eduKate Learning Manual: One NOAA-21 CrIS Infrared Photon | How Earth’s Thermal Glow Becomes an Atmospheric Temperature and Moisture Profile

eduKate Learning Manual · Science World | Continuation Route · Infrared Spectroscopy × Weather × Remote Sensing

Subtitle: Earth glows in infrared light all the time. CrIS measures that glow in thousands of spectral channels, but the vertical atmosphere shown in a weather product is not photographed layer by layer. It is reconstructed from radiance, absorption physics and a retrieval model.

Wait, What?

A satellite can estimate the temperature of air at different heights without carrying a thermometer through those layers.

NOAA-21’s Cross-track Infrared Sounder, CrIS, looks downward at the thermal infrared radiation leaving Earth and its atmosphere. Different gases absorb and emit at different wavelengths. Some wavelengths can escape from near the surface; others carry stronger contributions from higher layers. The resulting spectrum is therefore a coded vertical story.

The difficult part is decoding it without pretending the spectrum is a direct photograph of altitude.

Worth My While

This route teaches the logic behind atmospheric sounding and many other inverse problems. The instrument measures spectral radiance. Physics describes how temperature, water vapour, carbon dioxide and other gases affect that radiance. A retrieval combines the spectrum with prior information and often complementary microwave observations to estimate the atmosphere’s vertical state.

The map or profile is therefore scientifically grounded, but it is one step removed from direct observation. Knowing that distinction makes weather data more—not less—meaningful.

Big Question

How can one thermal-infrared photon emitted by Earth’s surface or atmosphere contribute to a NOAA-21 CrIS interferogram and spectrum, then support a vertical temperature or moisture profile without treating one spectral channel as a direct thermometer at one exact altitude?

Quick Answer

CrIS is a Fourier-transform infrared spectrometer. Incoming thermal infrared radiation is split and recombined so that different optical path differences create an interferogram. Mathematical Fourier transformation converts that interferogram into a high-resolution spectrum.

Atmospheric molecules imprint wavelength-dependent absorption and emission features on the spectrum. Carbon dioxide bands carry strong temperature information because CO₂ is relatively well mixed. Water-vapour bands carry humidity information. Retrieval systems calculate which atmospheric temperature and moisture profiles would produce spectra consistent with the observed radiance, while accounting for clouds, surface properties, instrument response and prior constraints.

NOAA’s current operational JPSS constellation uses NOAA-21 as the primary spacecraft for CrIS sounding products, with NOAA-20 supporting the constellation. The older Suomi NPP CrIS operational stream ended in February 2026, which is why this route names NOAA-21 explicitly.

What You Will Learn

Why Earth emits thermal infrared radiation day and night.
What a Fourier-transform spectrometer measures.
How an interferogram becomes a spectrum.
Why different wavelengths contain information about different atmospheric layers.
Why temperature and humidity profiles are retrievals rather than direct thermometer readings.
How clouds and surface emissivity create ambiguity.
Why microwave and infrared sounders complement each other.

Part 1 — Primary Foundation: Warm Things Glow Even When We Cannot See the Glow

Everything above absolute zero emits electromagnetic radiation. A human body emits mostly infrared. Earth’s land, ocean, clouds and atmosphere also emit infrared radiation because they are warm.

The amount and wavelength distribution depend on temperature and emissivity. On its way upward, that radiation also interacts with gases. By the time a photon reaches a satellite, its journey may contain clues about both where it came from and what it passed through.

Part 2 — Secondary Mechanism: Molecules Leave Spectral Fingerprints

Molecules absorb and emit infrared radiation at frequencies related to their rotational and vibrational energy states. Water vapour has many strong bands. Carbon dioxide has important infrared bands too.

At a wavelength where the atmosphere is relatively transparent, the satellite may receive substantial radiation from the surface or lower atmosphere. At a wavelength absorbed strongly by a well-mixed gas, radiation from lower levels is more likely to be absorbed and replaced by emission from higher, colder layers. Nearby wavelengths can therefore have different vertical sensitivities.

This is the foundation of sounding: not one wavelength equals one altitude, but a family of wavelengths has different weighting functions across altitude.

