lay a paper strip along the line, tick where each contour crosses it and note the elevations, then plot each tick at its height on a graph whose horizontal axis uses the map's scale. Vertical exaggeration is the ratio of the horizontal scale denominator to the vertical scale denominator: if the profile's horizontal scale is 1:24,000 and its vertical scale is 1 cm = 100 m (1:10,000), the vertical exaggeration is $24{,}000/10{,}000 = 2.4\times$. Exaggeration makes gentle relief visible but steepens every slope's appearance.

A digital elevation model (DEM) stores elevation as a grid of pixel values. Displayed in grayscale, low ground is usually dark and high ground bright; a hillshade simulates sunlight from one direction to show texture; color ramps (green to brown to white) show height bands. Slope and aspect maps are calculated from a DEM. SRTM (radar interferometry, 2000) and ASTER GDEM (optical stereo) are the Division B elevation sources; see Elevation and subsidence change for change over time.

Map projections and coordinate grids

Every flat map of the curved Earth distorts something: shape, area, distance or direction. A projection is the rule for flattening. Projection and grid vocabulary appears in map-reading questions; Division B does not calculate bearings or coordinate conversions.

Projections are also grouped by the surface the globe is projected onto: cylindrical (Mercator; true at the equator, area hugely exaggerated toward the poles, so Greenland looks as large as Africa), conic (good for mid-latitude countries such as the United States) and planar or azimuthal (good for polar regions).

Latitude is measured north or south from the equator (0° to 90°); longitude east or west from the prime meridian through Greenwich (0° to 180°). One degree of latitude is about 111 km everywhere. The Universal Transverse Mercator (UTM) grid divides the world between 80° S and 84° N into 60 north-south zones, each 6° of longitude wide, and gives positions in meters (eastings and northings) on a transverse Mercator (cylindrical) projection within each zone. Engineers and field scientists prefer UTM for small areas because distances come straight out in meters and distortion within a zone is small.

Two kinds of coordinate system. A geographic coordinate system gives positions as latitude and longitude in degrees on a model of the Earth (a datum such as WGS 84 or NAD 83). A projected coordinate system flattens that onto a plane with a projection and gives positions in linear units such as meters (UTM, state plane). Every projected system is built on a geographic one.

GIS basics. A geographic information system stores map data in layers. Raster data are grids of cells (satellite images, DEMs, classified maps); a raster's scale is set by its cell (pixel) size on the ground, such as 30 m, not by a representative fraction; vector data are points, lines and polygons (wells, roads, parcel boundaries). Georeferencing ties an image's pixels to map coordinates so layers line up.

A map is an orthographic (orthogonal) projection: every point is shown as if viewed from directly above, so tall objects do not lean and scale is uniform. An aerial photograph is a perspective (central) projection: all rays pass through the lens center, so the tops of tall objects are displaced outward from the photo center (relief displacement) and scale varies with terrain height. Orthorectification converts the photo into an orthographic view (an orthophoto).

A geographic information system (GIS) is a system of hardware, software, data and people for capturing, storing, managing, analyzing and displaying geographically referenced (location-tagged) data.

Classifying land cover

A land-cover map labels each pixel with a class such as water, forest, cropland, or built-up area. Land cover is what physically covers the surface (trees, grass, pavement, water); land use is what people do there (a park, a farm, a housing estate). A satellite sees cover directly and infers use. Two approaches dominate. In supervised classification, the analyst supplies training areas of known cover type, and the algorithm assigns every other pixel to the class it most resembles. In unsupervised classification, the algorithm first groups pixels with similar spectral signatures into clusters without any training data; the analyst then decides which land cover each cluster corresponds to. The difference is who supplies the labels: supervised methods need labeled examples before classification, and unsupervised methods need labels afterward.

Object-based classification first segments the image into patches of similar pixels (objects such as fields or roofs) and then classifies the objects using shape, size and texture as well as spectra. It suits fine-resolution imagery where single pixels are smaller than the objects of interest.

Accuracy assessment compares the classified map with reference ("ground truth") points in an error (confusion) matrix. Overall accuracy is the fraction of reference points classified correctly. Producer's accuracy for a class is correct ÷ reference total for that class; its complement is the omission error (real pixels of the class that were missed). User's accuracy is correct ÷ mapped total for that class; its complement is the commission error (pixels wrongly put in the class). The kappa coefficient adjusts overall accuracy for the agreement expected by chance.

Classification systems. The USGS Anderson system (1976) defines Level I classes (urban or built-up, agricultural, rangeland, forest, water, wetland, barren, tundra, perennial snow or ice) with finer Level II subclasses (for example deciduous, evergreen and mixed forest). Higher levels need finer resolution. The NLCD uses a modified Anderson Level II legend.

The National Land Cover Database (NLCD) is produced by the USGS with the Multi-Resolution Land Characteristics (MRLC) consortium of federal agencies, not by NASA. It classifies the conterminous United States (CONUS) at 30 m from Landsat imagery, with editions for 2001, 2004, 2006, 2008, 2011, 2013, 2016, 2019 and 2021 and, since 2024, an Annual NLCD series reaching back to 1985. Its 16 CONUS classes are:

(Alaska adds dwarf scrub 51, sedge 72, lichens 73 and moss 74.) The NLCD also maps percent impervious surface and tree-canopy cover. The NLCD Land Cover Change Index is a product that flags, for each 30 m pixel, whether its land cover changed at any point across the NLCD dates and what kind of change it was (for example to or from developed land, forest, agriculture, wetland, water or barren land). A pixel flagged as changed should mean a real change of cover, such as forest cleared for suburban housing, a new reservoir or a wetland drained for farmland. Differences that are not land-cover changes, such as a different sensor's radiometric resolution, seasonal greenness on the image date or haze, are not land-cover change and should not be read as change.

Land-cover change is measured by comparing classified maps from two dates (post-classification comparison), by subtracting an index or band between dates (image differencing, for example an NDVI difference map), or by analyzing a long time series of images to find when and where change happened. Describe the change the data show (forest to cropland, urban expansion) and remember the image shows a change in cover, not its cause. The 2027 rules group the human and environmental implications as their own item: habitat loss (forest or wetland pixels converted and fragmented into smaller patches), land-cover change in general, urbanization (impervious surface growing outward, hotter in thermal images, brighter in radar), deforestation (NDVI loss, often in fishbone patterns along new roads) and damage assessment and monitoring (before-and-after pairs after a flood, fire or earthquake, often from radar because it sees through cloud and smoke). A strong answer names the evidence in the image, the change it shows, and one consequence for people or ecosystems.

Division B platforms, bands and band combinations

The Division B list is short, so learn each platform's sensor type, bands, pixel size and coverage. ECOSTRESS gives fine-scale surface temperature (urban heat islands, crop water stress, lava flows), and GOES gives a near-continuous view of fires, smoke, volcanic ash and storms over the Americas. Figures below are standard published mission values; a test will give any value you need, but you should recognize them. Launch dates, agencies and full instrument specs are in the Satellites and instruments reference.

Combining Landsat 8, Landsat 9 and Sentinel-2 (the Harmonized Landsat Sentinel-2 product) gives a clear view of most places every two to three days, which helps catch short-lived changes such as floods, harvests and construction.

Landsat bands. Band numbers shifted when OLI added a coastal band, so learn both tables.

Sentinel-2 MSI bands (central wavelength, pixel size): 1 coastal aerosol 443 nm, 60 m; 2 blue 490 nm, 10 m; 3 green 560 nm, 10 m; 4 red 665 nm, 10 m; 5, 6, 7 red edge 705, 740, 783 nm, 20 m; 8 NIR 842 nm, 10 m; 8A narrow NIR 865 nm, 20 m; 9 water vapor 945 nm, 60 m; 10 cirrus 1,375 nm, 60 m; 11 SWIR 1,610 nm, 20 m; 12 SWIR 2,190 nm, 20 m. The red-edge bands sit on the steep rise in vegetation reflectance between red and NIR, so they are sensitive to chlorophyll content and early plant stress.

GOES ABI bands. ABI has no green band; true-color "GeoColor" images synthesize green from the other bands. Pixel sizes are at the point directly below the satellite and grow toward the edge of the disk. NOAA classes bands 3 to 6 as near-infrared; band 5 (1.61 µm, "snow/ice") separates ice and snow, which absorb strongly at 1.6 µm, from liquid-water clouds, which reflect it.

MODIS bands that come up: band 1 red (620 to 670 nm) and band 2 NIR (841 to 876 nm) at 250 m, used for NDVI; bands 3 to 7 (blue, green, NIR, two SWIR) at 500 m; bands 8 to 36 at 1 km for ocean color, atmosphere, clouds and thermal emission (bands 20 to 25 and 27 to 36 are thermal-emissive; band 31 near 11 µm is a standard surface-temperature band).

Spectral signatures are the key to using bands. Healthy vegetation absorbs blue and red for photosynthesis, reflects a little green (which is why leaves look green), then jumps to high reflectance in the NIR (the "red edge" near 700 nm), because the spongy mesophyll cells inside the leaf scatter NIR strongly. Its SWIR reflectance is lowered by leaf water, with absorption dips near 1.4 and 1.9 µm. Stressed or dying vegetation reflects more red (less chlorophyll) and less NIR (collapsing leaf structure), so its red edge flattens. The spectral symptoms are the same whatever the cause: water stress (drought), nutrient deficiency (especially nitrogen, which lowers chlorophyll), pests and insect outbreaks, disease and fungal infection, air pollution (ozone, sulfur dioxide), soil salinity, frost damage and natural senescence in autumn. Separating causes needs field checks or extra data: SWIR or thermal bands point to water stress, while patchy loss spreading through a forest suggests insects or disease. Water is dark everywhere, almost black in NIR and SWIR; clear water reflects a little blue and green, while sediment-laden water is brighter and browner and algae-rich water is greener. Bare soil reflectance rises smoothly with wavelength and drops when wet or rich in organic matter. Snow is bright in visible light but dark in SWIR, while cloud stays bright in SWIR, which is how the two are told apart. Pavement and roofs have flat, moderate signatures that depend on the material. Minerals such as clays and carbonates have narrow SWIR absorption features, which is why ASTER's SWIR bands were used for geology.

A band combination assigns three bands to the red, green and blue display channels. Band numbers shift between sensors: Landsat 8 and 9 add a coastal band at number 1, so TM and ETM+ band 3 (red) is OLI band 4. A standard false-color composite (color infrared) shows NIR as red, red as green and green as blue.

A composite that puts SWIR, NIR and red (or SWIR2, SWIR1 and NIR) in the display channels is used for vegetation moisture and stress, burn scars and also geology: rock and soil types differ most in the SWIR, so these combinations help map minerals and lithology.

Vegetation and spectral indices

The normalized difference vegetation index (NDVI) compares near-infrared and red reflectance. Healthy vegetation absorbs red light for photosynthesis and reflects near-infrared strongly, so it has high NIR and low red reflectance. NDVI is a ratio, so it ranges from $-1$ to $+1$ and partly cancels out differences in overall brightness.

The 2027 rules do not name NDVI, but they list vegetation change in forests and farmland, and NDVI is the standard way to measure it. Typical values:

A falling NIR and rising red reflectance over the years means vegetation is declining or stressed. Over one year, NDVI of a deciduous forest or a crop field rises in spring, peaks in summer and falls in autumn, a hump-shaped curve that tracks the growing season. A sudden regional drop means clearing, fire, drought, flood or harvest; check the date and a second index.

Limitations of NDVI. It saturates over dense, multilayered canopies (values plateau near 0.8 to 0.9 so further growth is invisible), it is sensitive to the brightness of the soil showing through sparse vegetation, it is changed by haze, aerosols and thin cloud (which raise red more than NIR), it depends on sun and view angles and on the sensor's exact bands (so values from different satellites are not identical), and a single value cannot tell crop type or the cause of stress. EVI (aerosol and soil corrections, less saturation) and SAVI (soil adjustment) were designed to reduce these problems.

Other indices. Each pairs a band where the target is bright with one where it is dark.

Band numbers for the index inputs differ by sensor:

Geohazards and what the image shows

The 2027 rules name four geohazards for Division B, and each leaves a different signature.

• Volcanoes: hot lava and vents are thermal-infrared (and SWIR) hot spots, even through thin cloud and at night; ash and sulfur dioxide plumes are visible in UV and thermal bands (GOES tracks ash for aviation, Sentinel-5P and Aura map SO₂); ground swelling or sinking before an eruption shows in radar interferometry; fresh lava flows appear as dark, rough, unvegetated tongues and can be outlined by comparing NDVI and thermal images over time.
• Landslides: a bare scar replaces vegetation (NDVI drops, the scar is brighter or darker than forest in visible bands depending on the exposed soil), a fan of debris spreads downslope, and comparing elevation models before and after shows lost height at the head and gained height at the toe. Steep slopes, soaked soil and cleared vegetation raise the risk.
• Flooding: smooth standing water reflects radar away from the sensor, so it is dark (low backscatter) in Sentinel-1 images, even under cloud and at night. In optical images water is dark in NIR and SWIR and positive in NDWI. Compare with a dry-season image. Flooded towns can look bright in radar where water and walls form corner reflectors. Images give flood extent; flood depth needs a DEM or ground data as well.
• Wildfires: active fire is hot in SWIR and especially in the midwave infrared (about 3.9 to 4 µm), the band used by NASA's FIRMS fire alerts from MODIS and VIIRS and by GOES. Fire algorithms compare that 4 µm band with a longwave thermal band near 11 µm: a fire pixel is far brighter at 4 µm than at 11 µm, while warm bare ground is bright in both, and clouds are screened with visible reflectance. NIR is not used to detect fire because flames emit little there compared with reflected sunlight; NIR is used afterward to map the burn scar. Smoke is visible in natural color (bluish-white, from scattering) but nearly transparent in SWIR, so SWIR combinations show the fire front through the smoke. A burn scar has low NIR and high SWIR, so NBR falls; dNBR measures severity. A single image is weak evidence; before-and-after pairs are strong.

Aerosols affect all of these. Smoke, dust and haze scatter sunlight into the sensor's view, adding a bright veil that changes measured reflectance, especially in blue. Longer wavelengths (NIR, SWIR, thermal) pass through haze better, and radar ignores it. Atmospheric correction removes the scattering and absorption of gases and aerosols before a surface property is estimated.

Damage assessment follows the same workflow for any disaster: acquire an image as close before the event as possible and one right after (radar if clouds or smoke persist), co-register them, then map change with an index difference, backscatter change or InSAR coherence loss, and finally check samples on the ground or in fine-resolution imagery. Damaged buildings often lose their bright double-bounce radar return; burned or flattened vegetation loses NIR.

The table summarizes the signal that identifies each common feature and the mistake tests try to provoke.

After heavy rain, clouds often hide new landslides from optical satellites; SAR (Sentinel-1) maps them anyway through amplitude change, loss of InSAR coherence where the ground surface was disturbed, and before-and-after elevation differences.

Surface properties: albedo, temperature and materials

Every material has a characteristic albedo, temperature behavior and spectral signature, and the 2027 rules ask you to use these to classify land, including engineered materials. Asphalt and dark roofs have low albedo, absorb most sunlight and run hot in thermal images on a summer afternoon; concrete and white roofs are brighter and cooler; vegetation stays cool through transpiration (evaporating water uses energy); water heats slowly, so it is cool by day and relatively warm at night, and a river or lake cools the land next to it in summer. These differences produce the urban heat island seen in Landsat thermal bands, ASTER and ECOSTRESS: cities are warmer than the countryside, by day at the surface and especially at night in the air, because pavement and buildings store heat and there is less vegetation and evaporation. Fresh snow is the brightest natural surface, fresh lava and active fires are the hottest, and wet soil is darker and cooler than dry soil of the same type.

Albedo is the fraction of incoming sunlight a surface reflects (0 to 1). Approximate values; they vary with sun angle, wetness, age and season:

Full table with clouds, sea ice and soils: Albedo.

The ranking that tests ask most often is fresh snow > desert sand > grass > conifer forest > ocean, with asphalt near the bottom; snow next to asphalt gives the greatest contrast. Thermal inertia is a material's resistance to temperature change: water and wet soil have high thermal inertia (small day-night swing), dry sand and rock low (large swing). Emissivity (0 to 1) is how efficiently a surface radiates compared with a blackbody: water and vegetation are about 0.96 to 0.99, bare metal roofs much lower, which is why a thermal sensor needs an emissivity map to turn radiance into true temperature.

Combining three measurements classifies land quickly. Dense vegetation: low albedo, low surface temperature, high NDVI (rural forest or farmland). Downtown: moderate albedo, high brightness temperature, low NDVI. Suburbs: moderate in all three. Snow or tundra: high albedo, low temperature, low NDVI. Desert: high albedo, high daytime temperature, low NDVI.

Elevation and subsidence change

Ground that rises or sinks shows up in three ways: a difference between two DEMs (meters of change, as after a landslide or volcanic eruption), radar altimetry heights from two dates, or InSAR fringes (centimeters to millimeters of change, as over pumped aquifers, mines and swelling volcanoes). Subsidence over a city is usually slow and spread over many square kilometers, with a bowl of concentric fringes; an earthquake gives an abrupt pattern bounded by a fault. Millimeter-per-year subsidence of single buildings is found by stacking many Sentinel-1 radar dates (persistent-scatterer InSAR), because buildings act as stable, bright reflectors. Lidar surveys from two dates also measure subsidence.

Earth processes: glaciers, rivers and stream patterns

Glacial movement is tracked by matching features (crevasses, debris bands) between two images and measuring how far they moved, by InSAR, and by DEM differencing that shows thinning. Retreat shows as a terminus farther up-valley and bare rock or new lakes where ice used to be. Glacial landscapes have U-shaped valleys, cirques, horns, arêtes and moraines (ridges of debris at the sides, middle and end of a glacier) visible in hillshaded DEMs. A glacier's mass balance is accumulation (snowfall, avalanching and refreezing that add mass, mostly in the high accumulation zone) minus ablation (melting, sublimation and calving that remove it, mostly in the low ablation zone). The equilibrium-line altitude separates the two and appears on late-summer images as the snowline. Fresh snow cover also insulates and brightens the ice surface (high albedo), protecting it from melt; once it melts away, darker bare ice absorbs more sunlight and melts faster. A glacier with negative mass balance thins and retreats even if it keeps flowing downhill.

Fluvial processes are what rivers do: erode the outside of bends, deposit on the inside (point bars), cut off meanders into oxbow lakes, build deltas and alluvial fans, and carry sediment that shows up as brown or tan plumes in natural color. Comparing two dates shows channel migration. Rivers flow downhill from high contours to low, away from the point of a contour V (the V points upstream).

Stream (drainage) patterns reveal the rock and structure beneath.

Climate processes and satellite data (Division C)

Division C uses satellites to study how the atmosphere, oceans and land store and move energy, carbon and water. Most archive questions in this area ask what an instrument measures and what a measured pattern means: about 35% of the Division C score is interpretation of imagery and data, and another 20% is climate processes. The 2027 rules list specific platforms for each data topic, so learn what each one contributes, which topic it is listed under, and how to read its maps. Launch dates, orbits and acronym expansions are in the Satellites and instruments reference; this chapter is about the physics behind each data product and how to interpret it.

Reading climate data products

Interpretation questions show a map, an image, a profile or a time series and ask what it shows and why. Work in the same order every time: read the title, units, date and color bar; decide whether the map shows an absolute value or an anomaly (the departure from the long-term average for that place and season); find the extreme regions; then name the physical process that produces them. A strong answer cites the evidence ("values above +2 °C along the equator east of 150° W") and the mechanism ("weakened trade winds let warm water spread east").

Cloud imagery: visible, infrared and water vapor

Weather satellites such as GOES show clouds three ways, and each answers a different question.

The infrared rule follows from Stefan-Boltzmann: a cloud top at 210 K emits $\sigma(210)^4 \approx 110$ W/m², while a 300 K ocean emits about 460 W/m², so on an infrared image the warm surface looks dark and the cold cloud top bright. Fog and low stratus are hard to see in infrared at night because their tops are almost as warm as the ground; forecasters subtract a 3.9 µm image from an 11 µm image to find them. A thunderstorm (cumulonimbus) is bright in both visible (thick, high albedo) and infrared (cold top); thin cirrus is bright in infrared but faint in visible; low stratocumulus is bright in visible but gray in infrared.

Rainfall from infrared. Early satellite rain estimates assumed that colder cloud tops mean taller, wetter storms: a pixel colder than a threshold (about 235 K in the classic GOES Precipitation Index) is counted as raining. Infrared works from geostationary orbit, so it gives frequent, fine-resolution coverage day and night. Its limits are that it measures the top of the cloud, not the rain: cold cirrus anvils that are not raining are counted, and warm shallow clouds that rain heavily (coastal and mountain rain) are missed. Microwave sensors (GPM's microwave imager, the DPR radar) sense the raindrops and ice themselves, because hydrometeors emit and scatter microwaves; they are more direct but coarser and fly in low orbit, so they pass less often. Modern products such as GPM's IMERG merge both.

Outgoing longwave radiation (OLR) maps show the infrared leaving the top of the atmosphere. Deep tropical convection (the ITCZ, the Amazon and Congo basins, the Indonesian warm pool, monsoons) appears as low OLR because the emission comes from cold cloud tops near the tropopause. The highest OLR is over clear subtropical oceans and deserts, where warm surfaces radiate through dry, cloud-free air. The poles have low OLR because the surface itself is cold. An OLR anomaly map is therefore a rainfall-anomaly map in disguise: negative OLR anomalies mean more convection than normal.

Hurricanes and severe storms

A tropical cyclone draws its energy from warm ocean water: it needs sea-surface temperatures of at least about 26.5 °C through a deep enough layer (roughly the top 50 m), moist mid-level air, low vertical wind shear (little change of wind with height) and enough Coriolis effect, so it forms at least about 5° from the equator. Its structure has a calm, clear eye (typically 20 to 65 km wide) of sinking air, a ring of tall thunderstorms in the eyewall with the strongest winds, heaviest rain and coldest infrared tops, spiral rainbands outside, and a warm core: the air in the center is warmer than its surroundings because of the sinking in the eye and the latent heat released in the eyewall. In the Northern Hemisphere the low-level winds spiral counterclockwise. Intensity is reported on the Saffir-Simpson scale of 1-minute sustained wind: Category 1 is 119 to 153 km/h (74 to 95 mph); Category 3 and above (178 km/h, 111 mph) is a "major" hurricane; Category 5 is 252 km/h (157 mph) or more.

