Materials on the Earth's surface differ in the way they reflect light of different wavelengths. These spectral differences are routinely exploited to provide maps of land cover, using satellite-borne instruments that collect light in a modest number of fairly broad spectral bands. The broad-band radiances collected by many satellites are generally adequate for rough discrimination into categories of general surface cover, but are rather limited in the amount of more quantiative information that can be inferred from the spectral content of their data.
Imaging spectrometry is a way to obtain more accurate information on the physical properties of the Earth's surface from remote platforms. The technique was originally developed to obtain chemical information of inaccessible planetary surfaces in the solar system but is now becoming of increasing importance in many application areas in Earth observation. Instead of data being acquired in a small number of fairly broad (~100nm) bands, the aim is to obtain data from a much larger number of spectral channels, with widths of perhaps 10nm, and which overlap to form a contiguous spectrum. Beyond the improved classification accuraces that might be expected from the greater number of channels, the acquisition of contiguous spectra opens the door to the use of spectroscopic techniques not possible with more familiar sources of satellite data.
During the last few years a new generation of airborne (such as AVIRIS - Airborne Visible and Infrared Imaging Spectrometer) optical remote sensing systems has evolved under the name of "imaging spectrometers". These new sensors can provide image data with detailed spectral resolution up to as many as 200 spectral bands and are powerful tools for Earth observing remote sensing. One of the ultimate goals of imaging spectrometry is to produce laboratory-like reflectance spectra for each pixel in an image.
Page last changed 9 March 2002