Optical spectral sensing

Optical spectral sensing is the passive measurement of solar radiation reflected or thermally emitted by Earth's surface, resolved by wavelength across the optical-to-thermal spectrum (approximately 0.4-14 micrometers). It is the most widely deployed EO sensing principle, encompassing panchromatic, multispectral, superspectral, and hyperspectral instrument classes.

Sensors measure spectral radiance at the aperture. Reflected sunlight drives VNIR and SWIR channels (0.45-2.5 micrometers); thermally emitted radiation from the surface drives TIR channels (8-14 micrometers). After atmospheric correction, measured radiance is converted to top-of-atmosphere or surface reflectance and emittance. Spectral signatures are matched to known material properties via band ratios, classification algorithms, or radiative-transfer inversion; TIR emission inversion yields land surface temperature.

Spatial resolution ranges from sub-metre commercial instruments to kilometre-scale global monitoring sensors. Hyperspectral instruments (100 or more contiguous bands at approximately 10 nm FWHM) enable spectroscopic identification of minerals, vegetation biochemistry, and atmospheric trace gases. Representative missions include Landsat 8/9 (NASA/USGS, OLI+TIRS, 30 m/100 m, 11 bands), Sentinel-2 A/B (ESA, MSI, 13 bands, 10-60 m), MODIS (Terra/Aqua, 36 bands, 250 m-1 km), PRISMA (ASI, 239 hyperspectral bands), and EnMAP (DLR, 230 bands, 420-2450 nm).

Principal applications include vegetation health and phenology mapping, land cover and land use classification, ocean colour and water quality, land surface temperature mapping, atmospheric trace gas column mapping (NO2, SO2, O3, CH4), sun-induced chlorophyll fluorescence for carbon cycle monitoring, wildfire thermal anomaly detection, and mineral mapping for geological surveys.

The technology's principal strength is its long mission heritage and the largest installed sensor ecosystem in EO, with commercial instruments now offering sub-metre spatial resolution. Core limitations are blockage of reflected-light bands by cloud cover and aerosol load, solar-illumination dependence, and the requirement for atmospheric correction for surface-level retrievals.