How each instrument sees a hurricane:

• GOES visible and infrared: a round central dense overcast, a clear or warm eye (a warm spot surrounded by a ring of very cold tops signals an intense storm), and a fan of high cirrus outflow. Infrared enhancement curves and the Dvorak technique turn this pattern into an intensity estimate. Rapid scans every minute show eyewall changes.
• Microwave (GPM GMI) and radar (GPM DPR, TRMM before it): see through the cirrus canopy to the rain itself: the eyewall ring, rainbands, and in 3D the height of the rain and the bright band, a layer of strong radar reflectivity at the melting level (about the 0 °C height) where wet snowflakes look like giant raindrops.
• SST (MODIS, GOES) and sea-surface height (Jason): warm water and positive height anomalies mark deep warm layers (ocean heat content) that let a storm intensify; a cold wake of upwelled water trails behind the storm.
• Scatterometers, Doppler radar and SAR: wind speed and direction; Doppler radar measures the radial velocity of raindrops, so inbound and outbound velocities side by side reveal rotation.
• Flood mapping after landfall: Sentinel-1 radar (Division B) or optical imagery once skies clear.

Vertical curtains: CloudSat, CALIPSO and the A-Train

CloudSat and CALIPSO look straight down and record a profile at each point along the orbit. Stacked side by side, the profiles make a curtain: horizontal axis = distance (or latitude) along the ground track, vertical axis = height (usually 0 to 20 or 30 km), color = signal strength.

CALIPSO's Vertical Feature Mask (VFM) is a classified curtain: each cell is labeled clear air, cloud, aerosol, stratospheric feature, surface, or "totally attenuated" (no signal). The aerosol subtype product labels dust, smoke, marine, polluted continental and so on. A high, thin layer with strong return at 8 to 12 km and weak return below is a cirrus or volcanic aerosol layer; a layer at 2 to 5 km off West Africa with high depolarization is Saharan dust.

The two satellites flew together in the A-Train (see the Satellites and instruments reference) so that one scene could be measured by several instruments within minutes. Typical synergy question: MODIS gives the horizontal cloud picture, cloud fraction, top temperature and optical thickness; CALIPSO gives cloud-top height and thin layers; CloudSat gives the internal structure and precipitation of thick cloud; CERES gives the cloud's effect on the energy budget (more reflected shortwave, less outgoing longwave).

Precipitation: GPM

The GPM Core Observatory carries the Dual-frequency Precipitation Radar (DPR) at Ku band (13.6 GHz) and Ka band (35.5 GHz) and a 13-channel passive microwave imager (GMI). Two radar frequencies help because the shorter Ka wavelength is more sensitive to light rain and snow but is attenuated more by heavy rain; the difference between the two returns gives the drop size distribution, which turns reflectivity into a rain rate more accurately. The radar gives 3D structure along a narrow swath; the radiometer gives wider coverage. GPM's 65° inclination reaches higher latitudes than TRMM did (35°), so it measures snowfall. On a GPM rain map, expect the heaviest long-term rain in the ITCZ, the western Pacific warm pool and monsoon regions, and the least over subtropical deserts and the cold eastern sides of the subtropical oceans.

Gas maps: OCO-2, TROPOMI and Aura

OCO-2 measures reflected sunlight in a weak CO₂ band (1.61 µm), a strong CO₂ band (2.06 µm) and the oxygen A-band (0.76 µm). The oxygen band tells it how much air the light crossed, so the ratio gives XCO₂, the column-averaged dry-air mole fraction of CO₂, in ppm: the average over the whole column from the surface to space, not a surface value. Because it needs sunlight and clear sky, there are gaps under clouds and in polar winter, and over oceans it points at the sun's glint to get enough signal. Its maps show:

OCO-2 data are trusted only after validation. Its retrievals are compared with the Total Carbon Column Observing Network (TCCON), ground-based Fourier-transform spectrometers that look straight at the Sun and measure the same column CO₂; the instrument is calibrated on board with lamps and solar observations and vicariously over bright desert sites; and soundings contaminated by cloud or thick aerosol are screened out, using the O₂ A-band (a cloud shortens or lengthens the apparent light path) before XCO₂ is reported. Each orbit yields up to about 24 soundings per second along a 10 km wide track, but after screening only a fraction are usable.

• A global mean near 423 ppm (2024), rising about 2.5 ppm per year (NOAA's global growth rate for 2024 was 3.76 ppm, the largest on record).
• A seasonal cycle strongest in the Northern Hemisphere: XCO₂ peaks in late spring and falls to a minimum in late summer and early fall as northern forests and crops photosynthesize, then rises again through winter. The Southern Hemisphere cycle is small because it has much less land vegetation.
• Higher values over the industrial Northern Hemisphere (eastern China, the eastern United States, Europe, the Middle East) and over fire regions (African savannas in their dry season), and slightly lower values over strong summer forest sinks and some ocean areas.
• Solar-induced fluorescence (SIF): a faint red glow from chlorophyll, retrieved near 0.76 µm, that tracks photosynthesis directly.

TROPOMI on Sentinel-5P maps several gases daily at a few kilometers. Read each by its lifetime: a short-lived gas outlines its sources.

Aura monitors the ozone layer and air quality. Its OMI maps total ozone, NO₂, SO₂ and the UV aerosol index by measuring backscattered ultraviolet sunlight; its MLS measures microwave emission from the limb (looking sideways through the atmosphere) to give vertical profiles of ozone, chlorine monoxide (ClO), HCl, water vapor and temperature, the measurements that tie ozone loss to chlorine.

Aerosol maps: MODIS, PACE and CALIPSO

Aerosol optical depth (AOD, symbol $\tau$) is a unitless measure of how much the aerosols in the whole column dim a vertical beam by scattering and absorption, usually reported at 550 nm. The direct beam is reduced by a factor $e^{-\tau}$: AOD 0.1 removes about 10%, AOD 1 leaves 37%.

Map patterns: summer dust off the Sahara across the tropical Atlantic (the Saharan Air Layer); smoke over central Africa (June to September) and the Amazon (August to October); pollution over northern India and eastern China; wildfire smoke over Canada, Siberia and the western US; moderate sea-salt AOD over the stormy Southern Ocean. High AOD over a remote ocean does not mean surface pollution: AOD counts aerosol at all heights, including sea salt and long-range dust or smoke aloft. MODIS retrievals are hardest over bright deserts and snow, because the surface reflection is confused with aerosol scattering.

Instruments add complementary information. MODIS gives column AOD over wide swaths (passive). CALIPSO gives the height of the layer (active). PACE's polarimeters (HARP2 and SPEXone) and multi-angle views give particle size, shape, absorption and type, which matters because the type decides whether an aerosol cools or warms.

How MODIS gets aerosol optical depth. It compares measured top-of-atmosphere reflectance with lookup tables computed for many aerosol types and amounts, reporting AOD at 0.55 µm. The Dark Target method works over dark vegetated land using 0.47, 0.66 and 2.12 µm (aerosol is nearly transparent at 2.1 µm, so that band estimates the surface's own reflectance in the visible) and over the dark ocean using bands from 0.55 to 2.1 µm. The Deep Blue method uses 0.41 and 0.47 µm, where even deserts are dark, so it works over bright arid land.

Temperature: SST, land surface temperature and profiles

Sea-surface temperature (SST) from infrared is the temperature of the top fraction of a millimeter (the skin). Instruments use two or more thermal window channels (the split window at about 11 and 12 µm) because water vapor absorbs them slightly differently, which lets the algorithm correct for the atmosphere. Infrared SST fails under cloud because clouds are opaque at those wavelengths; passive microwave SST sees through cloud but has coarse pixels (tens of kilometers) and cannot be used near coasts. MODIS on Aqua gives global SST every one to two days; GOES gives near-continuous SST, day and night, for its hemisphere. An SST anomaly map is the standard way to identify El Niño (warm equatorial eastern Pacific) and marine heat waves, and SST above about 26.5 °C marks hurricane fuel.

Land surface temperature (LST) is the temperature of the ground or canopy surface, which can be far hotter than the air on a sunny afternoon. Converting radiance to LST needs the surface emissivity (vegetation about 0.98, pavement about 0.93, some bare soils as low as 0.86 in parts of the thermal band); assuming emissivity 1 makes low-emissivity surfaces look colder than they are. ECOSTRESS (about 70 m) sees individual fields and city blocks, and because the Space Station's orbit is not sun-synchronous it samples different times of day, revealing the daily heating cycle: urban heat islands, crops that warm when they run short of water (their transpiration stops cooling them), and lava flows. ASTER gives 90 m thermal images and emissivity for geology and volcanoes. Plants under water stress show high canopy temperature and low evapotranspiration; ECOSTRESS products such as the Evaporative Stress Index compare actual to potential evapotranspiration.

CrIS is a hyperspectral infrared sounder with more than a thousand channels between about 3.9 and 15.4 µm. Channels near the center of a CO₂ absorption band see only the high, cold atmosphere; channels on the band's wings see deeper. Comparing them gives a temperature profile, and water-vapor channels give a humidity profile. These profiles feed weather forecasts and climate records, including the signature of greenhouse warming: a warming troposphere above a cooling stratosphere.

Vertical temperature structure. Temperature profiles from CrIS and radio occultation show four layers.

Passive microwave SST is lost within about 50 to 100 km of coasts because its footprint is tens of kilometers wide and mixes in radiance from land, which is much warmer in microwave brightness (land emissivity is high, water's is low). The rules list GOES, Suomi NPP (CrIS) and MODIS on Aqua for atmospheric and sea-surface temperature, plus ECOSTRESS and ASTER for land.

Energy flux maps: CERES

CERES measures three broad channels at the top of the atmosphere: total, shortwave (reflected sunlight, 0.3 to 5 µm) and an infrared window channel, from which it computes reflected shortwave (SW) and outgoing longwave (LW, OLR) flux in W/m². Albedo is reflected SW divided by incoming solar flux.

The net-radiation map explains the general circulation: the tropics gain energy and the high latitudes lose it, so the atmosphere and ocean must carry heat poleward. The global average of the net flux is Earth's energy imbalance, currently about +1 W/m² or a little more and rising, which CERES measures as a trend and ocean floats (Argo) confirm as rising ocean heat content. A CERES change of "less reflected SW with unchanged OLR" means the planet is gaining energy; "more OLR with unchanged absorbed SW" means it is losing energy.

Ocean color and phytoplankton: PACE and MODIS

Chlorophyll a absorbs blue and red light and reflects green, so water with more phytoplankton looks greener and less blue. Ocean-color sensors convert the ratio of blue to green reflectance into chlorophyll concentration, mapped on a logarithmic scale (about 0.01 to more than 10 mg/m³). High values appear in coastal waters, upwelling zones (off Peru, northwest and southwest Africa, California, along the equator in the Pacific), and high latitudes during the spring bloom; the lowest values are in the centers of the subtropical gyres, the "ocean deserts". PACE's Ocean Color Instrument is hyperspectral (ultraviolet to near-infrared at about 5 nm spectral resolution), so it can tell phytoplankton communities apart by their pigments, not just measure total chlorophyll. Coccolithophore blooms look milky turquoise because their chalk plates scatter light. Phytoplankton matter for climate because they perform about half of Earth's photosynthesis, feed the ocean food web and drive the biological carbon pump. Algal blooms fed by fertilizer runoff show up as bright green near river mouths and in lakes.

How deep a satellite sees. About 90% of the light leaving the water comes from above the penetration depth $z_{90} = 1/K_d$, where $K_d$ is the diffuse attenuation coefficient; at that depth downwelling light has fallen to $1/e$ (about 37%) of its surface value. In clear ocean water $z_{90}$ can be tens of meters, in turbid coastal water less than a meter. A subsurface chlorophyll maximum below this depth is invisible to the satellite, so surface-only estimates can underestimate the chlorophyll and productivity of the whole water column.

Greenhouse gases, aerosols and clouds

Greenhouse gases absorb outgoing infrared radiation and re-emit it in all directions, so part of it returns downward and the lower atmosphere stays warmer than it would otherwise be. They are nearly transparent to incoming sunlight, which is mostly visible and near-infrared, but absorb strongly in the thermal infrared where Earth emits. A molecule absorbs infrared if its vibration changes its dipole moment, which is why CO₂, H₂O, CH₄, N₂O and O₃ are greenhouse gases and N₂, O₂ and Ar (symmetric or single atoms) are not. The Division C list names the gases below.

(Nitrous oxide, N₂O, is a different gas: a long-lived greenhouse gas from fertilized soils, covered under the nitrogen cycle below.) The global warming potential (GWP) compares the heat a gas traps over a stated period with the same mass of CO₂. The standard time horizons are 20, 100 and 500 years (GWP-100 is the usual reporting value), and a short-lived gas such as methane has a much larger GWP-20 (about 80) than GWP-100 (about 27 to 30). GWP depends on three things: how strongly each molecule absorbs infrared, where its absorption bands fall (a gas absorbing inside the 8 to 14 µm atmospheric window, where little else absorbs, is far more effective than one absorbing where CO₂ and water vapor already saturate the spectrum) and how long it stays in the air. Five natural factors also move concentrations and climate: solar activity (an 11-year cycle changing total irradiance by about 0.1%), volcanic eruptions (sulfate aerosol cooled the planet by roughly 0.5 °C for a year or two after Mount Pinatubo in 1991), Earth's orbit (Milankovitch cycles over tens of thousands of years), the seasons (the Northern Hemisphere's growing season draws CO₂ down each summer, giving the sawtooth in the Mauna Loa record) and atmospheric and ocean circulation.

What the air is made of

Dry air is almost all nitrogen, oxygen and argon. Everything else is a trace gas, conventionally any gas other than those three (some sources define it as anything below 1% by volume, which makes argon a trace gas too; say which definition you use). CO₂ is a trace gas despite its importance.

Water vapor is left out of "dry air" because it varies from almost zero in polar and desert air to about 4% in humid tropical air; it is the most variable greenhouse gas in space and time and is not well mixed. Ozone near the ground is usually 20 to 60 ppb, below N₂O, and the CFCs and HFCs are measured in parts per trillion (ppt), so the usual abundance ranking of greenhouse gases is H₂O > CO₂ > CH₄ > N₂O > tropospheric O₃ > CFCs and HFCs. About half the mass of the atmosphere lies below about 5.5 km, and nearly all water vapor is in the troposphere, where all weather happens.

Contributions to the greenhouse effect. In the natural (total) greenhouse effect, water vapor is responsible for about half, clouds about a quarter, CO₂ about a fifth, and other gases the rest (a widely cited NASA GISS estimate). Other methods give water vapor anywhere from about 36% to 72%; Kiehl and Trenberth's often-quoted figure is about 60%, so pick "about 50 to 60%" if offered. In the human-caused change, CO₂ dominates, then methane. Keep the two rankings separate: "largest contributor to the greenhouse effect" is water vapor; "largest contributor to current warming" or "largest anthropogenic greenhouse gas" is CO₂.

Greenhouse and trace gas reference

Use this table for "which gas", "how long does it last", "how potent" and "which satellite" questions. Concentrations are 2024 global means from the WMO Greenhouse Gas Bulletin unless marked; GWP-100 values are from the IPCC Sixth Assessment Report (AR6, 2021), with the older AR5 value in brackets where tests often use it. A lifetime is the time for a pulse to decay to $1/e \approx 37\%$ of its starting amount.

ReferenceGreenhouse and trace gases· 18 rows

ppm = parts per million, ppb = per billion, ppt = per trillion (by number of molecules). "—" = not applicable.

18 rows

Gas	Formula	Now (2024)	Pre-industrial (1750)	Lifetime	GWP-100 AR6 [AR5]	Main sources	Main sinks	Absorbs at; mapped by
Water vapor	H₂O	0 to 4%, highly variable	Similar (rises about 7% per °C of warming)	About 9 days	— (a feedback, not a forcing agent)	Evaporation, transpiration	Condensation and precipitation	6.3 µm band, most of the far infrared beyond 20 µm, 2.7 µm; microwave lines at 22 and 183 GHz; CrIS, MODIS, GOES water-vapor bands
Carbon dioxide	CO₂	423.9 ppm global (Mauna Loa annual mean about 424.6 ppm)	About 278 to 280 ppm	No single value: half of a pulse is removed in decades, about 15 to 40% remains after 1,000 years	1 (reference gas)	Fossil fuels, cement, deforestation; natural respiration, decay, volcanoes (about 1% of human emissions)	Photosynthesis, ocean dissolution, rock weathering	15 µm (main), 4.3 µm; 1.6 and 2.06 µm used by OCO-2
Methane	CH₄	1,930 ppb	About 730 ppb	About 12 years (AR6 11.8)	27 non-fossil, 29.8 fossil [28]; about 80 over 20 years	Wetlands (largest natural), livestock (ruminants), rice, landfills, oil, gas and coal leaks, termites	Reaction with the hydroxyl radical (OH) in the troposphere (about 90%), soils, stratosphere	7.7 µm, 3.3 µm, SWIR 1.65 and 2.3 µm; TROPOMI
Nitrous oxide	N₂O	338.0 ppb	About 270 ppb	About 109 to 120 years	273 [265]	Fertilized agricultural soils (largest human source), manure, industry (nitric and adipic acid), burning; natural soils and oceans	Destroyed by UV in the stratosphere, producing NO that destroys ozone	7.8 µm, 17 µm, 4.5 µm
Ozone (troposphere)	O₃	About 20 to 60 ppb near the ground	Lower (by an estimated 30 to 50%)	Weeks	— (short-lived; total ozone forcing about +0.47 W/m²)	Not emitted: made from NOₓ + VOCs + sunlight	Photolysis, reaction, deposition	9.6 µm; UV and visible bands; Aura, TROPOMI
CFC-11	CCl₃F	About 0.22 ppb, falling	0	About 52 years	6,230 [4,660]	Foam blowing, refrigerants (banned)	UV photolysis in the stratosphere	8 to 12 µm window
CFC-12	CCl₂F₂	About 0.49 ppb, falling (most abundant CFC)	0	About 100 years	12,500 [10,200]	Refrigerants, spray cans (banned)	UV photolysis in the stratosphere	8 to 12 µm window
HCFC-22	CHClF₂	About 0.25 ppb	0	About 12 years	1,960 [1,760]	Air conditioning (being phased out)	OH	Window region
HFC-134a	CH₂FCF₃	About 0.13 ppb, rising	0	About 14 years	1,530 [1,300]	Car air conditioning, refrigerators	OH	Window region
HFC-23	CHF₃	About 0.04 ppb	0	About 230 years	14,600 [12,400]	By-product of HCFC-22 production	OH (slow)	Window region
PFC-14 (tetrafluoromethane)	CF₄	About 0.09 ppb	Small natural amount	About 50,000 years	7,380 [6,630]	Aluminum smelting, semiconductors	Almost none	Window region
Sulfur hexafluoride	SF₆	About 12 ppt, rising	0	About 850 to 3,200 years (estimates differ)	About 24,300 [23,500]: the most potent common greenhouse gas	Electrical switchgear leakage, magnesium production	Almost none (upper atmosphere)	10.5 µm
Nitrogen trifluoride	NF₃	About 3 ppt, rising	0	About 570 years	17,400 [16,100]	Flat-panel and semiconductor manufacturing	Almost none	Window region
Carbon monoxide	CO	About 50 to 150 ppb	Lower	About 1 to 2 months	Indirect only (uses up OH, so CH₄ lasts longer; makes CO₂ and ozone)	Incomplete combustion, wildfires, methane oxidation	OH	4.7 µm, 2.3 µm; TROPOMI, MOPITT
Nitrogen oxides	NO, NO₂ (NOₓ)	ppb in cities, ppt in remote air	Lower	Hours to a day	Indirect (makes ozone, nitrate aerosol)	Combustion (vehicles, power plants), lightning (main source in the upper troposphere), soils	Converted to nitric acid, rained out	UV-visible absorption near 400 to 465 nm; TROPOMI, OMI
Sulfur dioxide	SO₂	ppb near sources	Lower	About 1 to 2 days near the ground; weeks in the stratosphere before turning to sulfate	Indirect (sulfate aerosol cools)	Coal and oil burning, smelting, volcanoes	Oxidized to sulfate, rained out	UV bands near 310 to 320 nm, 7.3 µm; TROPOMI, OMI, ASTER
Ammonia	NH₃	ppb over farmland	Lower	Hours to days	Indirect (ammonium aerosol cools)	Fertilizer, livestock manure	Forms ammonium aerosol, deposition	10.5 µm; CrIS, IASI
Isoprene and other VOCs	C₅H₈ and others	ppb over forests	Similar	Hours	Indirect (ozone, secondary organic aerosol)	Broadleaf trees (isoprene is the most abundant non-methane hydrocarbon), solvents, fuels	OH	Formaldehyde as a TROPOMI proxy

Fluorinated gases that do not deplete ozone, because they contain no chlorine or bromine, are the four controlled under climate agreements: HFCs, PFCs, SF₆ and NF₃. They are made only by industry, so they are purely anthropogenic. CFCs and HCFCs contain chlorine and do both: they warm and they destroy ozone.

The Keeling curve. Charles David Keeling began continuous CO₂ measurements at the Mauna Loa Observatory in Hawaii in 1958, when the level was about 315 ppm; NOAA and the Scripps Institution continue them. The observatory sits high on a remote volcano in the middle of the Pacific, so it samples well-mixed air. Three features are tested. (1) The rise: from about 315 ppm to about 424 ppm (2024), roughly 50% above the pre-industrial 280 ppm; its monthly mean first topped 400 ppm in 2014 and its annual mean in 2015. (2) The acceleration: growth averaged about 0.8 ppm per year in the 1960s and about 2.4 ppm per year in 2011 to 2020, with a record jump of about 3.5 ppm from 2023 to 2024 (strong El Niño plus fires). (3) The sawtooth: about 6 to 7 ppm peak-to-trough at Mauna Loa, maximum in May (after winter decay with little photosynthesis) and minimum in September or October (after the summer growing season). The cycle is larger at high northern stations such as Point Barrow, Alaska, and is only about 1 ppm at the South Pole, because most land vegetation is in the Northern Hemisphere. Ice cores extend the record: CO₂ stayed between about 180 ppm (glacial periods) and 300 ppm (warm interglacials) for the past 800,000 years.

Why CO₂ is more evenly mixed than methane: CO₂ and methane both last long enough to mix around the globe (mixing between hemispheres takes about a year), but sources are concentrated in the Northern Hemisphere, so both are higher in the north: CO₂ by under 1% (a few ppm), methane by roughly 5% (about 100 ppb). Short-lived gases such as NO₂ and SO₂ never mix and stay near their sources.

