DERIVATION OF TEMPERATURE AND EMITTANCE FROM AIRBORNE MULTISPECTRAL THERMAL INFRARED SCANNER DATA

A new iterative algorithm is proposed to calculate surface temperature and emittance from data of an airborne scanner with reflective and thermal multispectral channels. For water areas the surface temperature can be determined from the thermal channels alone with an accuracy of 0.5-0.9-degrees-C, assuming a well-calibrated sensor and known atmospheric parameters. For non-water areas, information from the reflective bands of the airborne sensor is used for each image pixel to perform a pre-classification of the thermal band data by assigning a representative emittance value to one of the thermal bands. For a number of earth surface types this emittance value can be estimated with an accuracy of about +/-0.01, leading to a surface temperature accuracy of 0.5-1.5-degrees-C.