REMOTE SENSING/Passive Sensors 437 alteration) related to mineralisation, have various absorption features in the spectral range 2.0–2.4 mm These two SWIR spectral ranges, corresponding to Landsat TM band and 7, are preferred by geologists SWIR sensor systems are technically more difficult and complicated because the SWIR detectors have to operate at low temperatures, which therefore require a cooling system (a liquid nitrogen coolant or a cryocooler) to maintain the detectors at about 80 K With six broad reflective spectral bands, Landsat TM provided the best spectral resolution among broadband sensor systems for many years The six broad reflective spectral bands are very effective for discrimination of various ground objects but they are not adequate to achieve specific identification in particular rock types and major mineral assemblies relating to mineral deposits This requires a sensor system with much higher spectral resolution at a few nanometre bandwidth to resolve their subtle spectral signatures This demand has led to the development of the hyperspectral system ASTER (a push-broom scanner for VNIR and SWIR bands), on board the Terra-1 satellite, is a representative of a transitional sensor system between broadband multispectral and hyper-spectral narrowband sensing It is an integrated system of three scanners: a VNIR push-broom scanner with three broad spectral bands; a SWIR push-broom scanner with six narrow spectral bands; and a TIR across-track mechanical scanner with five thermal bands (Table 1) The system combines good spatial resolution in the VNIR bands and high spectral resolution in SWIR and thermal bands and was specifically designed for geological applications The three 15 m resolution VNIR bands are adequate for distinguishing broad categories of land cover such as vegetation, water, red soils, urban areas, superficial deposits, and rock outcrops, while the six narrow SWIR bands of 30 m resolution provide potential for mapping major mineral assemblies of rock types and alterations Another unique advantage of ASTER is that it has along track stereo capability The VNIR scanner has a backward viewing telescope to take NIR images beside its nadir telescope for the three VNIR bands Thus nadir and backward viewing NIR images are taken simultaneously, forming along track stereo image pairs The along track stereo image pairs enable generation of DEM (Digital Elevation Model) data Thermal Infrared (TIR) Sensors Various minerals, rock types, and other ground objects have different thermal properties, such as thermal inertia, thermal emission, and thermal absorption For instance, quartz has distinctive high emissivity around mm and strong absorption in 10–11 mm; dolomite can be distinguished from limestone by its significant lower emissivity in 8–11 mm There are several airborne thermal IR sensor systems for example, the Thermal Infrared Multispectral Scanner (TIMS) with bands in the 8.2–12.2 mm spectral region developed in 1982 Of the satellite systems, the Landsat Thematic Mapper system has a broad thermal band, TM band 6, at a wavelength range of 10.4– 12.5 mm So far the only space-borne multispectral thermal system is ASTER on board of Terra-1 satellite, which has thermal bands as shown in Table In general, a broadband TIR sensor operating in the 8–14 mm spectral range images the radiation temperature of the land surface while the narrower band of the multispectral thermal imagery data detect the thermal spectral signatures of materials on the land surface It is important to note that daytime thermal images are fundamentally different from the night time thermal images Daytime thermal images are dominated by topographic features, governed by the geometry between slopes and solar radiation described previously, while night time thermal images are solely determined by emission from the Earth surface They therefore detect the thermal properties of ground materials more efficiently For both systems, TM and ASTER, the spatial resolutions of their thermal bands are significantly lower than their reflective multispectral bands, as shown in Table One reason is that the interaction between thermal energy (or heat) and the atmosphere is more complicated than the case of VNIR and SWIR energy Heat can be transmitted in air, not only by radiation, but also by air circulation Both solar radiation to the Earth in TIR spectral range and direct thermal emission from the Earth are very weak, compared with the energy intensity of the Earth reflected solar radiation in the VNIR and SWIR spectral ranges Most thermal sensors are of cross-track mechanical scanner type, as shown in Figure The major difference of a thermal scanner from a reflective multispectral scanner is that it needs a cooler system to maintain the TIR detector at very low temperature for maximum sensitivity For instance, the thermal sensor of Landsat Thematic Mapper is surrounded by liquid nitrogen at 77 K stored in an insulated vessel In the ASTER system, a cryocooler is used to maintain the detectors for TIR bands at 80 K A blackbody plate is used as an on-board calibration reference that is viewed before and after each scan cycle, providing estimation of instrument drift This is essential to maintain the accuracy and consistency of a TIR instrument The temperature sensitivity of a modern TIR sensor system can be as high as 0.1 K