Air quality connections. The US EPA's six criteria air pollutants are ground-level ozone, particulate matter (PM2.5 and PM10), carbon monoxide, lead, sulfur dioxide and nitrogen dioxide; CO₂ is not one of them. PM2.5 means particles with aerodynamic diameter of 2.5 µm or less. Acid rain forms when SO₂ and NOₓ are oxidized to sulfuric acid (H₂SO₄) and nitric acid (HNO₃); natural rain is already slightly acidic (pH about 5.6) from dissolved CO₂, and acid rain is typically pH 4.2 to 4.4.

Water vapor has no defined GWP: it stays in the air only about 9 days and its amount is controlled by temperature, so it acts as a feedback, not a forcing agent. In a ranking of GWP, water vapor goes last: H₂O < CO₂ (1) < CH₄ < N₂O < CFCs and HFCs < SF₆.

Methane sources in the United States. In the EPA's greenhouse-gas inventory, the largest human methane sources are oil and natural gas systems (leaks and venting from wells, pipelines and processing, together roughly a third), enteric fermentation in cattle and other livestock (roughly a quarter to 30%), landfills (roughly 15 to 20%), manure management (about 10%) and coal mining (about 6%). Globally, agriculture (livestock and rice) is the largest human source, followed by fossil fuels and waste.

GWP fine print. Methane's GWP-20 is about 84 in the IPCC's AR5 (86 including climate-carbon feedbacks) and about 81 in AR6 (about 80 for non-fossil and 83 for fossil methane); many keys use 84 to 86. Tropospheric ozone has no standard GWP: it is not emitted but formed from NOx, CO, VOCs and methane, and lasts only weeks, so its warming is credited to those precursors; if a ranking includes it anyway, the order is CO₂ < CH₄ < tropospheric O₃ < CFCs. Agriculture's direct greenhouse emissions are mainly CH₄ (livestock digestion, manure, rice paddies) and N₂O (fertilized soils); its CO₂ comes mostly from land clearing and fuel, which inventories count separately. When a water molecule absorbs an infrared photon it vibrates and rotates faster; the energy is re-emitted as infrared or shared with neighboring molecules as heat.

Ozone: shield above, pollutant below

About 90% of atmospheric ozone is in the stratosphere, concentrated between about 15 and 35 km (the ozone layer). The rest is in the troposphere. The same molecule plays opposite roles in the two places.

Stratospheric ozone forms and is destroyed by sunlight in the Chapman cycle:

1. $\text{O}_2 + h\nu\ (\lambda < 242\ \text{nm}) \rightarrow \text{O} + \text{O}$ (photolysis of oxygen)
2. $\text{O} + \text{O}_2 + \text{M} \rightarrow \text{O}_3 + \text{M}$ (M is any third molecule that carries away energy)
3. $\text{O}_3 + h\nu\ (\lambda < 320\ \text{nm}) \rightarrow \text{O}_2 + \text{O}$ (ozone absorbs UV and is split, warming the stratosphere)
4. $\text{O} + \text{O}_3 \rightarrow 2\,\text{O}_2$ (loss)

Reaction 3 is the shield. Ozone absorbs ultraviolet strongly in the Hartley band (about 200 to 310 nm) and more weakly in the Huggins bands (about 310 to 360 nm), and absorbs weakly in the visible Chappuis band (about 400 to 650 nm, peaking near 600 nm, with a weak tail to about 850 nm); it also has an infrared band at 9.6 µm, which makes it a greenhouse gas. Absorbed UV heats the stratosphere, which is why temperature increases with height there (a temperature inversion that suppresses vertical mixing).

Total ozone is measured in Dobson units (DU): 1 DU is a layer of pure ozone 0.01 mm thick at 0 °C and 1 atmosphere. The global average is about 300 DU, so all the ozone overhead would make a layer only about 3 mm thick. Ozone is made mostly over the tropics but carried poleward and downward by the slow Brewer-Dobson circulation, so totals are highest at middle and high latitudes in late winter and spring.

Catalytic destruction. Chlorine and bromine atoms released when UV breaks up CFCs and halons in the stratosphere destroy ozone without being used up: $\text{Cl} + \text{O}_3 \rightarrow \text{ClO} + \text{O}_2$, then $\text{ClO} + \text{O} \rightarrow \text{Cl} + \text{O}_2$, net $\text{O}_3 + \text{O} \rightarrow 2\,\text{O}_2$. One chlorine atom can destroy tens of thousands of ozone molecules before it is locked up. Most stratospheric chlorine is normally stored in harmless reservoir molecules, hydrogen chloride (HCl, formed when Cl reacts with methane) and chlorine nitrate (ClONO₂). Bromine (from halons and methyl bromide) is far more destructive per atom. N₂O is the main natural source of the nitrogen oxides that destroy ozone in the middle stratosphere, and is now the largest remaining emission of an ozone-depleting substance. The ozone depletion potential (ODP) of a substance compares its ozone destruction with that of the same mass of CFC-11, whose ODP is 1 by definition; HFCs have ODP 0.

The Antarctic ozone hole is the area where total ozone falls below 220 DU. It was reported by British Antarctic Survey scientists in 1985 and appears every southern spring, deepening in September and reaching its largest area in late September or early October, then filling in by December. The sequence:

1. In the polar winter darkness, a strong polar vortex of westerly winds isolates the air over Antarctica, and it becomes extremely cold.
2. Below about −78 °C, polar stratospheric clouds (PSCs) of nitric acid and ice form.
3. On the cloud particles, the reservoirs react: $\text{HCl} + \text{ClONO}_2 \rightarrow \text{Cl}_2 + \text{HNO}_3$. The nitric acid stays in the particles, which can fall out (denitrification), so chlorine cannot be locked back up.
4. When sunlight returns in spring, it splits Cl₂ into free chlorine atoms, which destroy ozone rapidly through the ClO dimer cycle.
5. As the stratosphere warms and the vortex breaks up in late spring, ozone-rich air mixes back in.

The Arctic has smaller, irregular losses because its stratosphere is warmer and its vortex weaker and more disturbed; warmer winters mean fewer PSCs and less destruction. Aura's MLS showed the inverse relationship directly: where ClO is high, ozone is low.

Policy. The Montreal Protocol (signed 1987, in force 1989, ratified by every UN member) phased out CFCs, halons and other ozone-depleting substances; HCFCs are being phased out as transitional replacements. Stratospheric chlorine peaked around the late 1990s and is declining, and the ozone layer is projected to return to 1980 levels around 2040 globally, about 2045 in the Arctic and about 2066 over Antarctica (2022 WMO/UNEP assessment). Because CFCs are also strong greenhouse gases, Montreal has done more to slow warming than any climate treaty. The Kigali Amendment (agreed 2016, in force 2019) adds a phase-down of HFCs, the ozone-safe but high-GWP replacements; ways to reduce HFC emissions include switching to low-GWP refrigerants (ammonia, CO₂, hydrocarbons, HFOs), fixing leaks and recovering refrigerant from old equipment. HFC concentrations and emissions rose quickly from the 1990s because HFCs replaced CFCs and HCFCs in refrigerators, air conditioners, foams and aerosol cans; HFC-134a is the most abundant. In the United States, the AIM Act (2020) set an 85% HFC production and use phase-down by 2036, matching Kigali. Global HFC emissions are still rising, but more slowly, as the phase-down starts.

Ozone and temperature. Ozone loss cools the stratosphere (less UV absorbed) and is a small negative forcing at the surface. Rising CO₂ also cools the stratosphere, because in the thin upper air extra CO₂ radiates more energy to space than it absorbs from below. A warming troposphere under a cooling stratosphere is a fingerprint of greenhouse forcing: a brighter Sun would warm both layers.

Ground-level ozone is a secondary pollutant: it is not emitted but forms when sunlight splits NO₂ ($\text{NO}_2 + h\nu \rightarrow \text{NO} + \text{O}$, then $\text{O} + \text{O}_2 \rightarrow \text{O}_3$) and volatile organic compounds (VOCs) keep converting NO back to NO₂ without using up ozone, so ozone builds up. It peaks on hot, sunny, stagnant summer afternoons downwind of cities. It is the main ingredient of photochemical smog, damages lungs and crops, and is a greenhouse gas (positive forcing).

Total ozone also has a seasonal cycle outside the ozone hole. It is set by sunlight, which changes with Earth's axial tilt and the Sun's angle, together with the Brewer-Dobson circulation that carries ozone from the tropical stratosphere, where most is made, toward the poles in winter and spring. Mid- and high-latitude totals therefore peak in late winter and spring and are lowest in autumn. CFCs were invented in 1928 by Thomas Midgley Jr. (for General Motors' Frigidaire division) as non-toxic, non-flammable refrigerants to replace ammonia, sulfur dioxide and methyl chloride, and were later used in aerosol sprays, foam blowing and solvents.

Aerosols

An aerosol is a suspension of fine solid particles or liquid droplets in a gas; in climate science the word usually means the particles themselves (about 0.01 to 10 µm). SO₂ is a gas, not an aerosol, until it is oxidized to sulfate particles.

Aerosols also affect rainfall. By reflecting and absorbing sunlight they dim the surface (surface dimming), which reduces evaporation and the energy that drives convection, so a heavily polluted region can get less rain. Northern Hemisphere sulfate pollution in the twentieth century cooled that hemisphere and is thought to have pushed the tropical rain belt (ITCZ) south, contributing to the Sahel drought of the 1970s and 1980s. Dust from desertified land then adds more aerosol, a feedback that can sustain drought.

Aerosols are short-lived in the lower atmosphere (about a week before rain removes them), so unlike CO₂ they are concentrated near their sources and downwind: their forcing is patchy, strongest over and downwind of industrial and burning regions, and much larger in the Northern Hemisphere. In the stratosphere, where there is no rain, volcanic sulfate lasts one to two years, which is why only eruptions that reach the stratosphere change global climate.

How aerosols change the energy budget:

• Direct effect (aerosol-radiation interaction): particles scatter sunlight back to space (cooling) or absorb it (warming the air layer, cooling the surface beneath). The balance depends on the single-scattering albedo (SSA), the fraction of extinction that is scattering rather than absorption: about 1 for sulfate, about 0.8 to 0.95 for smoke, and much lower for pure black carbon. It also depends on the aerosol optical depth (how much there is), on the brightness of what is underneath (an absorbing layer over a dark ocean cools at the top of the atmosphere but over a bright cloud or snow warms) and on the layer height (absorbing aerosol above clouds warms most).
• Indirect effects (aerosol-cloud interaction): aerosols are cloud condensation nuclei (CCN, typically about 0.1 to 1 µm). The Twomey effect: more CCN spread the same water over more, smaller droplets, which have more total surface area, so the cloud's optical depth and albedo rise and it cools more. The Albrecht (lifetime) effect: smaller droplets collide and merge less efficiently, so drizzle is suppressed and clouds last longer and cover more area, cooling further. Ship tracks, bright lines in low clouds behind ships, are the visible example.
• Semi-direct effect: absorbing black carbon heats the cloud layer and can evaporate clouds, a warming effect.

Overall, IPCC AR6 assesses human aerosols' effective radiative forcing at about −1.1 W/m² (1750 to 2019), masking roughly a third of greenhouse-gas forcing. Reducing sulfur emissions for health reasons therefore unmasks warming: the 2020 International Maritime Organization rule that cut the sulfur limit in ship fuel from 3.5% to 0.5% reduced ship tracks and is estimated to have added a small positive forcing. Statements such as "the 40% drop in SO₂ emissions has cooled the planet" are false. This aerosol masking, and the dimming of sunlight at the surface it caused in the mid-20th century, is called global dimming.

Examples that tests use: Mount Pinatubo (Philippines, June 1991) injected about 15 to 20 million tonnes of SO₂ into the stratosphere, and global temperature fell by roughly 0.5 °C in 1992; Tambora (Indonesia, 1815) caused the "Year Without a Summer" in 1816. A CALIPSO curtain months after an eruption shows a thin, high layer in the stratosphere. Stratospheric aerosol injection, a proposed geoengineering method, would imitate this: it would cool the troposphere and surface but warm the stratosphere, where the sulfate absorbs infrared and near-infrared. The CLAW hypothesis (1987) proposed a negative feedback: warmer, sunnier oceans grow more phytoplankton, which release DMS; DMS forms sulfate CCN; more CCN brighten marine clouds and cool the surface.

The direct radiative forcing of an aerosol at the top of the atmosphere is the change in net downward flux it causes: for sunlight, the reflected (upward) flux without aerosol minus the reflected flux with it. Computed for clear sky it is usually negative (cooling); computed with clouds present it is smaller, and it turns positive (warming) when absorbing smoke or soot lies above bright cloud, because the aerosol darkens a bright scene. Marine cloud brightening is a proposed geoengineering method that sprays sea-salt particles into low marine clouds to exploit the Twomey effect, making them brighter (higher albedo) to cool the surface; stratospheric aerosol injection would mimic a volcanic eruption instead.

Absorbing aerosols also change circulation. Dust and black carbon absorb sunlight and warm the layer they sit in (diabatic heating) while dimming the surface. Over South Asia, absorbing aerosol piled against the Himalayas and Tibetan Plateau in spring heats the air aloft, which draws in moist air and can strengthen and advance the early monsoon rains (the "elevated heat pump" hypothesis), even though the dimmed surface weakens convection later.

Separating dust with lidar. Lidar uses two numbers. The depolarization ratio is near zero for spherical particles (cloud droplets, sulfate, most smoke: below about 0.05 to 0.1) and high for irregular particles (dust about 0.3; ice crystals higher). The lidar ratio (extinction-to-backscatter, in sr) is about 40 to 55 sr for dust, about 20 sr for sea salt and about 50 to 80 sr for smoke and pollution. CALIPSO's algorithm uses depolarization to flag dust, and research retrievals use it to split a mixed layer's extinction into dust and non-dust parts.

Aerosols and clouds in more detail. Many small cloud-condensation nuclei make many small droplets that are slow to collide and rain out, so warm rain is suppressed and more water is carried above the freezing level; when it freezes it releases extra latent heat, which can invigorate deep convection, giving fewer but heavier downpours (the invigoration hypothesis). The Southern Ocean shows the clean extreme: it has few, mostly natural nuclei (sea salt and sulfate from phytoplankton DMS), so its clouds have fewer, larger droplets, lower albedo and more drizzle than polluted Northern Hemisphere clouds, plus abundant supercooled liquid water. Cloud seeding for rain or snow uses silver iodide (sometimes dry ice or salt) as artificial nuclei.

Aerosol optical depth and visibility are linked: AOD below about 0.1 means very clear air (visibility tens of kilometers or more), about 0.3 to 0.5 noticeable haze, above about 1 thick smoke or dust, and values of 3 to 5 or more occur in dense wildfire smoke or dust storms, when visibility falls to about a kilometer or less.

Clouds and their radiative effects

At any time about two-thirds of Earth is covered by cloud. Clouds are named by height and form.

The mechanism behind the cooling-versus-warming split is two numbers: the cloud's albedo, which rises with thickness, liquid water and smaller droplets, and its top temperature, which falls with height. A low cloud has a high albedo and a warm top, so it reflects a lot of sunlight but barely reduces outgoing longwave (stratocumulus decks over cool subtropical oceans are the strongest cooling clouds on Earth). A high thin cirrus has a low albedo but a cold top, so it emits much less infrared to space than the warm surface below would have, a net warming. Cirrus and contrails, made by aircraft exhaust water vapor freezing in cold, humid air, are the best-known warming clouds.

The cloud radiative effect (CRE, also called cloud forcing) is the difference between all-sky and clear-sky top-of-atmosphere flux. CERES measures it globally at about −45 to −47 W/m² in the shortwave (cooling by reflection) and about +26 to +28 W/m² in the longwave (warming by trapping), a net of about −20 W/m²: clouds as a whole cool the planet. Clouds reflect about 20 to 23% of incoming sunlight, about two-thirds of Earth's total albedo of about 0.30. Clouds also narrow the daily temperature range: cloudy days are cooler (less sunlight in) and cloudy nights warmer (more infrared returned), so a sky cloudy both day and night gives the smallest range.

Cloud feedback asks how clouds change as the planet warms. AR6 assesses it as positive (amplifying), about +0.4 W/m² per K, with the largest uncertainty of any feedback: high clouds rise and their tops stay cold while the surface warms (positive), low clouds over the subtropical oceans thin and shrink (positive), while clouds at high latitudes contain more liquid and become brighter (negative). Higher humidity over oceans and the large contrast between dark ocean and bright cloud make the shortwave cloud effect larger over oceans than over land.

Cloud names are built from Latin roots: cirro- (curl or wisp of hair: high ice clouds), strato- (layer or sheet), cumulo- (heap or pile), nimbo-/-nimbus (rain-bearing) and alto- (middle level, despite meaning "high"). So nimbostratus is a rain-bearing layer and cumulonimbus a rain-bearing heap.

Nearly all clouds and weather are in the troposphere. The exceptions are polar stratospheric clouds (in the winter polar stratosphere, about 15 to 25 km, which drive ozone-hole chemistry) and noctilucent (polar mesospheric) clouds of ice near 80 to 85 km in the summer polar mesosphere.

Climate drivers and circulation

A climate forcing is an imposed change in Earth's energy budget from outside the climate system's own internal dynamics: changes in the Sun, Earth's orbit, volcanic aerosols, greenhouse gases, human aerosols and land-surface albedo. Internal variability, such as ENSO and the North Atlantic Oscillation, moves energy around within the system without an outside cause. The climate system has five components: atmosphere, hydrosphere (oceans, lakes, rivers, groundwater), cryosphere (all frozen water: sea ice, glaciers, ice sheets, snow cover, lake and river ice, and permafrost; it matters because ice and snow have high albedo and store fresh water), lithosphere (land surface) and biosphere. Weather is the state of the atmosphere at a given time and place; climate is the long-term statistics (usually 30-year averages) of weather. An anomaly is the departure from the average for the same place and time of year.

Internal or external? External forcings: solar variability, Milankovitch orbital cycles (which pace the ice ages, so glacial-interglacial cycles count as externally forced, amplified by internal CO₂ and ice-albedo feedbacks), volcanic eruptions, and human greenhouse gases, aerosols and land-use change. Internal variability: ENSO, the North Atlantic Oscillation, the Pacific Decadal Oscillation, monsoon variability and weather noise. A La Niña year is cooler because of internal variability, not a negative forcing.

Climate models and parameterization. Global climate models divide the atmosphere and ocean into grid cells tens to hundreds of kilometers wide and step the equations of fluid motion, energy and moisture forward in time. Processes smaller than a grid cell (individual clouds and convection, turbulence, radiation details, aerosol-cloud interactions) cannot be resolved, so they are parameterized: represented by simplified formulas tied to grid-scale variables. Cloud and convection parameterizations are the largest source of spread among models' climate sensitivity, which is why satellite cloud data (CloudSat, CALIPSO, MODIS, CERES) are used to test models.

Milankovitch cycles are the periodic changes in Earth's orbit and axis, named for Milutin Milanković.

Larger tilt means hotter summers and colder winters at high latitudes. Ice ages begin when northern summers are cool enough for winter snow to survive (low tilt, Northern summer at aphelion); feedbacks from ice albedo and from CO₂ released or absorbed by the ocean amplify the small orbital push. The last 800,000 years have had glacial cycles about 100,000 years long.

Seasons come from the 23.44° tilt, not from distance to the Sun: Earth is actually closest to the Sun (about 147 million km) in early January and farthest (about 152 million km) in early July, which changes the solar constant by only about 7%. Tilt changes both the angle of sunlight (low Sun spreads the same beam over more ground and through a longer path of air) and the length of day. That is also why the tropics receive more energy than the poles: flux on a horizontal surface is $S\cos\theta$, where $\theta$ is the solar zenith angle. The Southern Hemisphere has the lowest albedo in its summer, when snow cover is least and sea ice smallest.

Atmospheric circulation carries heat from the tropical energy surplus toward the poles.

• Hadley cells (equator to about 30°): air rises at the intertropical convergence zone (ITCZ), where the trade winds meet and rain is heaviest; it flows poleward aloft, sinks around 30° (the subtropical highs, where most great deserts lie) and returns as the easterly trade winds.
• Ferrel cells (about 30° to 60°): surface westerlies; the mid-latitude storm track.
• Polar cells (about 60° to 90°): cold air sinks at the pole and spreads out as polar easterlies.
• Jet streams are fast westerly winds near the tropopause: the polar jet (about 50° to 60°, strongest where the temperature contrast between polar and mid-latitude air is sharpest) and the subtropical jet (about 30°). A weaker, wavier polar jet lets weather systems stall, producing long heat waves under blocking highs and cold outbreaks further south.
• Monsoons are seasonal wind reversals driven by land heating faster than ocean in summer.
• High-pressure systems have sinking air and weak winds, which suppress vertical mixing and trap pollutants near the ground.

Ocean circulation carries the rest of the poleward heat. Wind-driven gyres circulate around each ocean basin, with fast warm western boundary currents such as the Gulf Stream. Winds blowing along the eastern edges of the oceans drive coastal upwelling (off Peru, California, northwest and southwest Africa), bringing cold, nutrient-rich water up and supporting rich fisheries. The thermohaline circulation, driven by differences in temperature and salinity, sinks cold, salty water in the North Atlantic (the Nordic and Labrador Seas) and around Antarctica and spreads it through the deep ocean over about a thousand years. Its Atlantic branch, the Atlantic Meridional Overturning Circulation (AMOC), warms northwestern Europe; meltwater and warming make the surface water lighter, which is expected to weaken it. The ocean is the largest heat buffer in the climate system: it has absorbed about 90% of the extra energy from the energy imbalance.

Polar amplification: the Arctic has warmed about three to four times as fast as the global average since 1979, because of ice-albedo feedback, a lapse-rate feedback that is positive at high latitudes (warming is trapped near the surface in the stable polar air) and heat carried poleward. The faint young Sun paradox: the Sun was about 25 to 30% dimmer about 4 billion years ago, yet geology shows liquid water on early Earth, which requires a much stronger greenhouse effect (more CO₂, and perhaps methane) or a lower albedo then.

A short history of greenhouse science. Joseph Fourier (1820s) argued that the atmosphere keeps Earth warmer than sunlight alone would. Eunice Foote (1856) showed that air with CO₂ warmed more in sunlight, and John Tyndall (1859) measured the infrared absorption of water vapor and CO₂. Svante Arrhenius (1896) made the first calculation linking CO₂ to temperature, estimating several degrees of warming for a doubling. Guy Callendar (1938) connected rising CO₂ from fossil fuels to observed warming, and Charles Keeling began the Mauna Loa record in 1958.