Part 3 — JC Depth: CrIS Measures an Interferogram Before It Produces a Spectrum

CrIS is not a simple prism camera. In a Fourier-transform spectrometer, incoming infrared radiation is divided into optical paths and recombined. Changing path difference makes the waves interfere constructively and destructively. The detector records this combined signal as an interferogram.

A Fourier transform converts the interferogram from optical-path-difference space into spectral radiance versus wavenumber. CrIS provides thousands of full-spectral-resolution channels. That dense spectrum contains far more information than a broad-band infrared temperature image.

Follow One NOAA-21 CrIS Infrared Photon

Thermal motion in Earth’s surface or atmosphere contributes infrared radiation.
Our photon travels upward and may encounter absorbing or emitting gases.
Clouds may block, absorb or emit strongly along the path.
The photon enters the CrIS optical system aboard NOAA-21.
Its electromagnetic field contributes to interference as optical paths are combined.
The detector does not label that one photon by altitude; it records the aggregate interferometric signal.
Calibration converts detector response into physically meaningful radiance.
A Fourier transform converts the interferogram into a spectrum.
Radiative-transfer models compare the measured spectrum with candidate atmospheric states.
Retrieval algorithms estimate vertical temperature, moisture and selected trace-gas information with uncertainty.
Those profiles enter weather analysis, forecasting and climate-quality data systems alongside other observations.

How Do We Know?

NOAA STAR describes CrIS as a Fourier-transform spectrometer measuring high-spectral-resolution upwelling infrared radiance. NOAA’s JPSS products combine CrIS infrared information with ATMS microwave observations to obtain atmospheric temperature, moisture and pressure profiles.

NASA and NOAA retrieval systems compare satellite-derived profiles with radiosondes, aircraft observations, numerical weather analyses and other satellite instruments. The agreement is not perfect—and that is useful. Differences reveal cloud effects, calibration problems, representativeness differences and conditions in which the retrieval has weaker information.

Observation vs Inference

Statement	Status
CrIS detected a calibrated spectral radiance in a particular channel.	Observation after instrument calibration.
The spectrum contains absorption/emission structure consistent with CO₂ and water vapour.	Physical interpretation grounded in spectroscopy.
Air at a range of pressure levels has a retrieved temperature and humidity profile.	Inverse-model result.
One channel measured the exact temperature at one exact height.	Incorrect simplification.
A later weather forecast will have a particular outcome.	Model prediction using many observations and dynamics, not a CrIS observation itself.

Misconceptions and Repairs

Misconception: CrIS is an infrared camera that directly photographs atmospheric layers. Repair: it measures spectral radiance and retrieves vertical profiles through radiative-transfer physics.
Misconception: one wavelength comes from one altitude. Repair: each channel usually has a broad vertical weighting function.
Misconception: brightness temperature is always the physical temperature of air. Repair: brightness temperature is a radiance-equivalent quantity and depends on where radiation originates and what absorbs it.
Misconception: clouds simply add noise. Repair: clouds are physical emitters and absorbers that can remove information from lower layers.
Misconception: infrared sounding and microwave sounding are duplicates. Repair: they have different spectral sensitivities and different responses to clouds, so their information is complementary.

Worked Reasoning: A Cold Infrared Channel Does Not Automatically Mean Cold Surface Air

Suppose one infrared channel has a low brightness temperature. It is tempting to say, “The ground must be cold.”

But if that wavelength lies in a strong CO₂ absorption band, much of the detected radiation may come from higher, colder atmospheric layers. A high cloud can also dominate the radiance. Surface emissivity can matter in atmospheric windows.

The correct diagnosis asks which part of the atmosphere the channel is sensitive to and what alternative emitters or absorbers could explain the measurement. Only then should it contribute to a surface or atmospheric temperature inference.

Checkpoint

What quantity does CrIS directly measure after calibration?
Why can nearby infrared wavelengths sense different parts of the atmosphere?
What is an interferogram?
Why is a retrieved profile an inverse problem?
Why do microwave observations improve atmospheric sounding in cloudy conditions?

Answer Key

Spectral infrared radiance.
Molecular absorption varies strongly with wavelength, changing the vertical region that contributes most to outgoing radiance.
The detected interference signal as a function of optical path difference before Fourier transformation.
Many possible atmospheric states can produce similar radiances, so physics and prior constraints are needed to infer the most consistent state.
Microwaves penetrate many non-precipitating clouds better than infrared and provide complementary temperature/moisture information.

