PREFIRE: Separating the Instrument's Heat from Earth's Signal

When Earth sends energy into space, the wavelengths carrying that energy matter alongside the total amount. Surface emission, water vapor, and clouds can leave different signatures in the same spectrum. Engineers trying to read those signatures from a small satellite face an additional problem: the instrument measuring infrared radiation is itself a warm object. PREFIRE's story concerns not only collecting a distant polar scene's signal, but separating the optical and electronic system's own contribution from the measurement. [1]

NASA's PREFIRE mission approaches this problem with thermal infrared spectrometers aboard two small CubeSats. The first satellite launched on 25 May 2024; the second launched on 5 June 2024 in New Zealand local time. The second event occurring on 4 June in US Eastern Daylight Time reflects a time-zone difference, not another launch. Observing regions at different times gives the pair additional opportunities to investigate short-lived changes. It does not produce a continuous movie of the poles. [2]

A measurement choice made by miniaturization

Instrument research published by Brian J. Drouin and colleagues in 2025 details the PREFIRE spectrometers' design and testing. Each completed TIRS instrument has a mass of 2.7 kilograms and consumes 4.3 watts. Those figures are not the mass and total power demand of the entire spacecraft. They describe the measuring instrument fitted within a limited spacecraft accommodation budget. [1]

A central choice was an ambient-temperature thermopile array rather than a detector requiring deep cooling. Making the system smaller and less power-hungry did not mean maximizing every form of resolution. The team considered which spectral distinctions were sufficient for the scientific questions and which spatial scales could sample cloudy and clear scenery appropriately. Instrument design therefore involved deciding which information to preserve within its constraints.

A thermopile sensor absorbs incoming radiation and produces a readable electrical response. The 64-by-eight-pixel array carries spectral and spatial information together with the optical system. Yet a pixel's signal is not simply an untouched copy of light arriving from Earth. Thermal contributions from optics, filters, and the detector system also enter the measurement. Building a compact instrument means managing and characterizing that background, not making it disappear.

Sampling intervals and resolution

A diffraction grating directs different wavelengths to different locations on the detector plane. The design provides dispersion across approximately 5–54 micrometers. This does not establish that every part of that range supplies equally usable scientific measurements. Optical design coverage, tested response, and calibration limitations are distinct properties. [1]

The research reports spectral resolution of 1.71 micrometers below 35 micrometers, with sampling at 0.86-micrometer increments. These numbers are not interchangeable. Taking closely spaced samples does not mean resolving neighboring spectral features equally sharply. Slit width, detector size, and diffraction jointly determine how each channel responds to nearby wavelengths.

The grating's different diffraction orders introduce another separation problem. Unwanted contributions at other wavelengths can reach the same detector as the desired spectral signal. Order-sorting filters reduce this mixing. They are not accessories adjusting the appearance of a colored photograph; they help define which radiation a channel measures. Their transmission and internal reflections must also be characterized and incorporated into the instrument model.

Earth, an internal target, and cold space

A pointing mirror moves the instrument between three viewing positions: the Earth scene, an internal calibration target, and cold space. Sensors monitor the internal target's temperature. Views of that target and space help establish the gain relating digitized detector values to physical spectral radiance. Repeated calibration during an orbit lets the Earth measurement be interpreted alongside the instrument's changing state. [1]

Radiance here describes radiation within a specified direction and wavelength range; it is not identical to the total energy emitted in every direction. Further inferences about surface and atmospheric properties require the spectral response and physical models. Calling a channel value a direct surface temperature under every condition would skip those inference steps. Clouds and gases can change the origin of the signal being observed.

An uncooled detector is not an instrument without thermal management. Thermal connections carry heat from electronics, while isolators and temperature measurements help track the background. If thermal changes between calibration views are not modeled sufficiently well, they can enter the imagery as noise or bias. Measuring from space therefore requires knowing the instrument's state as well as pointing it toward the target.

Similar instruments with separate responses

Ground tests identified spectral-position shifts in opposite directions for the two TIRS instruments. The researchers connected the measured responses to a geometric instrument model and determined the channel responses of each unit separately. Two instruments built to the same design were not assumed to share an identical calibration automatically. Small differences in optical placement can alter how much radiation from a particular wavelength enters a channel. [1]

An additional limitation appeared at longer wavelengths. Response lag in channels associated with a filter increased with wavelength. The paper reports that short calibration exposures produced reproducible gains without further correction up to 36 micrometers, while time-dependent calibration would need to be developed beyond that range. The performance table's 6–36-micrometer bandwidth and the broader optical design coverage should therefore not be combined into a claim of uniformly validated performance.

Laboratory quantities such as noise-equivalent difference temperature are likewise not identical to absolute accuracy in measuring every terrestrial surface temperature. Instrument response, calibration, pointing, and spacecraft thermal stability jointly determine data quality. The paper's ground characterization does not establish that every orbital operating condition has been independently validated. [1]

Observations with times and gaps

The instrument collects a measurement frame approximately every 0.707 seconds. As the satellite moves, however, the next frame is not a revisit to the same ground location. Eight separated scenes are sampled across track rather than a continuous strip. Locating their footprints combines the instrument's angular model with spacecraft attitude information and elevation correction. Rapid frame acquisition does not mean that the entire planet is observed in detail at every instant. [1]

NASA announced on 15 August 2025 that the mission was extended through September 2026 and that its analysis focus would broaden beyond the poles. That dated announcement does not by itself establish the satellites' operating status on 13 October 2026. [3]

PREFIRE's technological value does not lie in a small spectrometer resolving every climate question. It lies in sampling a difficult spectral region within specified constraints. Better models and forecasts are intended uses of the information, not universal outcomes demonstrated directly by the instrument tests. Reading design coverage, calibrated response, and sampling geometry together makes the meaning of these new observations clearer. The instrument's own heat is part of that account, because the route from a detector voltage to Earth's emitted radiation passes through the state of the measuring system itself.

Sources

[1] Brian J. Drouin and colleagues (2025), Thermal Infrared Spectrometers for the Polar Radiant Energy in the Far-Infrared Experiment (PREFIRE), Earth and Space Science 12, e2024EA003711. Primary research, DOI: 10.1029/2024EA003711, first published 26 September 2025; CC BY 4.0.

[2] NASA, NASA Launches Second Small Climate Satellite to Study Earth's Poles, 5 June 2024, correction 6 June 2024.

[3] NASA/JPL, NASA's PREFIRE CubeSat Mission Extended, 15 August 2025.

Information cutoff: 13 October 2026. Source access and review: 14 October 2026. The primary paper's design, testing, calibration, and conclusion text was read; raw data, supplements, and orbital performance were not independently reanalyzed. The article-specific CC BY 4.0 license was verified through publisher-deposited Crossref metadata. Original explanation and adaptation attributed to Drouin and colleagues.