Climate change effects on weather and ecosystems

Global mean surface temperature (GMST) is the area-weighted average of near-surface air temperature over land and sea-surface temperature over the oceans, reported as an anomaly from a baseline. IPCC AR6 measured 2011 to 2020 at about 1.1 °C above 1850 to 1900; 2024 was the warmest year on record at about 1.55 °C above that baseline (WMO), the first calendar year above 1.5 °C. Climate change is broader than global warming: it includes changes in precipitation, extremes, sea level and ice.

A biome shift of boreal forest into tundra is a positive feedback: the dark canopy sticking out above snow has a much lower albedo than snow-covered tundra, so the region absorbs more sunlight and warms further. Species most likely to survive have wide climatic tolerance ranges. CO₂ fertilization has made the planet greener overall, but heat and drought stress can turn forests from carbon sinks into carbon sources: a forest that releases more carbon than it absorbs is a carbon source.

Tipping points are thresholds beyond which a part of the climate system changes self-sustainingly and is hard to reverse. The 2025 Global Tipping Points Report concluded that warm-water coral reefs have already passed theirs (estimated near 1.2 °C of warming). The Greenland and West Antarctic ice sheets, permafrost and the North Atlantic subpolar gyre have estimated thresholds around 1.5 °C, close to current warming, and the Amazon rainforest (whose threshold deforestation may be lowering) and AMOC have higher, more uncertain central estimates. If a test asks which element is closest to or already past its threshold, coral reefs and then the Greenland and West Antarctic ice sheets are the expected answers.

Forests as carbon sinks. Boreal and temperate forests remain net carbon sinks, but intact tropical forests have weakened as sinks; with deforestation, fire and degradation the tropics as a whole are now roughly carbon-neutral or a net source in some estimates (parts of the southeastern Amazon emit more than they absorb). Hotter, drier conditions are expected to reduce local tree species richness as narrow-tolerance species decline or shift.

Coral bleaching. NOAA Coral Reef Watch flags thermal stress when sea-surface temperature exceeds the long-term maximum monthly mean by about 1 °C. Accumulated stress is measured in degree heating weeks (DHW, °C-weeks above that threshold over 12 weeks): about 4 DHW brings significant bleaching and about 8 DHW severe, widespread bleaching and mortality. For many tropical reefs the danger zone begins near 29 to 30 °C.

Photosynthesis pathways. Most plants (wheat, rice, trees) use C3 photosynthesis, which loses efficiency in heat because the enzyme RuBisCO increasingly grabs O₂ instead of CO₂ (photorespiration). C4 plants (corn, sugarcane, many tropical grasses) concentrate CO₂ around RuBisCO, so they do better in hot, sunny, drier climates. CAM plants (cacti, pineapple, agave) open their stomata only at night to save water, suiting deserts. Rising CO₂ benefits C3 plants most; rising heat and drought favor C4 and CAM plants.

El Niño–Southern Oscillation

The El Niño–Southern Oscillation (ENSO) is a coupled ocean-atmosphere cycle in the tropical Pacific. Normally, trade winds blow west along the equator, piling warm surface water in the west, deepening the warm layer there and letting cold, nutrient-rich water well up along South America. During El Niño, the trade winds weaken, warm water spreads east, upwelling along South America is suppressed, and the sea surface in the east becomes higher as well as warmer. During La Niña, the trade winds strengthen, the east cools further and upwelling increases.

Regional effects are the most tested part; they are tabulated in "ENSO impacts by region" below. In satellite questions, look for sea-surface temperature anomalies in the equatorial eastern Pacific (MODIS on Aqua, GOES) or sea-surface-height anomalies (JASON) to identify the phase.

The machinery. The Walker circulation is the east-west loop over the equatorial Pacific: air rises over the warm western Pacific (heavy rain over Indonesia), flows east aloft, sinks over the cool eastern Pacific (dry coastal Peru and Chile) and returns west as the trade winds. The thermocline, the boundary between warm surface water and cold deep water, normally tilts: deep (about 150 to 200 m) in the west and shallow (tens of meters) in the east, and sea level is roughly half a meter higher in the west than in the east. The Bjerknes feedback couples these: a warmer east weakens the east-west temperature contrast, which weakens the trade winds, which lets more warm water slosh east and deepens the eastern thermocline, which warms the east further. That positive feedback grows an El Niño (or, run the other way, a La Niña). Negative feedbacks end events: the ocean adjusts through slow Kelvin waves (eastward-moving bulges of warm water and a deepened thermocline along the equator, visible in Jason sea-surface height as a +5 to +20 cm ridge traveling east over about two months) and reflected Rossby waves, so each event eventually reverses. ENSO recurs irregularly every 2 to 7 years; events usually peak in northern winter and last about 9 to 12 months, La Niña sometimes longer. Between events conditions are ENSO-neutral.

How it is measured.

• Oceanic Niño Index (ONI): NOAA's official index, the 3-month running mean of SST anomalies in the Niño 3.4 region (5° N to 5° S, 120° W to 170° W). El Niño conditions: ONI ≥ +0.5 °C; La Niña: ≤ −0.5 °C; an episode is declared when the threshold holds for at least five consecutive overlapping 3-month seasons. Strength categories are often quoted as weak (0.5 to 0.9), moderate (1.0 to 1.4), strong (1.5 to 1.9) and very strong (2.0 or more).
• Southern Oscillation Index (SOI): the standardized difference in sea-level pressure between Tahiti (central Pacific) and Darwin, Australia (west). Normally Tahiti's pressure is higher; a sustained negative SOI (lower pressure at Tahiti, higher at Darwin) means El Niño, positive means La Niña. The "Southern Oscillation" is this atmospheric pressure seesaw.
• Satellites: SST anomaly maps (MODIS on Aqua for long consistent records from a sun-synchronous orbit; GOES for continuous coverage of the eastern Pacific) show a tongue of warm anomalies along the equator in the east during El Niño, cold during La Niña. Jason sea-surface-height anomaly maps show the same pattern as height: warmer water has expanded and the thermocline is deeper, so sea level is higher in the east (and lower in the west) during El Niño. TOPEX/Poseidon's tracking of the 1997 to 1998 El Niño made it famous. OLR and GPM rain maps show convection shifted into the central and eastern Pacific. OCO-2 and the Keeling curve show CO₂ growing faster during and after El Niño.

ReferenceENSO impacts by region· 14 rows

Typical patterns; any one event can differ. Winter means northern winter (December to February) unless stated.

14 rows

Region	El Niño	La Niña
Coastal Peru and Ecuador	Heavy rain and floods; warm water ends upwelling, collapsing anchovy fisheries and seabird colonies	Cooler, drier; strong upwelling and productive fisheries
Central and eastern tropical Pacific	Warmer SST, more rain and convection, higher sea level	Cooler SST, drier, lower sea level
Indonesia, Papua New Guinea, Philippines	Drought, peat and forest fires, haze	Wetter than normal, flooding
Eastern Australia	Drought, heat, bushfire risk; rainforests can switch from carbon sink to source	Wet, cool, flooding
India and South Asia	Weaker summer monsoon, drought risk	Stronger monsoon
Southern Africa	Drought in the summer rainy season	Wetter
East Africa	Wetter short rains (October to December)	Drier, drought
Northern South America (Colombia, Venezuela, northern Brazil)	Drought	Wetter
Southern United States and Gulf Coast	Cooler and wetter winter, more storms (subtropical jet strengthened and shifted south)	Warmer and drier winter, drought and fire risk
Pacific Northwest, northern Plains, Ohio Valley, Canada	Warmer, drier winter	Colder, wetter (snowier) winter
Atlantic hurricanes	Suppressed by stronger vertical wind shear	Enhanced (weaker shear)
Central and eastern Pacific hurricanes	Enhanced	Suppressed
Global mean temperature	Rises, by about 0.1 to 0.2 °C, peaking a few months after the event; El Niño years (1998, 2016, 2023 to 2024) set records	Falls slightly
Atmospheric CO₂ growth	Faster: tropical drought and fires weaken the land sink	Slower

Is ENSO a forcing? No. ENSO is internal variability: it redistributes heat between the ocean and atmosphere rather than changing the energy entering or leaving the planet. A volcanic eruption cools by changing albedo (an external forcing); El Niño warms the surface by releasing ocean heat. Some test keys still label La Niña a "negative forcing (cooling)": if forced to choose, give the direction of its surface effect, but note that it is internal variability. If a strong El Niño coincides with a large eruption, the effects partly cancel: the global surface warms less than El Niño alone would make it.

ENSO and climate change. Warming adds to every ENSO state: El Niño years are the record years. Because a warmer atmosphere holds more water vapor, rainfall anomalies during events are expected to intensify; whether events will become more frequent or stronger is still uncertain.

Teleconnections. ENSO changes weather far from the Pacific through teleconnections: the shift of tropical rainfall and heating alters the large-scale atmospheric circulation (the Walker and Hadley circulations and the waves they send into the midlatitudes), which bends the jet streams and storm tracks. That is how a warm eastern Pacific can bring a wet winter to the southern United States, a mild winter to Canada, drought to southern Africa and fewer Atlantic hurricanes.

Carbon and nitrogen cycles

The carbon cycle moves carbon among the atmosphere, oceans, plants and soils. Photosynthesis takes CO₂ out of the air; respiration, decomposition and fire return it. Fossil-fuel burning and land-use change add carbon faster than natural sinks absorb it, and the ocean takes up part of the excess, which makes seawater more acidic. Remote sensing measures pieces of this: column CO₂ from OCO-2, vegetation greenness from optical sensors, and ocean chlorophyll (a proxy for marine primary production) from ocean-color sensors such as PACE and MODIS.

Primary production. Gross primary production (GPP) is the total carbon fixed by photosynthesis; net primary production (NPP) is GPP minus the carbon the plants themselves respire (autotrophic respiration), so it is the carbon stored as new plant growth, available to the rest of the ecosystem, usually in g C/m² per year. Land NPP is roughly 50 to 60 Gt C per year and ocean NPP about the same. From space, NPP is estimated with a light-use-efficiency model: absorbed photosynthetically active radiation (from NDVI or similar indices) times an efficiency reduced by temperature and water stress. MODIS produces global GPP and NPP products (MOD17), Landsat-based versions give 30 m detail for farms and forests, and ocean NPP comes from chlorophyll maps (MODIS, PACE). Net ecosystem production subtracts the respiration of animals and microbes too; a positive value means the ecosystem is a carbon sink.

Reservoirs. Carbon is measured in gigatonnes of carbon (GtC, $10^{15}$ g of carbon; multiply by 3.67 for the mass of CO₂). One ppm of atmospheric CO₂ is about 2.12 GtC.

Fluxes and the human perturbation. Natural exchanges are large and nearly balanced: land plants take up roughly 120 GtC per year by photosynthesis and release about the same by plant and soil respiration (the largest natural flux of carbon into the atmosphere) and fire; the ocean and atmosphere exchange about 80 to 90 GtC per year each way. Human emissions are smaller but one-directional. From 2014 to 2023 (Global Carbon Budget 2024), fossil fuels and cement released about 9.7 GtC per year and land-use change about 1.1 GtC per year, close to 11 GtC per year in total (about 40 Gt of CO₂). Of that, about 48% stayed in the atmosphere (the airborne fraction, about 5 GtC per year), about 26% went into the ocean and about 30% into the land biosphere, the land and ocean taking roughly equal shares (the shares add to slightly more than 100% because the Global Carbon Budget has an unexplained imbalance of a few percent). Both sinks are projected to absorb a smaller fraction as emissions continue, because warmer water dissolves less CO₂ and drought and heat stress plants. Since 1750, CO₂ has risen about 50%, methane about 166% and N₂O about 25%.

Each process moves carbon in one direction; decide whether it removes carbon from the air (sink) or adds it (source).

A carbon sink is any reservoir or process that takes up more carbon than it releases; a source releases more than it takes up. Ocean, forests, soils, peat and carbonate rocks are sinks. Blue carbon is carbon stored by coastal and marine ecosystems: mangroves, seagrass meadows and salt marshes. A burned or cleared forest becomes a source; heat waves increase respiration and reduce photosynthesis, giving a net release. Converting tropical forest to cropland releases CO₂ (burning, soil carbon loss) and raises N₂O (fertilizer, disturbed soils).

Ocean acidification. Dissolved CO₂ forms carbonic acid, which releases hydrogen ions:

$$\text{CO}_2 + \text{H}_2\text{O} \rightleftharpoons \text{H}_2\text{CO}_3 \rightleftharpoons \text{H}^+ + \text{HCO}_3^- \qquad \text{H}^+ + \text{CO}_3^{2-} \rightarrow \text{HCO}_3^-$$

The extra H⁺ lowers pH, and it combines with carbonate ions, so less carbonate (CO₃²⁻) is available to corals, shellfish, pteropods and coccolithophores that build calcium carbonate shells and skeletons. Surface ocean pH has fallen from about 8.2 before industrialization to about 8.1 now; because pH is logarithmic, a drop of 0.1 means about $10^{0.1} = 1.26$ times as many hydrogen ions, an increase of about 26 to 30%. Satellites cannot measure pH directly; they monitor related quantities (SST, chlorophyll, coral reef condition) and ships, buoys and floats measure the chemistry. Warmer water holds less CO₂ (lower solubility) and a more stratified ocean mixes surface carbon down more slowly, both weakening the ocean sink, a positive feedback.

Nitrogen. The nitrogen cycle is linked through plant growth. Nitrogen-fixing bacteria (and lightning, and the industrial Haber-Bosch process for fertilizer) convert atmospheric N₂ into forms plants can use; microbes later return it as N₂ and as N₂O by denitrification. Excess fertilizer therefore raises N₂O emissions, and runoff feeds algal blooms that change ocean color. Combustion nitrogen oxides feed ozone and aerosol formation.

Nitrogen gas makes up 78% of the air, but its triple bond makes it unusable by almost all organisms. Converting N₂ into a reactive form is nitrogen fixation, and there are four main pathways:

The other steps move nitrogen that is already reactive; they are not fixation:

Human disruption. Synthetic fertilizer, cultivation of nitrogen-fixing crops and combustion now create more reactive nitrogen each year than all natural processes on land, roughly as much as natural fixation on land and sea combined. Consequences: N₂O emissions (agriculture is the largest human source; N₂O has a GWP of about 273 and also destroys ozone); NOₓ from fuel burning, which makes ozone, nitrate aerosol and acid rain; ammonia from fertilizer and manure, which forms ammonium aerosol; and runoff that causes eutrophication: excess nitrate and phosphate feed algal blooms, the bloom dies and its decay uses up oxygen, creating hypoxic dead zones (the Gulf of Mexico dead zone below the Mississippi is the best-known US example). Algal blooms appear as bright green water in satellite images of lakes and coasts. Nitrogen deposition can fertilize forests briefly but acidifies soils and stimulates microbial decomposition and N₂O release over time.

Ocean color makes the link visible. Chlorophyll a absorbs blue and red light and reflects green, so greener water means more phytoplankton. As warming moves the zones of ideal phytoplankton growth, regional ocean color shifts, which matters to fisheries and the coastal economies that depend on them.

Ocean carbon chemistry. Dissolved inorganic carbon in seawater is about 90% bicarbonate (HCO₃⁻), about 9% carbonate (CO₃²⁻) and only about 1% dissolved CO₂ plus carbonic acid (H₂CO₃), so "bicarbonate" is the right answer for the main form of ocean carbon. CO₂ dissolves better in cold water, so cold high-latitude seas (the North Atlantic and Southern Ocean) are the strongest CO₂ sinks, and where deep water forms there the carbon is carried down; warm equatorial upwelling regions release CO₂.

The Haber-Bosch fertilizer process is the largest human source of reactive nitrogen (about 120 Tg N per year), ahead of nitrogen fixation by legume crops (about 60 Tg N per year) and fossil-fuel combustion (about 25 to 30 Tg N per year as NOx). When tests rank carbon sinks they put the ocean first, then forests, then soils, with volcanoes as a source (in the annual budget the ocean and land sinks are similar in size).

Sea-level rise and surface waves

Sea level rises for two reasons. Warm water expands (thermal expansion, a volume change), and melting land ice adds mass. GRACE (Space gravimetry) sees ocean mass, ice-sheet loss, groundwater storage and hydrologic redistribution, but not thermal expansion. Altimeters such as JASON measure the sea-surface height, which combines both effects, and thermal expansion is measured with temperature profiles from ocean instruments such as Argo floats. The rise is a few millimeters per year, small enough that altimeters must average many measurements. Read each question's wording for whether it asks about mass, height or temperature.

Rates and contributions. Tide gauges show about 1.3 to 1.5 mm/yr through most of the 20th century; satellites since 1993 show an average of about 3.3 mm/yr, roughly doubling. In recent decades the budget is roughly: thermal expansion about 40%, mountain glaciers about 20%, the Greenland ice sheet about 20%, the Antarctic ice sheet about 10 to 15%, and changes in land water storage (groundwater pumping, reservoirs) the rest. Thermal expansion is still the largest single term (about 40%), which is the answer to most "leading cause" questions, but melting land ice from glaciers and both ice sheets combined has now overtaken it as the larger total contributor. The two ice sheets hold about 7.4 m (Greenland) and about 58 m (Antarctica) of sea-level equivalent.

How altimetry works and what can go wrong. Jason flies at 1,336 km in a 66° non-sun-synchronous orbit with a 9.9-day repeat, timing nadir radar pulses. Sea-surface height = orbit height (known to about a centimeter from DORIS, GPS and laser tracking) minus range. The range is corrected for the ionosphere, water vapor, tides, air pressure and sea state, as described in Radar altimetry and scatterometry. One pass is accurate to about 3 cm; averaging millions of measurements gives the global mean to millimeters. Jason measures absolute sea level; tide gauges measure relative sea level, which also includes land rising or sinking (subsidence makes local relative rise faster, as along the US Gulf Coast).

Reading a sea-surface-height anomaly map. Red means higher than average, blue lower. A positive anomaly usually means warmer, expanded water or a deeper thermocline (El Niño along the eastern equatorial Pacific, warm eddies, warm-core rings shed by the Gulf Stream); a negative anomaly means cooler, denser water or upwelling. Because warm water stands higher, SSH anomalies are a useful proxy for upper-ocean heat content, which is why Jason data help forecast hurricane intensification. A time series of global mean sea level shows a steady rise with a small seasonal wobble and dips during La Niña (when rain falls heavily on land, especially Australia, temporarily moving water out of the ocean) and bumps during El Niño.

Reading a GRACE map. GRACE and GRACE-FO report monthly mass changes as equivalent water thickness (the depth of water, in cm, that would produce the gravity change), at a resolution of a few hundred kilometers. Red or brown means mass loss, blue mass gain. The tested patterns:

Detecting surface melt on ice sheets. Liquid water changes how snow interacts with microwaves. Dry snow is nearly transparent to microwaves and scatters them, so its microwave brightness temperature is low; when meltwater appears, the snow absorbs and emits like a blackbody, so passive microwave brightness temperature (for example at 19 or 37 GHz from SSM/I or AMSR2, or at L band from SMAP) jumps sharply. Radar backscatter does the opposite: wet snow absorbs the pulse and looks dark (low σ⁰) in scatterometer and SAR images such as Sentinel-1. Daily melt-extent maps of Greenland and Antarctica come from these signals. Confounders are rain on the snow (also liquid water), refrozen meltwater forming ice layers (which change backscatter and emission after the melt ends) and melt ponds or lakes; thermal-infrared and optical images (dark blue lakes, darker wet snow) help confirm melt.

Waves, wind and surge. The significant wave height (SWH) is the average height of the highest one-third of waves, about four times the standard deviation of the sea-surface elevation ($\text{SWH} \approx 4\sigma$). An altimeter gets it from the leading edge of the returned pulse: over a flat sea the echo rises abruptly, over high waves the crests return first and troughs last, stretching the leading edge. The midpoint of the leading edge gives the range to mean sea level. Wind speed comes from the echo's strength: at nadir a calm sea reflects like a mirror and gives a strong return, while wind-roughened water scatters energy away, so stronger wind means a weaker echo (the opposite of side-looking radar, where rough water is brighter). Storm surge is the rise of water pushed ashore by a storm's winds, plus about 1 cm of rise for every hectopascal of pressure drop; rising sea level raises the base on which every surge builds.

Energy balance and feedbacks (Division C)

Planetary energy balance is the most mathematical area of the Division C scope, and it rewards clean, labeled diagrams. The idea is simple: Earth absorbs energy from the Sun, emits infrared radiation to space, and stays at a temperature where the two are equal, averaged over time. Everything else (albedo, clouds, greenhouse gases, aerosols) changes the terms in that balance.

Planetary energy balance and equilibrium temperature

The factor of 4 comes from geometry. Earth intercepts sunlight over a disk of area $\pi R^{2}$, but emits over a sphere of area $4\pi R^{2}$, so the average incoming power per unit area is one quarter of the solar irradiance times the fraction not reflected.

Written as total power rather than per square meter, the balance is $P_{in} = \pi R^{2} S(1-A)$ and $P_{out} = 4\pi R^{2}\varepsilon\sigma T^{4}$. The radius cancels, which is why a planet's equilibrium temperature does not depend on its size: doubling Mars's radius would not change its temperature. A planet's mass, density and composition do not enter either; you need only the stellar flux, albedo and emissivity. The common assumptions are that the planet is a uniform sphere, that it rotates fast enough to spread heat around, that it is heated only by its star (Earth's geothermal heat flow, about 0.09 W/m², is negligible) and that it emits as a blackbody or a uniform gray body.

Published energy budgets differ by a few W/m²; use the values a question gives, not the diagram's.

Albedo versus emissivity sensitivity. From $T_e \propto (1-A)^{1/4}\varepsilon^{-1/4}$, small changes give $\dfrac{\Delta T}{T} \approx -\dfrac{1}{4}\dfrac{\Delta A}{1-A}$ and $\dfrac{\Delta T}{T} \approx -\dfrac{1}{4}\dfrac{\Delta\varepsilon}{\varepsilon}$. With $A = 0.30$ and an effective $\varepsilon$ of about 0.61, an albedo change of 0.01 shifts temperature by about $-\tfrac14(0.01/0.70) = -0.36\%$ (about 0.9 K at 255 K), while an emissivity change of 0.01 shifts it by about $-\tfrac14(0.01/0.61) = -0.41\%$; the larger factor ($1/\varepsilon$ versus $1/(1-A)$) wins. (Division C calculation.)

