Spatial and Temporal Fusion for Urban Land Surface Temperature Image Mapping

Land surface temperature (LST) is one of the most important parameters in the physical processes of surface energy and water balance. Due to the payload limit and the technical bottleneck of the sensor, it is difficult to obtain land surface temperature products with both high spatial and high temporal resolutions, which restricts applications of land surface temperature products. Based on MODIS and ETM+ images, this paper combines the enhanced spatial and temporal adaptive reflectance fusion model(ESTARFM) with the non-linear disaggregation procedure for radiometric surface temperature (NL-DisTrad) to generate daily land surface temperature products of urban area at a resolution of 60m. We use the actual land surface temperature products in July 9 and October 13 of 2002 over Wuhan urban areas to verify the quality of fusion products. It is shown their R² are as high as 0.80 and 0.86, respectively, and the RMSE are 2.65 K and 1.78 K, respectively. The results show that the proposed algorithm has fine prospects in the spatial and temporal land surface temperature fusion applications in urban areas.