Improved land surface temperature retrieval method for the small satellite flying laptop

The enhanced algorithms developed in this study for land surface temperature retrieval combine the split-window method with the multi-angle method, employing a large number of view angles between nadir and 60 degrees zenith angle. The new algorithms are intended for analysing data from the Thermal Infrared Camera System of the small satellite Flying Laptop, a technology demonstration and earth observation satellite currently being developed at the Institute of Space Systems, Universitaet Stuttgart. The Thermal Infrared Camera System will observe at 8.0 to 9.2 μm and at 10.4 to 12.6 μm. The attitude control system of the satellite permits arbitrary pointing angles. This combination of bi-spectral camera system and multi-angle observation capability offers means of improving land surface temperature retrieval. In order to investigate the utility of the prospective observations, data has been modelled numerically for a wide range of atmospheric properties, using the radiation transport model MODTRAN. The resulting data was analysed statistically and a number of algorithms for temperature retrieval have been derived, depending on available angles. The multi-angle method improves accuracy of the retrieved temperature compared to the employment of the split-window method only, especially for very humid atmospheres. A standard deviation below one Kelvin can be achieved for land surface temperature.