Earth's energy budget, term by term

The table gives the global annual mean flows from IPCC AR6 (Figure 7.2, for the early 21st century), with the percentage of the 340 W/m² that arrives at the top of the atmosphere. Older textbook versions (Trenberth and others, 2009: 341 in, 102 reflected, 239 out, 396 surface emission, 333 back radiation, 80 latent, 17 sensible) differ by a few W/m² because of different data and periods. Tests that say "100 units of incoming sunlight" use these percentages.

The surface balance is: absorbed sunlight + back radiation = surface emission + sensible heat + latent heat + net storage, or $160 + 342 \approx 398 + 21 + 82 + 1$. The top-of-atmosphere balance is: $340 = 100 + 239 + 1$. Earth absorbs mainly shortwave radiation (sunlight, about 0.3 to 4 µm) and emits longwave radiation (about 4 to 100 µm), so albedo is defined over solar wavelengths, roughly 0.3 to 3 µm.

The classic 100-unit diagram. Older textbooks (and many tests) use NASA's version with incoming sunlight = 100 units. Reflected, 30 units: 6 backscattered by air, 20 by clouds, 4 by the surface (so air accounts for about 20% of the reflected light and clouds about two-thirds). Absorbed, 70 units: 16 by atmospheric gases and dust, 3 by clouds, 51 at the surface.

Solar flux at any distance

The Sun's luminosity $L$ (total power, about $3.83 \times 10^{26}$ W) spreads over a sphere of radius $d$, so the flux at distance $d$ is the inverse-square law. If you are given the Sun's temperature and radius instead, its luminosity comes from Stefan-Boltzmann.

Comparing planets

The same formula explains the inner planets, and the gap between the effective and actual surface temperature measures each greenhouse effect (the "Goldilocks" comparison).

Venus absorbs less sunlight per square meter than Earth because its clouds reflect three-quarters of it, so its effective temperature is lower than Earth's; its surface is the hottest in the solar system only because of its enormous greenhouse effect. If Venus became a perfect blackbody with no atmosphere, its surface would cool to about 330 K.

Greenhouse layer models

The simplest model that explains the greenhouse effect adds one atmospheric layer that is transparent to sunlight but absorbs infrared.

One perfectly absorbing layer. The layer absorbs all of the surface's infrared and emits $\sigma T_a^{4}$ both up and down. Balance at the top: $\sigma T_a^{4} = S(1-A)/4$, so $T_a = T_e = 255$ K (the planet still emits to space at its effective temperature). Balance at the surface: absorbed sunlight plus back radiation equals surface emission, $\sigma T_s^{4} = S(1-A)/4 + \sigma T_a^{4} = 2\sigma T_e^{4}$. So $T_s = 2^{1/4}T_e = 1.189 \times 255 = 303$ K: too warm, because the real atmosphere lets some infrared escape.

Leaky (gray) layer. Let the layer absorb a fraction $\varepsilon$ of the surface's infrared (its emissivity, by Kirchhoff's law equal to its absorptivity). Then:

With two opaque layers $T_s = 3^{1/4}T_e = 336$ K; with $n$ layers $T_s = (n+1)^{1/4}T_e$, the beginning of an explanation of Venus. A different "effective emissivity" shortcut treats the whole Earth as a gray body: $\varepsilon_{eff} = (T_e/T_s)^{4} = (255/288)^4 = 0.61$ (or 0.62 with 287 K). To warm the surface to 292 K with the same absorbed sunlight, $\varepsilon_{eff}$ would fall to $(255/292)^4 = 0.58$, about a 5% reduction in effective emissivity.

Two partly absorbing (gray) layers. Give each layer emissivity $\varepsilon$; each emits $\varepsilon\sigma T^4$ both up and down and absorbs a fraction $\varepsilon$ of the infrared reaching it. Write one balance per level, top down: at the top of the atmosphere, absorbed sunlight equals the infrared escaping (surface emission transmitted through both layers plus the layers' upward emission); then a balance for each layer (absorbed from below and above equals $2\varepsilon\sigma T^4$) and for the surface (sunlight plus back-radiation equals $\sigma T_s^4$). Solving gives a surface warmer than the one-layer gray case and, for $\varepsilon = 1$, $T_s = 3^{1/4}T_e \approx 1.32\,T_e$ (about 336 K for $T_e = 255$ K), the $n$-layer result $T_s = (n+1)^{1/4}T_e$.

Albedo

Albedo is the fraction of incoming sunlight a surface reflects. It is one of the largest controls on energy absorption. Fresh snow reflects most of the sunlight that reaches it, open ocean reflects very little, and forests and dark soils fall in between. Clouds raise planetary albedo a great deal, which is why their net effect on climate is so sensitive to their type and height.

Albedo is a ratio of fluxes, reflected over incident, measured across the solar wavelengths (about 0.3 to 3 µm), from 0 (a perfect blackbody absorbs everything) to 1 (a perfect reflector). Absorbed flux is $(1-A)$ times incident flux. It depends on the wavelength (snow reflects about 0.9 of visible light but much less near-infrared), on the sun angle (water reflects little at high sun but a lot near the horizon, as glint) and on wetness (wet soil is darker).

ReferenceAlbedo of surfaces and clouds· 23 rows

Broadband (solar-spectrum) values; ranges reflect sun angle, moisture, age and condition.

23 rows

Surface	Albedo	Notes
Thick cloud (cumulonimbus, deep stratus)	0.6 to 0.9	Brightest objects in visible imagery
Stratocumulus decks	About 0.4 to 0.7	Large area, so a large cooling effect
Thin cirrus	About 0.1 to 0.4	Mostly transparent to sunlight
Fresh, dry snow	0.8 to 0.9	The brightest natural surface
Old, wet or dirty snow	0.4 to 0.7	Grains grow and fill with meltwater; soot and dust darken it
Snow-covered sea ice	0.8 to 0.9	—
Bare sea ice	0.5 to 0.7	—
Melt ponds on sea ice	0.2 to 0.4	Pond water absorbs much more
Glacier ice (white, bubbly)	0.3 to 0.6	—
Clear or dirty ice	0.1 to 0.3	—
Desert sand	0.3 to 0.45	Deserts are the brightest snow-free land
Concrete	0.25 to 0.4 (new), lower when weathered	—
White or "cool" roofs	0.6 to 0.9	—
Grassland	0.15 to 0.26	—
Cropland	0.15 to 0.25	—
Deciduous forest	0.15 to 0.2	—
Coniferous forest	0.08 to 0.15	Dark needles and deep shadows; stays dark above snow
Dry light soil, clay	0.2 to 0.35	—
Dark or wet soil	0.05 to 0.15	—
Asphalt	0.04 to 0.12	Among the darkest surfaces
Open ocean	About 0.06 at high sun, higher at low sun	The darkest large surface
Earth's surface on average	About 0.13 to 0.15	—
Earth as a planet (with clouds and atmosphere)	About 0.29 to 0.30	Clouds supply about two-thirds

Why snow is so white and ice less so. Albedo depends not only on color but on how many times light is scattered back before it can be absorbed. Fresh snow is a jumble of tiny ice grains separated by air: every air-ice interface reflects and refracts a little light (ice and air have different refractive indices), and after many such multiple scattering events most photons emerge back out the top. Ice itself absorbs very little visible light, so very little is lost. As snow ages, grains grow larger and fewer, so light travels farther through ice between interfaces and more near-infrared is absorbed: albedo falls. Meltwater filling the pores removes the air-ice interfaces and lowers it further. Glacier ice that is full of bubbles stays fairly white; clear, bubble-free ice lets light penetrate deep and looks blue or dark. The same physics explains why sea foam and clouds are white although water is clear, and why a rough or porous surface usually reflects differently from a smooth, polished one of the same material. The refractive index of water does not set the albedo of the ocean directly; it controls how light is bent and reflected at the surface, which matters mainly at low sun angles.

Estimating cloud cover from albedo. If the ocean (albedo about 0.06) covers 71% of the surface, land (about 0.2 to 0.3 for ice-free land) 29%, and clouds have albedo $A_c$ and cover a fraction $f$, then the planetary albedo is approximately $f A_c + (1-f)A_{surface}$. With a surface average of $0.71(0.06) + 0.29(0.20) = 0.10$ and $A_c = 0.5$: $0.30 = 0.5f + 0.10(1-f)$, so $f = 0.5$. Real cloud cover is closer to two-thirds because many clouds are thinner and dimmer than 0.5.

Trends. Earth's albedo has been decreasing slightly in the CERES era, as low clouds over some oceans thinned and snow and sea ice declined; this has made the energy imbalance grow. Satellite datasets disagree most over the Northern Hemisphere, where seasonal snow and land surfaces vary the most. CERES found the Northern and Southern Hemispheres reflect almost the same total sunlight, because the Southern Hemisphere's extra cloud (especially over the Southern Ocean) makes up for the Northern Hemisphere's larger land area.

Radiative forcing

Radiative forcing is the change in net energy flux at the top of the atmosphere caused by a change in a driver, such as the concentration of a greenhouse gas, with the rest of the system held fixed. It is measured in W/m². Forcing is not temperature. A positive forcing means the planet gains energy and will warm until the balance is restored, which takes time because the oceans store heat. For carbon dioxide, a widely used approximation from a published radiative-transfer study is $\Delta F \approx 5.35 \ln(C/C_0)$ W/m² (Myhre et al., 1998).

The eventual temperature change is $\Delta T = \lambda\,\Delta F$, where $\lambda$ is the climate sensitivity parameter in K per W/m² and includes the feedbacks. The IPCC's best estimate for the warming after a CO₂ doubling is about 3 °C, with a likely range of 2.5 to 4 °C, so $\lambda \approx 0.8$ K/(W/m²).

The IPCC now reports effective radiative forcing (ERF), which lets the atmosphere and land adjust quickly (stratospheric temperature, clouds) before measuring the flux change, but keeps the ocean surface fixed; it predicts the eventual warming better than the instantaneous forcing. The doubled-CO₂ ERF in AR6 is about 3.9 W/m². Do not confuse forcing with the planet's energy imbalance, which is what remains of the forcing after the planet has partly warmed and radiates more: today's total forcing is about 2.7 W/m² but the imbalance is only about 1 W/m². Some tests use "radiative forcing" loosely for "absorbed minus emitted"; read the definition the question gives.

ReferenceEffective radiative forcing, 1750 to 2019 (IPCC AR6)· 15 rows

Best estimates in W/m², with the very likely (5 to 95%) range in brackets.

15 rows

Driver	ERF (W/m²)	Sign and mechanism
Carbon dioxide	+2.16 [1.90 to 2.41]	The largest single forcing
Methane	+0.54 [0.43 to 0.65]	Includes no indirect ozone effect
Nitrous oxide	+0.21 [0.18 to 0.24]	—
Halogenated gases (CFCs, HCFCs, HFCs and others)	+0.41 [0.33 to 0.49]	—
Ozone (troposphere increase minus stratosphere loss)	+0.47 [0.24 to 0.71]	Made from NOₓ, CO, VOCs and CH₄ emissions
Stratospheric water vapor (from methane oxidation)	+0.05	Small positive
Total well-mixed greenhouse gases	+3.32	—
Aerosol-radiation interactions	−0.22	Scattering minus black-carbon absorption
Aerosol-cloud interactions	−0.84	Twomey and lifetime effects
Total aerosols	−1.06 [−1.71 to −0.38]	Largest uncertainty
Land-use albedo change	−0.20	Clearing forest brightens the land
Black carbon on snow, contrails	About +0.08 and +0.06	Small positive
Total anthropogenic	+2.72 [1.96 to 3.48]	Positive: the planet is gaining energy
Solar change	About +0.01	Negligible since 1750
Volcanic	Episodic, up to several W/m² negative for a year or two after large eruptions	Not a long-term trend

Classifying forcings. A frequent question type gives a scenario and asks positive (warming) or negative (cooling).

Who caused it. Warming tracks cumulative CO₂ emissions, so historical responsibility follows cumulative totals: since 1850 the United States has emitted the most (about a fifth to a quarter of the total), followed by China, then Russia, Germany and the United Kingdom (fossil CO₂ only; counting deforestation, Brazil and Indonesia join the top five); by region, North America and Europe together account for roughly half (with land use) to 60% (fossil only) of cumulative emissions, although China now emits the most each year.

Feedbacks

A feedback is an effect that changes the system's response to a forcing. A positive feedback amplifies the change; a negative feedback reduces it. The cleanest example is ice-albedo feedback. Warming melts ice, exposing darker ocean or land, which absorbs more sunlight, which warms the planet further. Water vapor feedback works similarly: a warmer atmosphere holds more water vapor, which is itself a greenhouse gas, amplifying warming. The main negative feedback is the increase in outgoing thermal emission as the planet warms, which scales with the fourth power of temperature, and so tends to restore balance.

Feedbacks are measured by the feedback parameter $\alpha$, the change in net top-of-atmosphere flux per kelvin of surface warming, in W·m⁻²·K⁻¹. A negative value is stabilizing. The equilibrium warming for a forcing $\Delta F$ is $\Delta T = -\Delta F/\alpha_{net}$, so $\lambda = -1/\alpha_{net}$.

ReferenceClimate feedbacks and their signs· 17 rows

Fast-feedback values are IPCC AR6 best estimates. "+" amplifies warming, "−" damps it.

17 rows

Feedback	Sign	Size (W·m⁻²·K⁻¹)	Mechanism
Planck (blackbody) response	−	−3.22	A warmer planet radiates more ($\sigma T^4$); the basic stabilizer
Water vapor	+	About +1.8 (the largest positive feedback)	About 7% more vapor per K (Clausius-Clapeyron) absorbs more infrared
Lapse rate	− (globally)	About −0.5	In the tropics the upper troposphere warms more than the surface and radiates more efficiently; it is positive at the poles
Water vapor + lapse rate combined	+	+1.30	Usually quoted together because they are linked
Surface albedo (snow and sea ice)	+	+0.35	Melting exposes darker surfaces
Clouds	+ (likely)	+0.42 (−0.10 to +0.94)	Higher, colder high clouds; fewer subtropical low clouds; partly offset by brighter high-latitude clouds
Net fast feedbacks	−	−1.16	The climate is stable but its response is amplified
Permafrost carbon	+	Slow	Thawing soils release CO₂ and CH₄
Wetland and Arctic methane	+	Slow	Warmer, wetter soils produce more CH₄
Ocean CO₂ solubility and stratification	+	Slow	Warmer water holds less CO₂; less mixing
Forest dieback, fire and drought	+	Slow	Releases carbon; weakens the land sink
Boreal forest moving into tundra	+	Slow	Dark canopy lowers albedo
Ice-sheet albedo and size	+	Centuries to millennia	Lower, darker ice surfaces
CO₂ fertilization of plants	−	Decades	More CO₂ speeds photosynthesis (water-use efficiency rises; transpiration per plant falls)
Ocean dissolution of CO₂ (more CO₂ in air, more dissolves)	−	Decades to centuries	Takes up part of the excess
Silicate weathering	−	Hundreds of thousands of years	Warmer, wetter climate weathers more rock, drawing down CO₂
Desertification raising albedo; white daisies in Daisyworld	−	Model examples	A biological or surface change that raises albedo as warming proceeds

Climate sensitivity has two standard measures. Equilibrium climate sensitivity (ECS) is the eventual global surface warming after CO₂ doubles and the climate fully adjusts (fast feedbacks, deep ocean in balance): best estimate 3 °C, likely 2.5 to 4 °C, very likely 2 to 5 °C (AR6). Transient climate response (TCR) is the warming at the moment of doubling when CO₂ rises 1% per year (about 70 years): best estimate 1.8 °C, likely 1.4 to 2.2 °C. TCR is smaller because the ocean is still absorbing heat. Climate models disagree on ECS mainly through their cloud feedbacks; the spatial pattern of warming matters too, because warming in regions that radiate to space inefficiently (such as places with high clouds above) raises the effective sensitivity.

Heating and cooling rates

The Division C rules stress interpreting heating and cooling rates. Expect a graph or table of temperature, heat content or energy imbalance over time, and questions asking you to read the rate, identify the type of change and explain what physical change caused it. No calculus is needed: a rate is the slope between two points.

• Linear: a constant heating rate. Power $P$ into a mass $m$ with specific heat $c$ gives $\Delta T = Pt/(mc)$, so the rate is $P/(mc)$ in K/s.
• Piecewise linear: a different constant rate in each interval. Compute each segment starting from the temperature at the end of the one before. Example: starting at 280 K, a surface warms at 2.0 K/h for 6 h, then cools at 0.5 K/h for 10 h. After 6 h it is 292 K; after 16 h it is $292 - 5 = 287$ K.
• Exponential: a body relaxing toward equilibrium, with the rate proportional to the remaining gap.

The same law is often written with a rate constant $k = 1/\tau$ (units of inverse time), as Newton's law of cooling: the rate of temperature change is proportional to the temperature difference from the surroundings, so $T(t) = T_{env} + (T_0 - T_{env})e^{-kt}$. Any quantity that decays in proportion to how much is left (a pulse of methane, an anomaly, a radioactive sample) follows $N = N_0 e^{-kt}$. Useful relations:

Reading the shape. A straight line means a steady imbalance (constant forcing, no feedback yet). A piecewise-linear record has breakpoints where something changed: a volcanic eruption that cools for a year or two, a step in emissions, or the shift between seasons. An exponential approach that flattens out means a negative feedback (more outgoing radiation as the planet warms) is closing the gap. Exponential growth that steepens means the driver itself is compounding: CO₂ rising 0.5% per year doubles in $\ln 2 / 0.005 \approx 139$ years. Because forcing depends on the logarithm of concentration, steady exponential growth in CO₂ gives a forcing that rises roughly linearly in time.

Match the shape of a curve to the physics that produces it:

Breadth reference

Tests reward breadth: many questions are one-off recall of a single term, person or date that the topic sections above mention only in passing, or not at all. This section collects those facts in two searchable tables. Type a word into the filter box above a table (for example "radar", "ozone" or "1972") to show only matching rows, then follow the "See also" column to the section that explains the idea in full.

