Theoretical split-window algorithms for determining the actual surface temperature

Summary

Two split-window strategies for obtaining the actual surface temperature (AST) from an infra-red sensor system are investigated by means of simulations of radiometer signals for a wide range of different geophysical situations. The differences between strategies are closely related to the strong or weak presence of atmospheric effects in the split-window coefficient. It has been demonstrated that the strong presence makes it necessary to know the total atmospheric water vapour as well as the effective emissivities in the two thermal channels for applying the strong split-window method, whereas for the weak presence one only needs to know the emissivities. Simplified algorithms have been obtained from these methods when some of the input data are unknown. Likewise, included for all these algorithms is a rigorous evaluation of their accuracies that takes into account the uncertainties in the emissivity and the noise-equivalent temperature. In this way several algorithms are provided for determining the AST, leaving to the choice of the reader the algorithm that best responds to the required accuracy and the available input data.