DEVELOPMENT OF PRACTICAL MULTIBAND ALGORITHMS FOR ESTIMATING LAND-SURFACE TEMPERATURE FROM EOS/MODIS DATA

A practical multiband, hierarchical algorithm for estimating land-surface temperature from NASA's future EOS instruments MODIS and ASTER is developed through comprehensive, accurate, radiative transfer simulations at moderate spectral steps of 1-5 cm(-1) for wide ranges of atmospheric and surface conditions. The algorithm will accept empirical or estimated information about the surface emissivity and reflectivity and the atmospheric temperature and water-vapor profiles. Ground-based and aircraft measurements are necessary to validate and improve the algorithm and to establish its quality. Its accuracy depends on the calibration accuracy of thermal infrared data, uncertainties in surface heterogeneity, and temperature-dependent atmospheric absorption coefficients. Better knowledge of land-surface spectral emissivities and more accurate coefficients for atmospheric molecular band absorption and water vapor continuum absorption are needed to develop global land-surface temperature algorithms accurate to 1-2 degrees K.