Glossary

ReferenceRemote sensing and climate glossary· 398 rows

398 rows

Term	Meaning	See also
Absorbance	Logarithmic measure of how much light a sample absorbs: $A = \log_{10}(I_0/I)$; optical depth uses the natural log instead	Attenuation and the Beer-Lambert law
Absorptance (absorptivity)	Fraction of incident radiation a surface absorbs; for an opaque body, absorptance + reflectance = 1	Emissivity, Kirchhoff's law and brightness temperature
Absorption band	Wavelength range where a gas absorbs strongly because photons match its vibration, rotation or electronic transitions	Absorption bands and atmospheric windows
Accuracy assessment	Checking a classified map against reference data with a confusion (error) matrix, overall accuracy and kappa	Classifying land cover
Across-track scanner	Whiskbroom scanner: a rotating or oscillating mirror sweeps a few detectors side to side, perpendicular to the flight path	Scanning systems
Active sensor	Sensor that supplies its own energy and records the return (radar, lidar, sonar, scatterometer, altimeter)	Active and passive sensors classified
Adjacency effect	Light reflected by bright neighboring ground scattered into a pixel's line of sight, blurring contrast at edges	Scattering and absorption in the air
Aerosol	Solid or liquid particles suspended in air (dust, sea salt, sulfate, smoke, soot, pollen)	Aerosols
Aerosol indirect effect	Aerosols acting as cloud condensation nuclei change cloud droplet number, brightness and lifetime (Twomey and Albrecht effects)	Aerosols
Aerosol optical depth (AOD)	Column-integrated extinction by aerosol at a given wavelength (often 550 nm); 0.1 is clean air, above 1 is very hazy	Aerosol maps: MODIS, PACE and CALIPSO
Albedo	Fraction of incoming solar radiation a surface or planet reflects, over all wavelengths (Earth about 0.30)	Albedo
Albrecht effect	Aerosol-polluted clouds with smaller droplets rain less, so they last longer and cover more area (cloud lifetime effect)	Aerosols
Along-track scanner	Pushbroom scanner: a linear array of detectors images a whole line across the swath at once as the platform moves forward	Scanning systems
Altimeter	Instrument that measures height by timing a radar or laser pulse to the surface and back	Radar altimetry and scatterometry
Ammonification	Decomposers convert organic nitrogen in dead matter and waste into ammonia/ammonium (NH₃/NH₄⁺)	Carbon and nitrogen cycles
Anaglyph	Stereo pair printed in red and cyan so each eye sees one image through colored glasses, giving a 3-D view	Elevation, altitude and slope
Analemma	Figure-8 ground track traced by a geosynchronous satellite whose orbit is inclined (eccentricity makes the 8 lopsided); also the Sun's yearly figure-8	Geostationary versus geosynchronous
Anomaly	Departure of a value from its long-term average (reference period) for that place and time of year	Reading climate data products
Aperture	Opening (lens, mirror or antenna) that collects radiation; a larger aperture gives finer diffraction-limited resolution	The diffraction limit
Aphelion	Point in Earth's orbit farthest from the Sun (early July)	Kepler's laws and orbital calculations
Apogee	Point of an Earth orbit farthest from Earth	Kepler's laws and orbital calculations
Ascending node	Point where a satellite crosses the equator heading north	Sun-synchronous orbits and equator crossing times
Atmospheric correction	Removing atmospheric scattering and absorption (path radiance, transmittance) to recover surface reflectance	Image processing pipeline
Atmospheric drag	Friction from thin upper air that slowly lowers low orbits, worse when solar activity heats and expands the thermosphere	Other orbits: Molniya, Lagrange points, graveyards and decay
Atmospheric river	Long narrow plume of concentrated water vapor transport that brings heavy rain or snow when it hits land	Climate change effects on weather and ecosystems
Atmospheric window	Wavelength range where the atmosphere is largely transparent (visible, about 3 to 5 µm, 8 to 14 µm, microwave)	Absorption bands and atmospheric windows
Attenuation	Weakening of a beam by absorption plus scattering along its path	Attenuation and the Beer-Lambert law
Azimuth	Horizontal direction measured clockwise from north; in radar, the along-track direction	Radar imaging, interferometry and elevation models
Azimuth resolution	Radar resolution along the flight track; about $\lambda R / L$ for real aperture, about $L/2$ for SAR	Radar imaging, interferometry and elevation models
Backscatter	Energy scattered back toward the sensor; radar images are maps of backscatter coefficient (sigma-nought, σ⁰)	Radar imaging, interferometry and elevation models
Band	A wavelength range recorded by one detector channel; also called a channel	The four resolutions and their trade-offs
Band ratio	Dividing one band by another pixel by pixel to suppress illumination and topographic effects and enhance a target	Vegetation and spectral indices
Bandwidth	Width of the wavelength (or frequency) range a band records; narrower bands mean finer spectral resolution	The four resolutions and their trade-offs
Baseline	Distance between two antennas or two camera positions; sets the height sensitivity of InSAR and stereo	Radar imaging, interferometry and elevation models
Bathymetry	Measurement of water depth and the shape of the seafloor or lakebed	Earth processes: glaciers, rivers and stream patterns
Beer-Lambert law	$I = I_0 e^{-\tau}$: intensity falls exponentially with optical depth along the path	Attenuation and the Beer-Lambert law
Bit depth (radiometric quantization)	Number of bits per pixel value; $n$ bits give $2^n$ levels (8-bit = 256)	Bit depth, dynamic range and saturation
Black carbon	Soot from incomplete combustion; absorbs sunlight and warms; darkens snow and ice	Aerosols
Blackbody	Ideal body that absorbs all incident radiation and emits the maximum possible at every wavelength for its temperature (emissivity 1)	Blackbodies and thermal emission
Blue carbon	Carbon stored by coastal ecosystems: mangroves, salt marshes and seagrass beds	Carbon and nitrogen cycles
Blue Marble	Whole-Earth photograph taken by the Apollo 17 crew in 1972; NASA later reused the name for mosaics	People and milestones
Bolometer	Thermal detector whose electrical resistance changes as absorbed radiation heats it	Detectors
Brewster's angle	Angle of incidence at which reflected light is completely polarized; $\tan\theta_B = n_2/n_1$ (about 53° for water)	Reflection and polarization
Brightness temperature	Temperature a blackbody would need to emit the observed radiance; below true temperature when emissivity is under 1	Emissivity, Kirchhoff's law and brightness temperature
Brillouin scattering	Inelastic scattering of light by sound waves (acoustic phonons) in a medium; tiny frequency shift	Scattering and absorption in the air
Calibration	Converting raw DNs to physical units (radiance) using known references (onboard lamps, blackbodies, the Sun, vicarious sites)	Image processing pipeline
Canopy height model	Lidar surface model minus terrain model: first-return elevation minus ground elevation	Lidar
Carbon dioxide fertilization	Faster plant growth and carbon uptake under higher CO₂, part of today's land sink	Carbon and nitrogen cycles
Carbon sequestration	Long-term removal and storage of carbon in plants, soils, ocean or geologic formations	Carbon and nitrogen cycles
CCD (charge-coupled device)	Silicon detector array that stores charge in each pixel and shifts it out line by line; visible and near-IR	Detectors
Change detection	Comparing images of the same place from different dates to find what changed	Image processing pipeline
Chappuis bands	Weak ozone absorption bands across the visible, strongest near 600 nm	Absorption bands and atmospheric windows
Chlorophyll	Plant and phytoplankton pigment that absorbs blue and red and reflects green	Vegetation and spectral indices
Chlorophyll-a (Chl-a)	Main photosynthetic pigment; ocean-color sensors estimate it from blue-to-green reflectance ratios	Ocean color and phytoplankton: PACE and MODIS
Choropleth map	Map with areas (counties, countries) shaded by a data value per area	Display types
Classification	Assigning each pixel to a land-cover class by its spectral (and sometimes spatial) properties	Classifying land cover
Climate model (GCM)	General circulation model: computer simulation of atmosphere, ocean, land and ice physics on a 3-D grid	Climate drivers and circulation
Climate sensitivity	Equilibrium global warming for doubled CO₂ (IPCC AR6 likely range 2.5 to 4 °C, best estimate 3 °C)	Feedbacks
Cloud condensation nuclei (CCN)	Small particles on which water vapor condenses to form cloud droplets	Aerosols
Cloud radiative effect	Difference in top-of-atmosphere net flux between all-sky and clear-sky; globally about −20 W/m² (clouds cool overall)	Clouds and their radiative effects
Clutter	Unwanted radar echoes from the ground, sea, buildings, birds or insects that mask targets	Weather, precipitation and Doppler radar
Coherence	Degree to which the phase of two radar images (or waves) stays correlated; loss signals change in the surface	Radar imaging, interferometry and elevation models
Color composite	Image made by displaying three bands as red, green and blue	Display types
Color infrared (CIR)	False-color display with NIR as red, red as green and green as blue; healthy vegetation looks red	Division B platforms, bands and band combinations
Column (total column)	Amount of a gas in a vertical column of the atmosphere above an area (e.g., Dobson units for ozone)	Gas maps: OCO-2, TROPOMI and Aura
Commission error	Pixels wrongly included in a class (false positives); its complement is user's accuracy	Classifying land cover
Compton scattering	Inelastic scattering of X-ray or gamma photons by electrons, showing light's particle nature	Scattering and absorption in the air
Conformal projection	Map projection that preserves local angles and shapes (e.g., Mercator) but distorts area	Map projections and coordinate grids
Constellation	Group of coordinated satellites working together, shortening revisit time	Coverage, swath and revisit
Contour line	Line joining points of equal elevation; closely spaced contours mean steep slopes	Elevation, altitude and slope
Contrast stretch	Remapping DNs to use the full display range (linear, histogram-equalized) to make features easier to see	Image processing pipeline
Coral bleaching	Corals expel their symbiotic algae under heat stress and turn white; mapped from SST anomalies	Climate change effects on weather and ecosystems
Coriolis effect	Apparent deflection of moving air and water by Earth's rotation, to the right in the Northern Hemisphere	Climate drivers and circulation
Corner reflector	Two or three perpendicular surfaces that send radar straight back to the source; very bright in images	Radar imaging, interferometry and elevation models
Critical angle	Smallest incidence angle in the denser medium that gives total internal reflection; $\sin\theta_c = n_2/n_1$	Refraction and the inverse-square law
Cross-polarization	Radar transmit and receive in different polarizations (HV, VH); sensitive to volume scattering in vegetation	Radar imaging, interferometry and elevation models
Cryosphere	Frozen water on Earth: sea ice, glaciers, ice sheets, snow and permafrost	Earth processes: glaciers, rivers and stream patterns
Daisyworld	Lovelock and Watson's 1983 model planet where black and white daisies regulate temperature through albedo feedback	Feedbacks
Decibel (dB)	Logarithmic ratio unit, $10 \log_{10}(P_1/P_2)$; radar backscatter is quoted in dB	Radar imaging, interferometry and elevation models
Declination (magnetic)	Angle between true north and magnetic north, shown in the map margin	Marginal information and symbology
Denitrification	Bacteria convert nitrate back to N₂ (and some N₂O) gas under low-oxygen conditions	Carbon and nitrogen cycles
Density slicing	Assigning DN ranges of one band to discrete colors or gray levels	Display types
Depolarization ratio	Lidar ratio of cross-polarized to co-polarized return; high for ice crystals and dust, low for spherical droplets	Lidar
Descending node	Point where a satellite crosses the equator heading south	Sun-synchronous orbits and equator crossing times
Dielectric constant	Electrical property of a material that controls radar reflectivity; water's is very high, so moist soil is radar-bright	Radar imaging, interferometry and elevation models
Differential absorption lidar (DIAL)	Lidar comparing returns at an on-line (absorbed) and off-line wavelength to measure a gas concentration profile	Lidar
Diffraction	Spreading of waves as they pass an aperture or edge; sets the resolution limit $\theta \approx 1.22\lambda/D$	The diffraction limit
Diffuse reflection	Reflection scattered in many directions from a rough surface; ideal diffuse reflector is Lambertian	Reflection and polarization
Digital elevation model (DEM)	Raster grid of ground elevations	Elevation, altitude and slope
Digital number (DN)	Integer pixel value recorded by the sensor before calibration	Bit depth, dynamic range and saturation
Digital surface model (DSM)	Elevation of the top surface including trees and buildings (lidar first returns)	Lidar
Digital terrain model (DTM)	Bare-earth elevation with vegetation and buildings removed (lidar last or ground returns)	Lidar
Dimethyl sulfide (DMS)	Gas released by phytoplankton that oxidizes to sulfate aerosol and cloud nuclei over oceans	Aerosols
Dispersion	Dependence of refractive index on wavelength, which splits white light into colors in a prism	Refraction and the inverse-square law
Dissolved inorganic carbon (DIC)	CO₂(aq), carbonic acid, bicarbonate and carbonate dissolved in water	Carbon and nitrogen cycles
Dobson unit (DU)	Column ozone unit: 0.01 mm of pure ozone at standard temperature and pressure; about 300 DU is typical	Ozone: shield above, pollutant below
Doppler effect	Shift in received frequency caused by relative motion of source and observer	Weather, precipitation and Doppler radar
Doppler radar	Radar that measures the frequency shift of echoes to get radial velocity of rain or wind	Weather, precipitation and Doppler radar
Double-bounce scattering	Radar signal reflecting off ground then a vertical wall (or trunk) and back; very bright in cities and flooded forest	Radar imaging, interferometry and elevation models
Drainage pattern	Map pattern of streams (dendritic, trellis, radial, rectangular, annular) that reflects underlying rock and structure	Earth processes: glaciers, rivers and stream patterns
Dual-polarization radar	Weather radar transmitting both horizontal and vertical pulses to distinguish rain, hail, snow and debris	Weather, precipitation and Doppler radar
Dwell time	Time a detector views each ground cell; longer dwell (pushbroom) gives more signal and better SNR	Scanning systems
Eccentricity	How elongated an ellipse is: 0 is a circle, near 1 very elongated; Earth's orbit about 0.017	Kepler's laws and orbital calculations
Effective radiating level	Altitude (about 5 to 6 km) from which Earth's outgoing longwave radiation effectively escapes, at about 255 K	Greenhouse layer models
El Niño	Warm phase of ENSO: weak trade winds and warm SST in the central and eastern equatorial Pacific	El Niño–Southern Oscillation
Elastic scattering	Scattering with no change in photon energy or wavelength (Rayleigh, Mie)	Scattering and absorption in the air
Electromagnetic radiation	Energy carried by coupled oscillating electric and magnetic fields, traveling at $c$ in a vacuum	Wave basics and photon energy
Electromagnetic spectrum	Full range of EM radiation by wavelength: gamma, X-ray, UV, visible, IR, microwave, radio	The electromagnetic spectrum region by region
Elements of image interpretation	Tone/color, size, shape, texture, pattern, shadow, height, site and association	Elements of image interpretation
Emissivity	Ratio of a body's emitted radiance to a blackbody's at the same temperature and wavelength (0 to 1)	Emissivity, Kirchhoff's law and brightness temperature
Endmember	Spectrum of a pure material used to unmix mixed pixels	Pixels, mixed pixels and off-nadir growth
Energy imbalance	Absorbed solar minus outgoing longwave at the top of the atmosphere; currently about +1 W/m² or a little more, and rising	Earth's energy budget, term by term
ENSO	El Niño–Southern Oscillation: coupled ocean-atmosphere cycle of the tropical Pacific, every 2 to 7 years	El Niño–Southern Oscillation
Ephemeris	Table of a satellite's predicted (or precise) positions over time	Kepler's laws and orbital calculations
Equal-area projection	Map projection that preserves area ratios but distorts shape	Map projections and coordinate grids
Equilibrium temperature	Temperature at which absorbed sunlight equals emitted IR; Earth about 255 K	Planetary energy balance and equilibrium temperature
Eustatic sea-level change	Global change in sea level from changing ocean water volume (melt, thermal expansion)	Sea-level rise and surface waves
Eutrophication	Nutrient overloading (nitrogen, phosphorus) causing algal blooms and oxygen loss	Carbon and nitrogen cycles
Evapotranspiration	Evaporation from soil and water plus transpiration from plants	Climate change effects on weather and ecosystems
EVI	Enhanced Vegetation Index; adds a blue band and coefficients to reduce soil and atmosphere effects and avoid NDVI saturation	Vegetation and spectral indices
Exitance (radiant exitance)	Radiant flux emitted per unit area of a surface, W/m²; $\sigma T^4$ for a blackbody	Radiometric quantities and units
Exosphere	Outermost layer of the atmosphere, above about 500 to 1,000 km, where atoms escape to space	What the air is made of
External forcing	Climate driver from outside the climate system (solar output, orbital cycles, volcanoes, human emissions)	Climate drivers and circulation
Extinction	Total removal of light from a beam by scattering plus absorption	Attenuation and the Beer-Lambert law
Extinction coefficient	Fractional loss of beam intensity per unit path length	Attenuation and the Beer-Lambert law
Faint young Sun paradox	Early Sun was about 25 to 30% dimmer, yet early Earth had liquid water	Climate drivers and circulation
False color	Display in which bands are assigned to colors that differ from what the eye would see (e.g., NIR shown as red)	Display types
Feedback	Process in which a change triggers effects that amplify (positive) or damp (negative) the original change	Feedbacks
Fiducial marks	Reference marks in the corners or edges of an aerial photo frame that locate the principal point	Photographic scale
Field of view (FOV)	Total angle viewed by a sensor; with altitude it sets swath width	IFOV, GSD, FOV and swath
First return	Earliest lidar return from a pulse, usually from treetops or rooftops	Lidar
Fluorescence	Emission of light at a longer wavelength after absorbing shorter; plants emit solar-induced fluorescence (SIF) near 685 and 740 nm	Spectroscopy and gas retrieval
Flux	Energy flow per unit area per unit time (W/m²)	Radiometric quantities and units
Focal length	Distance from lens to focal plane; photo scale = focal length / flying height	Photographic scale
Footprint	Ground area illuminated or viewed by one pulse or one measurement	Radar altimetry and scatterometry
Foreshortening	Radar distortion in which slopes facing the sensor appear compressed and bright	Radar imaging, interferometry and elevation models
Fusion (pan-sharpening)	Merging a sharp panchromatic band with coarser color bands to make a sharp color image	Image processing pipeline
Geographic information system (GIS)	System for storing, analyzing and displaying spatial data in layers	Image processing pipeline
Geoid	Equipotential surface of Earth's gravity that best matches mean sea level	Space gravimetry
Geometric correction	Removing geometric distortions so image pixels match map positions	Image processing pipeline
Georeferencing	Assigning map coordinates to an image, usually with ground control points	Image processing pipeline
Geostationary orbit (GEO)	Circular equatorial orbit at 35,786 km altitude with a 1 sidereal day period; satellite appears fixed in the sky	Geostationary versus geosynchronous
Geosynchronous orbit	Orbit with a period of one sidereal day; may be inclined or eccentric	Geostationary versus geosynchronous
Global warming potential (GWP)	Heat trapped by 1 kg of a gas over a period (usually 100 years) relative to 1 kg of CO₂	Greenhouse and trace gas reference
Graticule	Grid of latitude and longitude lines on a map	Map projections and coordinate grids
Graveyard orbit	Disposal orbit about 300 km above GEO where retired geostationary satellites are moved	Other orbits: Molniya, Lagrange points, graveyards and decay
Gravity anomaly	Difference between measured and expected gravity; GRACE maps changes caused by mass shifts	Space gravimetry
Gray body	Body whose emissivity is the same at all wavelengths but below 1	Emissivity, Kirchhoff's law and brightness temperature
Greenhouse effect	Warming of a planet's surface because gases absorb and re-emit outgoing infrared	Greenhouse layer models
Greenhouse gas	Gas that absorbs and emits thermal infrared (H₂O, CO₂, CH₄, N₂O, O₃, CFCs)	Greenhouse and trace gas reference
Ground control point (GCP)	Feature with known map coordinates that is identifiable in an image; used to georeference	Image processing pipeline
Ground range	Horizontal distance on the ground from the radar's nadir track to a target	Radar imaging, interferometry and elevation models
Ground sample distance (GSD)	Distance on the ground between centers of adjacent pixels	IFOV, GSD, FOV and swath
Ground track	Path traced on Earth's surface by the point directly below a satellite	Coverage, swath and revisit
Ground truth	Field observations used to train or check image classification	Classifying land cover
Haber-Bosch process	Industrial fixation of N₂ into ammonia for fertilizer; the largest human source of reactive nitrogen	Carbon and nitrogen cycles
Hadley cell	Tropical circulation: air rises near the equator, flows poleward aloft, sinks near 30°	Climate drivers and circulation
Halocarbons	Carbon compounds containing chlorine, fluorine or bromine (CFCs, HCFCs, HFCs, halons)	Greenhouse and trace gas reference
Harmful algal bloom	Rapid growth of toxin-producing algae or cyanobacteria, often fed by nutrient runoff; seen in ocean-color images	Ocean color and phytoplankton: PACE and MODIS
Heat capacity	Energy needed to raise a body's temperature by 1 K; water's is high, which slows ocean warming	Heating and cooling rates
Heat island (urban)	Cities warmer than surroundings because of dark impervious surfaces, less vegetation and waste heat	Surface properties: albedo, temperature and materials
High-pass filter	Spatial filter that keeps rapid brightness changes, sharpening edges	Image processing pipeline
Hillshade	Shaded-relief image computed from a DEM for a chosen sun angle	Elevation, altitude and slope
Histogram	Graph of how many pixels have each DN; used to choose contrast stretches and thresholds	Image processing pipeline
Histogram equalization	Stretch that spreads DNs so each display level holds about the same number of pixels	Image processing pipeline
Hotspot	Pixel much hotter than surroundings in thermal or mid-IR, usually an active fire or lava	Geohazards and what the image shows
Hyperspectral	Imaging in many (often 100 or more) narrow contiguous bands, giving a near-continuous spectrum per pixel	Spectral classes of imagers
Hypoxia (dead zone)	Water with too little dissolved oxygen for most animals, often after eutrophication	Carbon and nitrogen cycles
Ice core	Cylinder of glacier ice whose trapped air bubbles and isotopes record past atmosphere and temperature	Climate drivers and circulation
Ice sheet	Continental-scale glacier (Greenland, Antarctica); its melt raises sea level	Sea-level rise and surface waves
Ice shelf	Floating extension of an ice sheet; its melt does not directly raise sea level but releases inland ice flow	Sea-level rise and surface waves
Ice-albedo feedback	Warming melts bright ice and snow, exposing darker surfaces that absorb more sunlight; positive	Feedbacks
IFOV (instantaneous field of view)	Angle seen by one detector at one instant; IFOV (radians) × altitude ≈ ground cell size at nadir	IFOV, GSD, FOV and swath
Illuminance	Visible light falling on a surface weighted by the eye's response, in lux (lumens/m²)	Radiometric quantities and units
Image registration	Aligning two images so the same pixel covers the same ground	Image processing pipeline
Incidence angle	Angle between the incoming beam and the surface normal (or local vertical)	Radar imaging, interferometry and elevation models
Inclination	Angle between an orbital plane and the equatorial plane; 0° equatorial, 90° polar, above 90° retrograde	Kepler's laws and orbital calculations
Index of refraction	$n = c/v$, ratio of light speed in vacuum to speed in a medium	Refraction and the inverse-square law
Inelastic scattering	Scattering that changes the photon's energy (Raman, Brillouin, Compton)	Scattering and absorption in the air
InSAR	Interferometric SAR: compares phase of two radar images to measure elevation or centimeter-scale ground motion	Radar imaging, interferometry and elevation models
Insolation	Incoming solar radiation received at a place (W/m²)	Solar flux at any distance
Interferogram	Image of phase differences between two radar images; each color cycle (fringe) is half a wavelength of line-of-sight motion	Radar imaging, interferometry and elevation models
Inverse-square law	Intensity from a point source falls as $1/d^2$	Refraction and the inverse-square law
Ionosphere	Ionized upper atmosphere (about 60 to 1,000 km) that reflects and delays radio waves	What the air is made of
Irradiance	Radiant flux arriving per unit area of a surface, W/m²	Radiometric quantities and units
Isostatic adjustment	Vertical land motion as the crust re-balances after loads change (glacial rebound after ice melts)	Sea-level rise and surface waves
Jamming	Deliberate or interfering transmission on a radar's frequency that masks real echoes	Weather, precipitation and Doppler radar
Jet stream	Narrow band of fast westerly winds near the tropopause	Climate drivers and circulation
Keeling curve	Continuous record of atmospheric CO₂ at Mauna Loa since 1958, rising with a seasonal sawtooth	Carbon and nitrogen cycles
Kelvin wave	Ocean wave trapped at the equator that carries warm water east during El Niño onset	El Niño–Southern Oscillation
Kirchhoff's law	At thermal equilibrium a body's emissivity equals its absorptivity at each wavelength	Emissivity, Kirchhoff's law and brightness temperature
La Niña	Cold phase of ENSO: strong trade winds and cold SST in the eastern equatorial Pacific	El Niño–Southern Oscillation
Lagrange point	One of five points where a small body keeps a fixed position relative to two large ones; L1 between Sun and Earth hosts DSCOVR	Other orbits: Molniya, Lagrange points, graveyards and decay