Can You Explain WHY?

Why is carbon dioxide useful for temperature sounding even though CrIS is not primarily measuring CO₂ concentration in every channel?
Why can a cloud top hide information about the air and surface below it?
Why does high spectral resolution improve vertical information?
Why must retrieved profiles be validated against independent instruments?

Singapore and the World

Singapore’s atmosphere is warm, humid and frequently cloudy. That makes the CrIS–ATMS partnership especially instructive: infrared sounders provide rich spectral information, while microwave measurements can retain useful sensitivity through many cloudy scenes. Operational weather prediction combines many receivers because no single instrument sees the atmosphere perfectly under every condition.

For a Singapore learner, the deeper lesson is that modern weather forecasting begins with measurement diversity: radiosondes, aircraft, radar, satellites, surface stations, ocean observations and numerical models each constrain different parts of the same moving atmosphere.

Deep Science Window — Retrieval Means Solving the Radiative-Transfer Equation Backwards

The forward problem asks: if we know the atmospheric temperature, humidity, gases, clouds and surface emissivity, what spectrum should CrIS observe?

The inverse problem asks: given the observed spectrum, what atmospheric state most plausibly produced it? That reversal is difficult because the atmosphere has many variables and the measurement contains finite information. Retrieval algorithms therefore use weighting functions, error covariances, prior estimates and physical constraints.

A good retrieval does not hide uncertainty. It reports what the spectrum can actually constrain.

Counterexamples and Model Limits

Thick cloud can remove infrared information about layers below the cloud top.
Surface emissivity varies with land type and wavelength.
Temperature inversions can challenge simple assumptions about vertical structure.
Water vapour is highly variable and can create nonlinear spectral effects.
A retrieval can be pulled toward its prior where observations contain little information.
Instrument noise and calibration uncertainty propagate into the profile.
Satellite footprints and radiosonde paths sample different volumes, so apparent disagreement can include representativeness error.

Evidence Boundaries

This page owns the route from one thermal-infrared contribution through CrIS spectroscopy to a bounded atmospheric-profile inference. Molecular spectroscopy belongs to Physics and Chemistry. Numerical weather prediction, data assimilation, cloud retrieval, trace-gas retrieval and forecast operations remain specialist owners.

A CrIS profile is scientific evidence, not a stand-alone severe-weather warning or aviation decision.

KNOW → CONNECT → EXPLAIN → APPLY → CHECK

KNOW: CrIS measures spectral radiance, not direct vertical temperature.
CONNECT: thermal emission → molecular absorption/emission → interferogram → spectrum → retrieval.
EXPLAIN: why weighting functions are broad rather than one wavelength per altitude.
APPLY: compare an atmospheric-window channel with a strong CO₂-band channel.
CHECK: test clouds, emissivity, calibration and prior sensitivity before over-interpreting a profile.

eduKateAI Direction Graph

Thermal emission (radiation-physics owner) → molecular absorption and emission (spectroscopy owner) → CrIS interferogram and radiance spectrum (instrument owner) → atmospheric profile retrieval (remote-sensing owner) → data assimilation and forecast (meteorology owner). Science Route owns only the traversal.

Where to Go Next

Compare this passive thermal-infrared route with a radiosonde, which physically travels through the atmosphere, and with GNSS radio occultation, which infers atmospheric structure from signal bending. Different instruments reach similar state variables through completely different observables.

Authoritative Sources

Teaching Guide for Parents, Tutors and Teachers

Use three transparent sheets labelled surface, lower atmosphere and upper atmosphere. Give each sheet a different pattern of coloured “absorption lines”. Shine an imaginary broadband spectrum upward and ask which wavelengths survive from which layer.

The diagnostic question is: “What did CrIS observe before any profile existed?” The learner should answer “spectral radiance” or “an infrared spectrum”. If they answer “temperature at 5 kilometres”, return to the observation-versus-inference table.

For advanced transfer, ask why two different atmospheric profiles might produce similar spectra and what extra information could discriminate them. That question opens the door to inverse problems, uncertainty and data assimilation without losing the reader in jargon.