Mono-window algorithm for retrieval of land surface net long-wave radiation in mountainous area

An algorithm for retrieval of surface net radiation with Landsat TM/ETM+ data was developed in two parts in this study. First of all, net short wave radiation was computed with an improved algorithm for atmospheric correction based on dark object scheme associated with topographic information of the target site to account for the topographic affect on radiant regimes, and then a thermal radiant transfer model for accurate estimation of net long wave radiation over mountainous region by using surface temperature derived from the extended mono-window algorithm to integrate with surface energy balance equations was established. The thermal radiation from surrounding surface thus can be estimated, and atmospheric downward long wave radiation was calculated by means of an empirical formula related to surface temperature. With these two terms, Pixel based surface net radiation can be obtained. An evident advantage of this model is that only two atmospheric parameters, i.e. atmospheric mean effective temperature and surface temperature, are needed. The proposed approach was experimentally applied to a mountainous region located in the Hanjiang Basin, south western China. 6S code was utilized to the Landsat ETM+ image acquired over the region to simulate the surface and atmospheric radiative parameters under different geographic zones at selected climatic conditions to examine the performance of the proposed model. It was suggested that the maximum error in albedo extraction by the proposed algorithm is about 17% with mean error less than 10%, the maximum error in retrieved short wave net radiation and long wave net radiation is about 15% and 12% with mean error less than about 11% and 9% respectively, which suggested the proposed approach an valid and operational mean in quantitative reversion of surface long wave radiation with satisfied accuracy.