Lambertian surface	Ideal diffuse surface whose radiance is the same from every viewing direction	Reflection and polarization
Land cover	Physical material at the surface (forest, water, pavement)	Classifying land cover
Land use	How people use land (residential, agriculture, recreation)	Classifying land cover
Lapse rate	Rate temperature drops with height, about 6.5 °C/km on average in the troposphere	Feedbacks
Last return	Final lidar return from a pulse, usually from the ground	Lidar
Latency	Delay between observation and data delivery, or signal travel delay; GEO adds about 0.24 s one way	Orbits and satellites
Latent heat	Energy absorbed or released when water changes phase; released when vapor condenses	Earth's energy budget, term by term
Layover	Radar distortion where the top of a tall feature is imaged before its base, leaning toward the sensor	Radar imaging, interferometry and elevation models
Lidar	Light detection and ranging: active sensor timing laser pulses for range	Lidar
Limb sounding	Viewing the atmosphere edge-on at the horizon to profile gases with fine vertical resolution	Viewing geometry: nadir, limb, occultation and multi-angle
Longwave radiation	Thermal infrared emitted by Earth and atmosphere, about 4 to 100 µm	Earth's energy budget, term by term
Low Earth orbit (LEO)	Orbit below about 2,000 km; period about 88 to 127 minutes	Orbits and satellites
Low-pass filter	Spatial filter that averages neighbors, smoothing noise and blurring edges	Image processing pipeline
Luminosity	Total power radiated by a star (Sun about 3.8 × 10²⁶ W)	Solar flux at any distance
Map projection	Mathematical transformation of the curved Earth onto a flat map	Map projections and coordinate grids
Marginal information	Map border items: title, scale, legend, north arrow, date, datum, contour interval	Marginal information and symbology
Marine cloud brightening	Proposed geoengineering: spraying sea-salt particles into marine clouds to raise their albedo (Twomey effect)	Aerosols
Medium Earth orbit (MEO)	Orbits between LEO and GEO (about 2,000 to 35,786 km), e.g., GPS at about 20,200 km	Orbits and satellites
Mercator projection	Conformal cylindrical projection; straight compass lines but large area exaggeration at high latitudes	Map projections and coordinate grids
Mesosphere	Atmospheric layer from about 50 to 85 km, where temperature falls with height	What the air is made of
Methane (CH₄)	Greenhouse gas from wetlands, livestock, rice, landfills and fossil fuels; lifetime about 12 years	Greenhouse and trace gas reference
Mie scattering	Scattering by particles about the size of the wavelength (haze, smoke, dust); weak wavelength dependence, forward-peaked	Scattering and absorption in the air
Milankovitch cycles	Cyclic changes in eccentricity (about 100,000 yr), obliquity (41,000 yr) and precession (about 23,000 yr) that pace ice ages	Climate drivers and circulation
Mixel (mixed pixel)	Pixel containing more than one land-cover type, giving a blended spectrum	Pixels, mixed pixels and off-nadir growth
Molniya orbit	Highly elliptical, 63.4° inclined, 12-hour orbit that dwells over high northern latitudes	Other orbits: Molniya, Lagrange points, graveyards and decay
Multi-angle viewing	Observing a scene from several angles to retrieve aerosols, cloud heights and surface reflectance anisotropy (MISR)	Viewing geometry: nadir, limb, occultation and multi-angle
Multispectral	Imaging in several (typically 3 to 15) separate broad bands	Spectral classes of imagers
Nadir	Point on the ground directly below the sensor	Viewing geometry: nadir, limb, occultation and multi-angle
Natural color (true color)	Red, green and blue bands displayed as red, green and blue	Display types
NBR (Normalized Burn Ratio)	(NIR − SWIR)/(NIR + SWIR); drops sharply after fire; dNBR maps burn severity	Vegetation and spectral indices
NDSI	Normalized Difference Snow Index, (Green − SWIR)/(Green + SWIR); snow is bright in green and dark in SWIR	Vegetation and spectral indices
NDVI	Normalized Difference Vegetation Index, (NIR − Red)/(NIR + Red); dense healthy vegetation is high	Vegetation and spectral indices
NDWI	Normalized Difference Water Index; one form uses (Green − NIR)/(Green + NIR) to highlight water	Vegetation and spectral indices
Near-infrared (NIR)	About 0.7 to 1.3 µm; healthy leaves reflect strongly, water absorbs	The electromagnetic spectrum region by region
Net primary productivity (NPP)	Carbon fixed by photosynthesis minus plant respiration	Carbon and nitrogen cycles
Nitrification	Bacteria oxidize ammonium to nitrite and then nitrate	Carbon and nitrogen cycles
Nitrogen fixation	Conversion of N₂ gas to ammonia by bacteria, lightning or industry	Carbon and nitrogen cycles
Nitrous oxide (N₂O)	Long-lived greenhouse gas and ozone-depleting gas from fertilized soils, manure and industry	Greenhouse and trace gas reference
Nodes	The two points where an orbit crosses the equatorial plane (ascending and descending)	Sun-synchronous orbits and equator crossing times
Nonselective scattering	Scattering by particles much larger than the wavelength (cloud droplets); all visible wavelengths equally, so clouds look white	Scattering and absorption in the air
Oblique photograph	Aerial photo taken with the camera tilted; high oblique includes the horizon	Photographic scale
Occultation	Viewing the Sun, a star or a GPS signal through the atmosphere's limb as it sets or rises to profile it	Viewing geometry: nadir, limb, occultation and multi-angle
Ocean acidification	Drop in seawater pH as the ocean absorbs CO₂	Carbon and nitrogen cycles
Ocean color	Spectrum of light leaving the ocean, used to estimate chlorophyll, sediment and dissolved matter	Ocean color and phytoplankton: PACE and MODIS
Off-nadir	Viewing away from straight down; pixels grow larger and relief displacement increases	Pixels, mixed pixels and off-nadir growth
Omission error	Pixels of a class missed by the map (false negatives); its complement is producer's accuracy	Classifying land cover
Opacity	Degree to which a medium blocks radiation; the complement of transmittance	Absorption bands and atmospheric windows
Optical depth	Dimensionless measure of extinction along a path; transmittance $= e^{-\tau}$	Attenuation and the Beer-Lambert law
Orbital decay	Gradual lowering of an orbit from atmospheric drag until reentry	Other orbits: Molniya, Lagrange points, graveyards and decay
Orbital period	Time to complete one orbit; $T^2 \propto a^3$ (Kepler's third law)	Kepler's laws and orbital calculations
Orthophoto (orthoimage)	Aerial photo corrected for tilt and relief displacement so it has uniform map scale	Geometric distortions of scanned and photographed images
Outgassing	Release of dissolved gas from water (e.g., CO₂ from warming or upwelled ocean water)	Carbon and nitrogen cycles
Ozone hole	Springtime severe depletion of stratospheric ozone over Antarctica, below 220 DU	Ozone: shield above, pollutant below
Ozone layer	Stratospheric ozone, mostly 15 to 35 km, that absorbs UV-B and UV-C	Ozone: shield above, pollutant below
Panchromatic	Single broad band spanning much of the visible (often into NIR), recorded as grayscale at fine resolution	Spectral classes of imagers
Parallax	Apparent shift in an object's position when viewed from two points; used to measure heights from stereo pairs	Elevation, altitude and slope
Passive microwave	Radiometers that measure natural microwave emission (sea ice, soil moisture, rain, SST)	Imagers, radiometers and spectrometers
Passive sensor	Sensor that records naturally reflected sunlight or emitted radiation	Active and passive sensors classified
Path radiance	Radiance scattered into the sensor's view by the atmosphere without reaching the ground; causes haze	Scattering and absorption in the air
Peak wavelength	Wavelength of maximum blackbody emission; $\lambda_{max} = b/T$ with $b \approx 2898$ µm·K	Blackbodies and thermal emission
Perigee	Point of an Earth orbit closest to Earth	Kepler's laws and orbital calculations
Perihelion	Point in Earth's orbit closest to the Sun (early January)	Kepler's laws and orbital calculations
Permafrost	Ground that stays at or below 0 °C for at least two consecutive years	Climate change effects on weather and ecosystems
Phenology	Timing of seasonal biological events (green-up, flowering, leaf fall)	Vegetation and spectral indices
Photoelectric effect	Light ejects electrons from a metal only above a threshold frequency; Einstein's 1905 evidence for photons	Wave basics and photon energy
Photogrammetry	Measuring distances, heights and positions from photographs	Elevation, altitude and slope
Photon	Quantum of EM energy, $E = hf = hc/\lambda$	Wave basics and photon energy
Phytoplankton	Microscopic floating photosynthetic organisms; about half of global primary production	Ocean color and phytoplankton: PACE and MODIS
Pixel	Smallest picture element of a digital image, holding one value per band	Pixels, mixed pixels and off-nadir growth
Planck's law	Spectral radiance of a blackbody as a function of wavelength and temperature	Planck's law and the shape of the curves
Point cloud	Set of 3-D points (x, y, z) produced by lidar or photogrammetry	Lidar
Polar amplification	Faster warming at high latitudes than the global average, mainly from ice-albedo and lapse-rate feedbacks	Climate drivers and circulation
Polar orbit	Orbit with inclination near 90° passing over or near both poles	Orbits and satellites
Polar stratospheric clouds	Very cold (below about −78 °C) stratospheric clouds whose surfaces activate chlorine to destroy ozone	Ozone: shield above, pollutant below
Polar vortex	Strong westerly winds circling the pole in winter; the stratospheric vortex isolates cold air where polar stratospheric clouds form	Ozone: shield above, pollutant below
Polarization	Orientation of a wave's electric field oscillation (horizontal H, vertical V, circular)	Reflection and polarization
Positive feedback	Response that amplifies the initial change (water vapor, ice-albedo)	Feedbacks
Precession (axial)	Wobble of Earth's spin axis, about 26,000-year cycle (about 23,000 yr climatic precession)	Climate drivers and circulation
Principal point	Center of an aerial photograph, where lines between opposite fiducial marks cross	Photographic scale
Prograde orbit	Orbit in the direction of Earth's rotation (inclination under 90°)	Kepler's laws and orbital calculations
Proxy data	Indirect climate records (tree rings, ice cores, corals, sediments) used before instruments	Climate drivers and circulation
Pseudocolor	Color ramp applied to a single band or data layer	Display types
Pulse length	Duration of a radar or lidar pulse; range resolution is $c\tau/2$	Radar imaging, interferometry and elevation models
Pulse repetition frequency (PRF)	Number of radar or lidar pulses sent per second	Lidar
Pushbroom scanner	Along-track scanner using a linear detector array; no moving mirror	Scanning systems
Quantum efficiency	Fraction of incident photons that a detector converts into electrons	Detectors
Radar	Radio detection and ranging: active microwave sensor measuring echo time and strength	Radar imaging, interferometry and elevation models
Radar shadow	Area behind a steep slope that the radar beam cannot reach; appears black	Radar imaging, interferometry and elevation models
Radial velocity	Component of motion toward or away from a radar, measured by Doppler shift	Weather, precipitation and Doppler radar
Radiance	Radiant flux per unit area per unit solid angle, W/(m²·sr); what a sensor measures	Radiometric quantities and units
Radiative forcing	Change in net top-of-atmosphere energy flux caused by a driver, in W/m²; positive warms	Radiative forcing
Radiometer	Instrument that measures radiant energy quantitatively in one or more bands	Imagers, radiometers and spectrometers
Radiometric correction	Removing sensor and atmospheric errors in brightness values (calibration, striping, atmospheric effects)	Image processing pipeline
Radiometric resolution	Ability to distinguish small differences in energy, set by bit depth and noise	The four resolutions and their trade-offs
Radiosity	Total radiant flux leaving a surface per unit area (emitted plus reflected), W/m²	Radiometric quantities and units
Radiosonde	Balloon-borne instrument package measuring temperature, humidity, pressure and wind	Reading climate data products
Raman scattering	Inelastic scattering where photons gain or lose energy to molecular vibrations; used in Raman lidar	Scattering and absorption in the air
Range	Distance from sensor to target, found from echo time: $R = ct/2$	Radar imaging, interferometry and elevation models
Range resolution	Smallest separation along the radar look direction that can be resolved, $c\tau/2$ (slant)	Radar imaging, interferometry and elevation models
Rayleigh scattering	Scattering by molecules much smaller than the wavelength; strength $\propto 1/\lambda^4$, so the sky is blue	Scattering and absorption in the air
Rayleigh-Jeans law	Classical long-wavelength approximation to Planck's law; diverges at short wavelengths (ultraviolet catastrophe)	Planck's law and the shape of the curves
Real aperture radar (RAR)	Side-looking radar whose azimuth resolution is set by its physical antenna length	Radar imaging, interferometry and elevation models
Reanalysis	Consistent gridded record of past weather made by assimilating observations into a fixed weather model	Reading climate data products
Red edge	Steep rise in vegetation reflectance between about 680 and 750 nm	Vegetation and spectral indices
Reflectance	Fraction of incident energy reflected by a surface at a wavelength	Surface properties: albedo, temperature and materials
Reflectivity (radar, Z)	Weather radar measure of echo strength from precipitation, in dBZ	Weather, precipitation and Doppler radar
Refraction	Bending of a wave as it crosses between media of different refractive index	Refraction and the inverse-square law
Relief displacement	Tall objects appear to lean outward from the photo center (principal point) in vertical photos	Geometric distortions of scanned and photographed images
Remote sensing	Acquiring information about an object without physical contact, usually by recording EM radiation	Event format
Repeat cycle	Time for a satellite to return to exactly the same ground track	Coverage, swath and revisit
Resampling	Computing new pixel values on a new grid (nearest neighbor, bilinear, cubic convolution) during geometric correction	Image processing pipeline
Residence time	Average time an atom or molecule stays in a reservoir: reservoir size / flux in or out	Carbon and nitrogen cycles
Retrieval	Inferring a geophysical quantity (temperature, gas amount, AOD) from measured radiances, usually by model inversion	Spectroscopy and gas retrieval
Retrograde orbit	Orbit opposite to Earth's rotation (inclination over 90°); sun-synchronous orbits are slightly retrograde	Sun-synchronous orbits and equator crossing times
Revisit time	Time between successive observations of the same place (may use off-nadir pointing)	Coverage, swath and revisit
Roughness (radar)	Surface texture relative to wavelength; rough surfaces backscatter strongly, smooth ones reflect away (dark)	Radar imaging, interferometry and elevation models
Saturation	Pixel recording the maximum DN because the signal exceeded the sensor's range; also color purity	Bit depth, dynamic range and saturation
SAVI	Soil-Adjusted Vegetation Index; adds a soil factor L (often 0.5) to NDVI for sparse vegetation	Vegetation and spectral indices
Scale (map)	Ratio of map distance to ground distance (e.g., 1:24,000)	Scale, distance, perimeter and area
Scan line corrector (SLC)	Landsat 7 ETM+ mirror that compensated for forward motion; its 2003 failure left wedge gaps	Platforms
Scattering	Redirection of radiation by particles or molecules	Scattering and absorption in the air
Scatterometer	Radar that measures backscatter precisely from several angles; ocean winds from surface roughness	Radar altimetry and scatterometry
Sea ice	Frozen seawater floating on the ocean; extent is mapped daily by passive microwave	Earth processes: glaciers, rivers and stream patterns
Sea-surface temperature (SST)	Temperature of the top layer of the ocean, measured by thermal IR and microwave radiometers	Temperature: SST, land surface temperature and profiles
Semi-synchronous orbit	Orbit with a period of half a sidereal day (about 12 h), e.g., GPS	Other orbits: Molniya, Lagrange points, graveyards and decay
Sensible heat	Heat transferred by conduction and convection that changes temperature	Earth's energy budget, term by term
Shortwave infrared (SWIR)	About 1.3 to 3 µm; sensitive to leaf and soil moisture, minerals, burn scars	The electromagnetic spectrum region by region
Shortwave radiation	Solar radiation, about 0.2 to 4 µm	Earth's energy budget, term by term
Side-looking radar (SLAR)	Imaging radar pointed to one side of the flight path so range differences separate targets	Radar imaging, interferometry and elevation models
Sidereal day	One rotation relative to the stars, 23 h 56 min 4 s	Geostationary versus geosynchronous
Signal-to-noise ratio (SNR)	Ratio of signal strength to noise; higher means cleaner data	Signal-to-noise ratio and sensitivity
Sink	Reservoir or process that removes a substance from the atmosphere (ocean and land absorb CO₂)	Carbon and nitrogen cycles
Site	Location of a feature relative to its surroundings, used as an interpretation clue	Elements of image interpretation
Slant range	Direct line-of-sight distance from radar antenna to target	Radar imaging, interferometry and elevation models
Snell's law	$n_1 \sin\theta_1 = n_2 \sin\theta_2$	Refraction and the inverse-square law
Soil moisture	Water held in the soil; measured by L-band passive microwave (SMAP) and radar	Platforms
Solar constant (total solar irradiance)	Solar flux at 1 AU above the atmosphere, about 1,361 W/m²	Solar flux at any distance
Solar zenith angle	Angle between the Sun and the local vertical	Sun-synchronous orbits and equator crossing times
Solar-induced fluorescence (SIF)	Faint red and far-red glow emitted by chlorophyll during photosynthesis, measured by OCO-2	Spectroscopy and gas retrieval
Sonar	Sound navigation and ranging: active acoustic sensing used for bathymetry	Active and passive sensors classified
Sounder	Instrument that measures vertical profiles of temperature, humidity or gases using many channels	Imagers, radiometers and spectrometers
Southern Oscillation	Seesaw in surface pressure between Tahiti and Darwin; atmospheric half of ENSO	El Niño–Southern Oscillation
Southern Oscillation Index (SOI)	Standardized Tahiti minus Darwin pressure difference; strongly negative during El Niño	El Niño–Southern Oscillation
Spatial resolution	Size of the smallest feature a sensor can distinguish, often quoted as ground pixel size	The four resolutions and their trade-offs
Speckle	Grainy salt-and-pepper noise in radar images from interference of returns within a cell	Radar imaging, interferometry and elevation models
Spectral reflectance curve	Graph of reflectance versus wavelength for a material	Surface properties: albedo, temperature and materials
Spectral resolution	Number and narrowness of bands a sensor records	The four resolutions and their trade-offs
Spectral signature	Characteristic pattern of reflectance or emission across wavelengths that identifies a material	Surface properties: albedo, temperature and materials
Spectral unmixing	Estimating the fractions of endmember materials within a mixed pixel	Pixels, mixed pixels and off-nadir growth
Spectrometer	Instrument that splits light into many narrow wavelengths to measure a spectrum	Imagers, radiometers and spectrometers
Spectroradiometer	Spectrometer calibrated to measure radiance at each wavelength (MODIS = Moderate Resolution Imaging Spectroradiometer)	Imagers, radiometers and spectrometers
Specular reflection	Mirror-like reflection from a smooth surface; angle of reflection equals angle of incidence	Reflection and polarization
Stefan-Boltzmann law	Total power per area from a blackbody, $\sigma T^4$ with $\sigma = 5.67 \times 10^{-8}$ W/(m²·K⁴)	Blackbodies and thermal emission
Steradian (sr)	Unit of solid angle; a sphere is $4\pi$ sr	Radiometric quantities and units
Stereo pair	Two overlapping images of the same area from different positions, viewed together for 3-D	Elevation, altitude and slope
Steric sea-level change	Sea-level change from density changes (thermal expansion and salinity) with no change in water mass	Sea-level rise and surface waves
Storm surge	Rise of sea level pushed ashore by storm winds and low pressure	Hurricanes and severe storms
Stratosphere	Layer from the tropopause (about 8 to 16 km) to about 50 km; temperature rises with height because ozone absorbs UV	What the air is made of
Stratospheric aerosol injection	Proposed geoengineering: adding sulfate aerosols to the stratosphere to reflect sunlight, mimicking volcanoes	Aerosols
Striping	Image artifact of regular lines caused by detectors with mismatched calibration	Image processing pipeline
Sulfate aerosol	Fine particles from SO₂ (fossil fuels, volcanoes); reflect sunlight and brighten clouds, cooling	Aerosols
Sun glint	Specular reflection of sunlight from water toward the sensor, making water bright	Reflection and polarization
Sun-synchronous orbit (SSO)	Near-polar, slightly retrograde orbit whose plane precesses about 1° per day so it crosses each latitude at the same local solar time	Sun-synchronous orbits and equator crossing times
Superspectral	Between multispectral and hyperspectral: more than about 10 narrow bands (e.g., MODIS with 36)	Spectral classes of imagers
Supervised classification	Classification trained on analyst-chosen samples of known classes	Classifying land cover
Surface roughness	Height variation of a surface relative to wavelength; controls radar backscatter and specular reflection	Radar imaging, interferometry and elevation models
Swath	Width of the ground strip imaged on one pass	IFOV, GSD, FOV and swath
Synthetic aperture radar (SAR)	Radar that combines echoes along the flight path to simulate a long antenna, giving fine azimuth resolution independent of range	Radar imaging, interferometry and elevation models
Target	Object or area being sensed	Active and passive sensors classified
Teleconnection	Linked climate anomalies in distant regions (e.g., ENSO affecting U.S. winter weather)	El Niño–Southern Oscillation
Temperature inversion	Layer where temperature increases with height, trapping pollutants below	Climate drivers and circulation
Temporal resolution	How often a sensor images the same place	The four resolutions and their trade-offs
Texture	Frequency of tonal change in an image: smooth or rough appearance	Elements of image interpretation
Thermal expansion	Increase in seawater volume as it warms; a major part of sea-level rise	Sea-level rise and surface waves
Thermal inertia	Resistance of a material to temperature change; high (water, rock) means small day-night swings	Surface properties: albedo, temperature and materials
Thermal infrared (TIR)	About 3 to 14 µm in remote sensing (mainly 8 to 14 µm), where Earth emits	The electromagnetic spectrum region by region
Thermocline	Layer where ocean temperature drops quickly with depth	El Niño–Southern Oscillation
Thermohaline circulation	Deep ocean circulation driven by density differences from temperature and salinity	Climate drivers and circulation
Thermosphere	Layer above about 85 km where temperature rises sharply with height due to absorption of extreme UV	What the air is made of
Tipping point	Threshold beyond which a system shifts to a new state, often irreversibly	Feedbacks
Total internal reflection	Complete reflection inside a denser medium when the incidence angle exceeds the critical angle	Refraction and the inverse-square law
Trade winds	Easterly surface winds between about 30° N and 30° S	El Niño–Southern Oscillation
Training site	Area of known land cover used to define class signatures in supervised classification	Classifying land cover
Transmittance	Fraction of radiation that passes through a medium	Attenuation and the Beer-Lambert law
Tropopause	Boundary between troposphere and stratosphere, about 8 km at the poles to 16 km at the equator	What the air is made of
Troposphere	Lowest atmospheric layer, where weather occurs and temperature usually drops with height	What the air is made of
Turbidity	Cloudiness of water from suspended sediment or particles	Ocean color and phytoplankton: PACE and MODIS
Twomey effect	More aerosol makes more, smaller cloud droplets, so clouds reflect more sunlight (cloud albedo effect)	Aerosols
Ultraviolet catastrophe	Classical prediction that blackbody emission grows without limit at short wavelengths; resolved by Planck	Planck's law and the shape of the curves
Ultraviolet (UV)	About 10 to 400 nm; UV-A 315 to 400, UV-B 280 to 315, UV-C 100 to 280 nm	The electromagnetic spectrum region by region
Unsupervised classification	Computer groups pixels into spectral clusters first; the analyst labels them afterward	Classifying land cover
Upwelling	Rise of cold, nutrient-rich deep water to the surface (coasts, equator)	Climate drivers and circulation
Vegetation index	Band combination (NDVI, EVI, SAVI) sensitive to vegetation amount or health	Vegetation and spectral indices
Vertical exaggeration	Ratio of vertical to horizontal scale in a profile or stereo view, making relief look steeper	Elevation, altitude and slope
Volatile organic compounds (VOCs)	Carbon gases that evaporate easily; with NOx and sunlight they form ground-level ozone	Ozone: shield above, pollutant below
Volume scattering	Multiple scattering inside a volume (forest canopy, dry snow); gives cross-polarized radar return	Radar imaging, interferometry and elevation models
Walker circulation	East-west overturning in the tropical Pacific atmosphere: rising over Indonesia, sinking over the eastern Pacific	El Niño–Southern Oscillation
Water vapor feedback	Warmer air holds more water vapor (about 7% per °C), a greenhouse gas, amplifying warming; strongest positive feedback	Feedbacks
Water vapor imagery	Imagery near 6.2 to 7.3 µm showing mid- and upper-tropospheric moisture, not the surface	Cloud imagery: visible, infrared and water vapor
Waveform	Recorded strength of a radar or lidar return versus time	Lidar
Wavelength	Distance between successive wave crests; $\lambda = c/f$	Wave basics and photon energy
Wavenumber	Number of waves per unit length, $1/\lambda$, usually in cm⁻¹ in spectroscopy	Wave basics and photon energy
Whiskbroom scanner	Across-track scanner with a rotating or oscillating mirror	Scanning systems
Wien's displacement law	$\lambda_{max} = 2898\ \mu\text{m·K} / T$; hotter bodies peak at shorter wavelengths	Blackbodies and thermal emission
Zenith	Point directly overhead an observer	Viewing geometry: nadir, limb, occultation and multi-angle

