Regresar

The land surface temperature retrieval in arid and semi-arid regions

Abstract:

Land surface temperature (LST) is the net result of the climatic energy balance to the environmental process. The LST with high accuracy can offer important information to monitor the climate changes. The C. Coll's split-window algorithm for LST has been used in European regions with good accuracy based on the AVHRR images. In this paper, the further modification and development was made with the C. Coll's Algorithm for retrieving from MODIS data, which combines surface emissivity with corresponding Normalized Difference Vegetation Index (NDVI) threshold. In Arid/Semi-Arid region, the complex land covers in spatio-temporal make the accuracy retrieving more difficult. As a validation of the model, the retrieved LST is compared with meteorological data and MOD11 which are released by LP DAAC. The results show that the accuracy of the Coll's Algorithm for retrieving LST is significantly improved with of the MOD11 products and similar with the observation data, while they display the clear trapezoidal relationship between LST and NDVI, while MODIS products lost this information. The developed Coll's Algorithm is suitable for the LST retrieval and the retrieval map can properly present LST characteristics of spatio-temporal changes with high accuracy in arid and semi-arid regions.