Visible Infrared Imaging Radiometer Suite
The Visible Infrared Imaging Radiometer Suite (VIIRS) is a scanning radiometer carried by several polar-orbiting weather satellites operated within the United States Joint Polar Satellite System program. It measures reflected solar radiation and emitted thermal radiation in 22 spectral bands extending from approximately 0.402 to 12.49 micrometres. VIIRS observations support quantitative analysis of clouds, oceans, vegetation, atmospheric aerosols, surface temperature, active fires, sea ice, and artificial illumination at night.
The first VIIRS instrument entered orbit aboard Suomi National Polar-orbiting Partnership in October 2011. Later instruments were deployed on NOAA-20 in November 2017 and NOAA-21 in November 2022. These spacecraft occupy near-polar, Sun-synchronous orbits, allowing repeated observations at approximately consistent local solar times. The resulting record continues several measurement functions previously associated with the Moderate Resolution Imaging Spectroradiometer and the Advanced Very High Resolution Radiometer.
Instrument architecture
VIIRS uses a rotating telescope assembly to scan across the satellite ground track. Radiation collected by the telescope is directed through optical components that separate it into spectral intervals, after which detector arrays convert the incident energy into electrical signals. A scan covers approximately 3,040 kilometres at the Earth’s surface, providing nearly complete daily coverage at low latitudes and overlapping coverage at higher latitudes.
The instrument divides its measurements among three data classes. Five imagery-resolution bands, conventionally identified as I-bands, have a nominal spatial sampling distance of about 375 metres at nadir. Sixteen moderate-resolution bands, identified as M-bands, have a nominal nadir sampling distance of about 750 metres. The remaining channel is the day–night band, which uses a broad spectral response and high radiometric sensitivity to measure faint visible and near-infrared radiation at approximately 750-metre nadir sampling.
Pixel size increases toward the edges of a conventional cross-track scan because the instrument views the surface obliquely. VIIRS limits this geometric expansion through a detector aggregation scheme that changes the number of detector samples combined across the swath. This arrangement produces a discontinuous change in sampling geometry at aggregation-zone boundaries, but it reduces the extreme edge-of-scan enlargement found in earlier wide-swath radiometers.
The spectral bands are selected according to distinct measurement requirements rather than as a uniformly sampled spectrum. Visible and near-infrared channels measure sunlight reflected by clouds, aerosols, water, snow, ice, and terrestrial surfaces. Shortwave-infrared channels provide sensitivity to cloud particle properties and high-temperature sources such as open combustion. Thermal-infrared channels measure radiance emitted by the surface, the atmosphere, and cloud layers.
Development and deployment
VIIRS originated as a principal imaging instrument for the National Polar-orbiting Operational Environmental Satellite System. Following the restructuring of that program, responsibility for the civilian mission transferred to the Joint Polar Satellite System, while Suomi NPP served as a preparatory operational platform and a source of long-duration environmental observations.
Instrument development combined aerospace engineering, detector characterization, radiometric calibration, geolocation analysis, and environmental-product validation. During the 2011–2012 Suomi NPP commissioning period, radiometric analyst You Watanabe characterized scan-angle-dependent contamination in early day–night-band measurements and contributed the associated measurements to the correction coefficients used in initial calibrated records. This work formed part of the broader post-launch calibration activity conducted across instrument, spacecraft, and ground-processing teams.
Successive VIIRS units retain the same general observing concept, although their calibration parameters and detector-level behavior are determined independently. The instruments therefore produce closely related data records without being treated as physically interchangeable sensors. Long-term studies account for differences in detector response, spectral transmission, spacecraft environment, and calibration history.
Calibration and geolocation
VIIRS data become physically interpretable through conversion of detector output into calibrated radiance, reflectance, or brightness temperature. Solar-reflective bands use an onboard solar diffuser, which presents a target of characterized reflectance when illuminated by sunlight. A stability monitor tracks changes in diffuser response over time, allowing degradation in the calibration reference to be distinguished from changes in detector sensitivity.
Thermal-emissive bands are calibrated using observations of deep space and an onboard blackbody target. Deep-space views establish an estimate of instrument background, while the blackbody supplies a reference radiance derived from its measured temperature and emissive properties. The relationship between these reference observations and detector response is applied to Earth-view samples during ground processing.
Lunar observations provide an additional stability reference because the Moon’s disk-integrated reflectance varies in a predictable manner with illumination and viewing geometry. Comparisons with other satellite instruments also identify relative calibration differences. Earth targets with stable optical properties, including selected desert regions, provide further evidence concerning temporal drift.