People and milestones

ReferencePeople and milestones· 80 rows

80 rows

Year	Person or event	Why it matters
1609, 1619	Johannes Kepler publishes his laws of planetary motion (first two in 1609, the third in 1619)	Orbits are ellipses, equal areas in equal times, and $T^2 \propto a^3$: the basis of every orbit calculation
1621	Willebrord Snell (Snellius) finds the law of refraction	$n_1 \sin\theta_1 = n_2 \sin\theta_2$ governs lenses and atmospheric bending of light
1666	Isaac Newton splits white light with a prism (published in Opticks, 1704)	Showed white light is a mixture of colors: the first look at the visible spectrum
1687	Newton's Principia	Law of universal gravitation, which explains why satellites orbit and fixes orbital speed and period
1800	William Herschel discovers infrared	A thermometer beyond the red end of a prism spectrum warmed most: invisible "calorific rays"
1801	Johann Wilhelm Ritter discovers ultraviolet	Silver chloride darkened fastest beyond the violet end of the spectrum
1808	Étienne-Louis Malus discovers polarization by reflection	Light reflected from glass is polarized: the basis of polarimetry
1814	Joseph von Fraunhofer maps dark lines in the solar spectrum	Absorption lines that Kirchhoff and Bunsen later tied to specific elements
1815	David Brewster states Brewster's law	At $\tan\theta_B = n_2/n_1$ reflected light is fully polarized
1824	Joseph Fourier argues the atmosphere warms Earth	Earliest statement of what became the greenhouse effect
1835	Gaspard-Gustave de Coriolis describes the Coriolis effect	Apparent deflection by Earth's rotation, which shapes winds and ocean currents
1842	Christian Doppler describes the Doppler effect	Frequency shift with motion, used by Doppler weather radar and scatterometers
1852	August Beer relates absorption to concentration	Completes the Beer-Lambert law used in gas retrievals
1856	Eunice Foote shows CO₂-filled air heats more in sunlight	First experiment linking CO₂ to warming
1858	Gaspard-Félix Tournachon ("Nadar") takes the first aerial photograph, from a tethered balloon near Paris	Start of aerial remote sensing (the original image is lost)
1859	John Tyndall measures infrared absorption by gases	Showed water vapor and CO₂ absorb heat radiation while N₂ and O₂ do not
1859 to 1860	Gustav Kirchhoff's law of thermal radiation (and spectroscopy with Robert Bunsen)	Emissivity equals absorptivity; introduced the idea of a blackbody
1860	James Wallace Black photographs Boston from a balloon	Oldest surviving aerial photograph
1861	James Clerk Maxwell (with Thomas Sutton) shows the first color photograph, a tartan ribbon	Three-color (red, green, blue) principle behind every color composite
1865	Maxwell's theory of electromagnetism	Predicted EM waves traveling at the speed of light; light is an EM wave
1871	Lord Rayleigh explains the blue sky	Scattering by particles much smaller than the wavelength goes as $1/\lambda^4$
1879, 1884	Josef Stefan finds, and Ludwig Boltzmann derives, the $T^4$ law	Stefan-Boltzmann law: total emitted power $\sigma T^4$
1887	Heinrich Hertz produces and detects radio waves	Confirmed Maxwell's EM waves; the unit of frequency is named after him
1893	Wilhelm Wien's displacement law	Peak wavelength $\propto 1/T$: hotter bodies peak at shorter wavelengths
1896	Svante Arrhenius calculates CO₂ warming	First quantitative estimate of warming from more CO₂ (several °C for doubling)
1900	Max Planck's radiation law	Energy comes in quanta $E = hf$; fixed the classical ultraviolet catastrophe
1904	Christian Hülsmeyer demonstrates the "telemobiloscope"	Earliest radio-echo device for detecting ships, a forerunner of radar
1905	Albert Einstein explains the photoelectric effect	Light behaves as particles (photons); basis of photon detectors
1907 to 1908	Julius Neubronner's pigeon camera (patent applied 1907, granted 1908)	Small timed cameras strapped to homing pigeons took aerial photos
1908	Gustav Mie's scattering theory	Exact solution for scattering by spheres comparable to the wavelength: Mie scattering
1909	Wilbur Wright carries a cameraman in Italy	Earliest photography (motion pictures) from an airplane
1923	Arthur Compton's X-ray scattering	Inelastic photon-electron scattering confirmed photons carry momentum
1924	Gilbert Walker names the Southern Oscillation	The atmospheric half of ENSO; Bjerknes later named the Walker circulation for him
1920s	Gordon Dobson builds the ozone spectrophotometer	Ground network for total ozone; the Dobson unit is named for him
1928	C. V. Raman discovers Raman scattering	Inelastic molecular scattering, used by Raman lidar (Nobel Prize 1930)
1928	Thomas Midgley Jr. develops CFCs for Frigidaire (General Motors)	Non-toxic, non-flammable refrigerants that later destroyed stratospheric ozone
1930	Sydney Chapman's ozone cycle	Photochemical creation and destruction of stratospheric ozone by UV
1935	Robert Watson-Watt demonstrates radio detection of aircraft	Led to practical radar in Britain before World War II
1938	Guy Callendar links rising CO₂ to observed warming	First to tie fossil-fuel CO₂ to measured temperature rise
1941	Milutin Milankovitch publishes his Canon of Insolation	Orbital cycles (eccentricity, tilt, precession) pace the ice ages
1940s	Color-infrared film developed in World War II	Healthy vegetation shows red, which exposed painted camouflage; became the CIR display
1957	Sputnik 1 (October 4)	First artificial satellite (Soviet Union)
1957	Roger Revelle and Hans Suess	Showed the ocean cannot absorb fossil CO₂ as fast as emitted; humans conducting a "geophysical experiment"
1958	Explorer 1 (January 31)	First U.S. satellite; discovered the Van Allen radiation belts
1958	Charles David Keeling begins CO₂ measurements at Mauna Loa	The Keeling curve: rising CO₂ with a seasonal cycle
Late 1950s	Evelyn Pruitt (Office of Naval Research) coins "remote sensing"	Replaced "aerial photography" as sensing moved beyond film and visible light (dates given range 1958 to 1960)
1959	Explorer 6 (August 14)	First photograph of Earth from orbit
1959	Explorer 7 (October 13) carries Verner Suomi's radiometer	First successful satellite measurement of Earth's radiation budget; Suomi became "father of satellite meteorology"
1960	TIROS-1 (April 1)	First successful weather satellite (TV cameras)
1960	Theodore Maiman builds the first laser	Made lidar possible
1960	Corona (Discoverer 14) returns the first film from orbit	First satellite reconnaissance imagery (declassified 1995)
1963, 1964	Syncom 2 and Syncom 3, designed by Harold Rosen's Hughes team	First geosynchronous (1963) and first geostationary (1964) satellites
1966	ATS-1 and Suomi's spin-scan cloud camera	First full-disk images of Earth from geostationary orbit
1967	Syukuro Manabe and Richard Wetherald's model	First realistic estimate of warming for doubled CO₂ (about 2 °C); Manabe shared the 2021 Nobel Prize in Physics
1969	Jacob Bjerknes links El Niño to the Southern Oscillation	Showed the ocean and atmosphere act as one coupled system: ENSO
1972	Landsat 1 (ERTS-1), July 23	First civilian Earth-resources satellite; Virginia Norwood ("Mother of Landsat") designed its Multispectral Scanner
1972	Blue Marble photographed on Apollo 17 (December 7)	Full sunlit Earth in one frame
1974	Mario Molina and F. Sherwood Rowland	Showed CFCs release chlorine that destroys stratospheric ozone (Nobel Prize 1995 with Paul Crutzen)
1974	Sean Twomey describes the aerosol cloud-albedo effect	More aerosol, more and smaller droplets, brighter clouds: the Twomey effect
1974, 1975	SMS-1 and GOES-1	First NOAA geostationary weather satellites
1978	Seasat (June 27)	First civilian satellite SAR and first ocean-dedicated satellite; failed after 106 days
1978	TIROS-N and Nimbus-7	TIROS-N flew the first AVHRR; Nimbus-7 flew TOMS (ozone) and CZCS (first ocean color sensor)
1979	Charney report (U.S. National Academy of Sciences)	Climate sensitivity of 3 °C ± 1.5 °C for doubled CO₂
1982	Landsat 4	First Thematic Mapper (7 bands, 30 m)
1983	Andrew Watson and James Lovelock's Daisyworld	Toy model of biological albedo feedback regulating climate
1985	Joseph Farman, Brian Gardiner and Jonathan Shanklin (British Antarctic Survey)	Reported the Antarctic ozone hole from ground Dobson data
1986	SPOT 1 (France)	First satellite pushbroom imager; off-nadir stereo
1987	Montreal Protocol	Phases out ozone-depleting substances; the most successful environmental treaty
1988	IPCC founded by WMO and UNEP; James Hansen's Senate testimony	Climate change becomes a world policy issue
1991	Mount Pinatubo erupts; ERS-1 launched	Pinatubo's sulfate cooled the globe about 0.5 °C in 1992; ERS-1 was ESA's first SAR satellite
1992	UNFCCC signed at Rio; TOPEX/Poseidon launched	Climate treaty framework; start of the precise satellite sea-level record
1997	Kyoto Protocol; TRMM launched	First binding emissions targets (in force 2005); first space-borne rain radar
1999	IKONOS (September 24), Terra, Landsat 7	First commercial 1 m imagery; flagship of NASA's Earth Observing System
2000	Shuttle Radar Topography Mission	Near-global InSAR elevation model
2002	GRACE and Aqua launched	Gravity measurement of ice and groundwater mass; Aqua starts the A-Train
2006	CloudSat and CALIPSO launched together	Radar and lidar vertical profiles of clouds and aerosols
2014	Sentinel-1A and OCO-2	First Copernicus satellite; NASA's first CO₂-dedicated satellite to reach orbit
2015	Paris Agreement	Goal of holding warming well below 2 °C, pursuing 1.5 °C (in force 2016)
2016	Kigali Amendment; GOES-16 launched	HFC phase-down under the Montreal Protocol; first GOES-R satellite with ABI
2024	PACE launched (February 8)	Hyperspectral ocean color to identify phytoplankton types

Common pitfalls

How tournaments test this

What the bank asks most. In scio.ly's question bank (about 6,200 Remote Sensing questions from both divisions), the largest topic groups are radar and lidar (about 11%), sensor types and active versus passive (8%), greenhouse gases and aerosols (7%), what each Division C satellite and instrument measures (7%), the electromagnetic spectrum (6%), orbit types (6%), scattering and absorption (6%), blackbody laws (5%) and energy balance and albedo (4%). Map measurement, land cover, geohazards and vegetation indices are each 1 to 3% of the bank but carry most of the Division B interpretation points, which are usually built around images the bank cannot store. Expect many recall questions on acronyms, launch years and "which satellite carries which instrument": the Satellites and instruments reference is built for them.

Scio.ly's question archive holds 7,568 archived questions in the remote sensing scope. The practice bank's 6,245 questions are mostly exports of these archive questions (about 4,900); about 400 come from a 2025 import of tests the archive also holds, and about 900 from older tests the archive does not have. Of the archive, 5,568 are tagged Division C, 1,812 Division B and 188 shared. The 2026 season supplies 4,571 archive questions (about 60%), with a tail from earlier seasons, including 1,168 from 2018 and 946 from 2023. Older questions were written for earlier topics, so do not assume each matches the 2027 rules. In particular, the 2026 Division B list included azimuths, bearings and latitude-longitude conversion, which the 2027 Observable Earth list no longer includes, and the 2026 Division C list included the Rayleigh-Jeans law and ATMS, which are also gone; skip archive questions on those.

Question types. Free response makes up 3,961 archive items (52%) and multiple choice 3,181 (42%); the rest are multiple-select (186), numeric (133), true or false (100) and matching (7). The bank is 55% multiple choice (3,465 of 6,245) and 39% free response (2,414).

Task profile. The most common task is explaining why something happens (12.5% of archive items), then "which of the following" (8.6%), interpreting a graph, table or curve (5.8%) and calculating a value (4.4%). Comparison (1.8%) and definitions (1.2%) are smaller. The bank shows a similar pattern.

Subtopics. Archive tags overlap: climate processes 2,878 (about 38%), data analysis 1,947 (26%), sensors 1,716 (23%), imaging 1,520 (20%), satellites 1,268 (17%) and mapping only 447. The tags skew toward Division C because that division supplies most of the archive; plan study time from the 2027 percentages, not tag counts.

Recurring terms. Among two-word terms, "remote sensing" itself leads (649 questions), then surface temperature (253), climate change (188), electromagnetic spectrum (175), greenhouse gas (165), spatial resolution (163), radiative forcing (105), equilibrium temperature (104), Rayleigh scattering (90), radiometric resolution (71) and Stefan-Boltzmann law (69).

Most common correct answers. Rayleigh scattering (54 questions), MODIS (45), infrared (44), Mie scattering (41), passive (35), water vapor (34), OCO-2 (32), active (30), visible (27), geostationary orbit (25), lidar (25) and carbon dioxide (24). Classification words (active, passive, infrared, visible) and instrument names dominate: when a question describes a measurement, it often expects you to name the platform or instrument.

Test yourself

The questions below are real archive questions whose official keys were re-solved while writing this guide, chosen because their content is inside the 2027 rules. Each links to its public practice test. Try them before reading the explanations.

Test yourselfTest yourself: sensors, light and orbits

0 / 0 correct

1. 1

   Which of the following instruments are classified as Active Sensors?

   Question 1 options

   ARadar (radio detection and ranging)BLiDAR (light detection and ranging)CImaging RadiometerDLaser altimeter
   Check answer

   From: Raisbeck Aviation Invitational - WA, 2025-26, Division C

2 questions from this set are no longer available.

Practice more Remote Sensing questions →

Test yourselfTest yourself: light, scattering and sensors

0 / 0 correct

1. 1

   According to the Stefan-Boltzmann law, the total energy radiated per second per unit by a black body is directly proportional to which of the following?

   Question 1 options

   AThe square of its absolute temperatureBThe cube of its absolute temperatureCThe fourth power of its absolute temperatureDThe logarithm of its absolute temperature

   From: CSE (Captains Summer Exchange) 2025-26, Division C

Practice more Remote Sensing questions →

Test yourselfTest yourself: orbits, resolution and satellites

0 / 0 correct

1. 1

   Which orbit provides global coverage with nearly constant local solar time?

   Question 1 options

   AGeostationaryBHighly ellipticalCSun-synchronousDMedium Earth

   From: Mira Loma Invitational - CA 2025-26, Division C

1 question from this set is no longer available.

Practice more Remote Sensing questions →

Test yourselfTest yourself: maps, land cover and vegetation

0 / 0 correct

1. 1

   What does the term "nadir" refer to?

   Question 1 options

   AThe ground surface area that forms one pixel in the imageBThe point on the Earth's surface directly below the observing satelliteCA known point that can be used as a reference point for all other locationsDA coordinate-based local, regional, or global system used to locate geographical entities

   From: Pembroke Hill Invitational 2025-26, Division B

Practice more Remote Sensing questions →

Test yourselfTest yourself: climate processes and energy balance

0 / 0 correct

1. 1

   Which gas is responsible for the majority of anthropogenic greenhouse warming?

   Question 1 options

   AOzoneBMethaneCCarbon dioxideDSulfur dioxide

   From: Flint Hill Invitational - VA 2025-26, Division C

1 question from this set is no longer available.

Practice more Remote Sensing questions →

Condensed facts

A condensed page of the facts and formulas tests ask most. Put it at the front of your binder.

ReferenceFormula summary· 26 rows

26 rows

Formula	Use	Div
$c = \lambda f$; $E = hf = hc/\lambda$	Wavelength, frequency, photon energy	B, C
$\lambda_{max} = b/T$	Peak emission wavelength	B, C
$M = \varepsilon\sigma T^{4}$	Emitted power per area	B concept, C calc
$I \propto 1/\lambda^{4}$	Rayleigh scattering ratio	C
$I = I_0 e^{-kx}$	Attenuation (Beer-Lambert)	C
$n_1 \sin\theta_1 = n_2 \sin\theta_2$; $v = c/n$	Refraction	B concept, C calc
$E \propto 1/d^{2}$	Inverse-square law	B concept, C calc
$R = c\,\Delta t/2$	Radar or lidar range	B, C
$\Delta f = 2v_r f/c$	Doppler radar	C
One InSAR fringe $= \lambda/2$ of line-of-sight motion	Deformation	B (read), C
$T = 2\pi\sqrt{a^{3}/GM}$; $v = \sqrt{GM/r}$	Orbital period and speed ($a$ from Earth's center)	B concept, C calc
GSD $\approx H \times$ IFOV; swath $= 2H\tan(\text{FOV}/2)$	Pixel size and swath	C
$\Delta\lambda = 360° \times T/1440$ min; frames/s $= v_g/L$	Ground-track shift, frame rate	C
$\theta \approx 1.22\lambda/D$	Diffraction limit	C
Levels $= 2^{n}$ bits	Radiometric resolution	C
SNR$_{dB} = 10\log_{10}(P_S/P_N)$	Signal-to-noise	C
Scale $= f/H$	Air-photo scale	B
Ground distance $=$ map distance × denominator; area × denominator²	Map measurements	B
Slope $= \Delta h/d$ (× 100 for %)	Gradient	B
NDVI $= (\text{NIR} - \text{Red})/(\text{NIR} + \text{Red})$	Vegetation	B
NBR $= (\text{NIR} - \text{SWIR2})/(\text{NIR} + \text{SWIR2})$; dNBR = before − after	Burn severity	B
$S(1-A)/4 = \varepsilon\sigma T_e^{4}$	Planetary equilibrium temperature	C
$T_s = T_e[2/(2-\varepsilon)]^{1/4}$; $T_s = (n+1)^{1/4}T_e$	One-layer and n-layer greenhouse models	C
$\Delta F = 5.35\ln(C/C_0)$; $\Delta T = \lambda\Delta F$	CO₂ forcing and warming	C
$T(t) = T_{eq} + (T_0 - T_{eq})e^{-t/\tau}$	Exponential approach (Newton cooling)	C
$t_{1/2} = \ln 2/k = 0.693\tau$	Half-life, e-folding	C

References and further reading

• Science Olympiad, soinc.org: the 2027 rules manual (Remote Sensing B, Observable Earth; Remote Sensing C, Climate Change).
• NASA, TROPOMI and Sentinel-5P overview and ECOSTRESS: the instruments new to the 2027 lists.
• USGS, Earth Resources Observation and Science Center: the NLCD and SRTM elevation products.
• NASA Earthdata, earthdata.nasa.gov: satellite missions, instrument descriptions and data products, including MODIS, CERES, GRACE and OCO-2.
• NASA Earth Observatory, earthobservatory.nasa.gov: accessible explanations of Earth's energy budget, albedo, ENSO and the carbon cycle.
• USGS, Landsat Missions: the Landsat sensors (TM, ETM+ and OLI) named in the Division B scope.
• Copernicus, copernicus.eu: the Sentinel missions, including Sentinel-1 radar and Sentinel-2 multispectral imaging.
• NOAA NESDIS, nesdis.noaa.gov: geostationary and polar-orbiting weather satellites, including GOES and Suomi NPP.
• Kopp, G. and Lean, J. L. (2011), "A new, lower value of total solar irradiance," Geophysical Research Letters 38, L01706: doi:10.1029/2010GL045777. Source for the solar irradiance of about 1,361 W/m².
• Myhre, G., Highwood, E. J., Shine, K. P. and Stordal, F. (1998), "New estimates of radiative forcing due to well mixed greenhouse gases," Geophysical Research Letters 25, 2715: doi:10.1029/98GL01908. Source for the CO₂ forcing approximation.
• Wikipedia, Remote sensing: a general overview of platforms, sensors and applications. Use it to orient yourself, then check the primary sources above.
• SciOly Wiki, Remote Sensing: the community event page. Read it for context; this guide does not reproduce its content.
• USGS, Landsat satellite missions and band designations: launch dates, orbits and band wavelengths for Landsat 1 to 9.
• ESA, SentiWiki: Sentinel-1, Sentinel-2 and Sentinel-5P mission and instrument specifications.
• NASA, MODIS, ASTER, CERES, CALIPSO, GPM, PACE, GRACE-FO and OCO-2 mission pages.
• NOAA, GOES-R series: ABI bands, scan modes and satellite positions.
• MRLC consortium, National Land Cover Database: NLCD classes, products and the land cover change index.
• IPCC, Sixth Assessment Report, Working Group I: radiative forcing, feedbacks, climate sensitivity and greenhouse-gas lifetimes and GWPs.
• WMO, Greenhouse Gas Bulletin: current greenhouse-gas concentrations.
• NOAA Global Monitoring Laboratory, Trends in atmospheric CO₂: the Mauna Loa (Keeling) record.
• NOAA Climate Prediction Center, ENSO resources: the Oceanic Niño Index and ENSO impacts.
• Lillesand, T., Kiefer, R. and Chipman, J., Remote Sensing and Image Interpretation (Wiley), and Jensen, J. R., Remote Sensing of the Environment (Pearson): standard textbooks for sensors, image interpretation and classification.