Estimation and analysis of land surface water and heat fluxes in mountain-plain area based on remote sensing and DEM

Abstract

The income and expenses status of water along with heat on land surface determine the form and evolvement of environment to a great extent. This paper focus on the topic of estimation and analysis of land surface water and heat fluxes in mountain-plain area based on remote sensing and DEM, selecting Sunan Mountain and Zhangye Plain in China as a case study. All the biophysical parameters quantitatively retrieved from ASTER data were as the inputs of SVAT-based model—SEBS (Surface Energy Balance System), to derive the latent heat flux and daily evaporation, in which the parameterization schemes for calculating the instantaneous solar radiation and daily integrated radiation were improved by accounting for the variations in slope and azimuth of land surface and terrain shadow in mountainous areas. After preliminary validation and comparison, we consider that the estimation of land surface water and heat fluxes as well as daily evaporation shows a reliable result, which also implies that the parameterization methods and SEBS model are feasible. Then the spatial patterns of land surface heat fluxes as well as daily evaporation on these two landform systems were cross-compared. Different types of land use land cover (LULC) have their corresponding threshold concentrations, and the high vegetation density corresponds to high evaporation (daily evaporation ranges from 0mm to 7.524 mm day-1). Forest, prairie and grassland, irrigated field have higher daily evaporation, while cold desert, Gobi and droughty riverside land have low daily evaporation, some of which reach to 0mm. Results show that land surface heterogeneity has great influence on the spatial pattern of land surface water and heat fluxes, and vegetation coverage influence surface water and heat's characteristic, which figures that it is better to introduce vegetation fraction and Ts to describe land surface water and heat fluxes rather than the commonly used NDVI-Ts, though needed further analysis and validation. Besides the vegetation coverage, terrain complexity dominates the spatial distribution of land surface heat fluxes in mountainous area, which are distinctly different from plain in quantity and spatial law. The standard deviation of the net radiation flux error and latent heat flux error are almost linearly with the standard deviation of height. In certain range, they have significant correlativity with the standard deviation of slope and aspect. These correlations with net radiation flux are even more obviously. In addition, the utility and limitations of SEBS's applicability to watersheds with various land cover types and complex terrain were analyzed.