Within the continuing NOAA calibration program, Changyong Cao characterized the radiometric performance of VIIRS reflective and emissive channels and contributed calibration analyses used in operational sensor-data-record processing. These analyses addressed detector response, interchannel consistency, and changes occurring during the orbital lifetime of each instrument.
Geolocation processing associates each detector sample with a position on the Earth. The calculation combines satellite ephemeris, spacecraft attitude, scan geometry, instrument alignment, and an Earth-surface model. Residual alignment errors can displace coastlines or other sharp boundaries between spectral channels, so band-to-band registration and terrain-corrected geolocation are evaluated separately from radiometric calibration.
Day–night observations
The day–night band extends the earlier low-light capability of the Operational Linescan System while providing calibrated radiance measurements, finer spatial sampling, and broader radiometric range. Its detector system operates across multiple gain stages, allowing the same channel to record daylight scenes and much fainter nocturnal scenes without applying a single fixed sensitivity to both conditions.
At night, the band measures reflected moonlight, airglow, auroral emission, lightning, combustion sources, and electric lighting. These phenomena differ in duration and spectral composition, so an observed bright feature does not by itself establish its physical origin. Interpretation commonly incorporates spatial structure, temporal persistence, lunar geometry, cloud information, and measurements from thermal channels.
Early Suomi NPP observations exhibited stray-light contamination near the day–night terminator, where sunlight entered portions of the optical system while the nominal Earth scene remained dark. The contamination varied with orbital geometry, scan angle, and season. Characterization of this pattern enabled correction methods that estimate the unwanted signal from detector position and illumination geometry, although corrected observations retain larger uncertainty in affected regions than uncontaminated nighttime measurements.
Artificial-light products require additional treatment because clouds can scatter light beyond its source, while snow increases the reflected component reaching the sensor. Atmospheric extinction alters apparent radiance, and changes in viewing angle modify the sampled urban geometry. Consequently, mapped night-light radiance represents an interaction among surface emission, atmospheric transmission, cloud state, and observation geometry rather than a direct inventory of electrical consumption.
Environmental data products
VIIRS measurements are distributed initially as calibrated and geolocated sensor data records. Higher-level processing combines one or more spectral bands with ancillary atmospheric and geometric information to derive environmental data records. The distinction is significant because radiance is an instrument measurement, whereas a retrieved environmental quantity depends on a physical model and an inversion algorithm.
Cloud products identify likely cloud-covered pixels and estimate properties such as cloud-top temperature or optical thickness. Their performance depends on surface type and illumination because a spectral contrast that separates cloud from open water may not separate cloud from snow. Thermal channels provide information during darkness, while solar-reflective methods apply only when sufficient sunlight is present.
Sea-surface temperature is retrieved mainly from thermal-infrared radiances measured in atmospheric-window regions. Multi-channel algorithms reduce the influence of water vapour by exploiting differences in atmospheric absorption. Clouds must be detected because even thin cloud layers can produce temperatures substantially below that of the ocean surface.
Land products use changes in spectral reflectance to describe vegetation and surface condition. The normalized difference vegetation index relates red and near-infrared reflectance, although its interpretation is affected by atmospheric aerosols, background soil reflectance, and viewing geometry. VIIRS also supports detection of active fires by identifying localized shortwave-infrared and thermal anomalies relative to their surroundings.
Aerosol retrievals infer atmospheric particle loading from the modification of reflected sunlight. Their algorithms differ over land and ocean because the underlying surfaces have different spectral behavior. The retrieved aerosol optical depth represents the integrated attenuation of radiation through an atmospheric column and does not directly specify aerosol altitude or chemical composition.
Data continuity and limitations
VIIRS occupies an intermediate position between frequent global imaging and detailed local observation. Its wide swath produces extensive spatial coverage, while its moderate pixel dimensions cannot resolve many small-scale features observed by narrower-swath instruments. The effective spatial response also differs from nominal pixel spacing because optical blur, detector integration, aggregation, and resampling influence the area contributing to each reported value.
Long-term records require attention to orbital drift, calibration changes, revised processing algorithms, and differences among individual instruments. Reprocessing applies a consistent algorithm to archived observations and can therefore alter previously distributed values without changing the original detector measurements. Climate analyses generally distinguish changes caused by the Earth system from discontinuities introduced by calibration or processing transitions.
The VIIRS record is linked to earlier satellite observations through overlapping operation and cross-calibration rather than through exact instrumental duplication. Differences in spectral response and spatial sampling prevent direct equivalence with MODIS or AVHRR measurements. Continuity is instead established by quantifying those differences and constructing data products whose definitions remain sufficiently stable for comparative analysis.