LAUSR

Abstract Measuring high spatial/temporal land surface temperature (LST) is a significant key in urban climate studies. Current satellite thermal sensors have a trade-off in their spatial and temporal resolution. The objective of this study is to develop an adaptive thermal sharpening (TsHARP) algorithm to downscale LST by utilizing impervious surface (IS) index in urban areas. The study utilizes the aggregated sharpened LST retrieved from the Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER) L1B image derived over Kuala Lumpur, Malaysia on the 14th Feb 2003. Kuala Lumpur is among the fastest growing metropolitan regions in terms of population and economy in South-East Asia. With the application of the proposed downscaling procedure on ASTER data covering Kuala Lumpur, it was found that all the evaluated resolutions attained superior results compared to the traditional methods of over urban imagery particularly for environments with heterogeneous land covers. Furthermore, the proposed adaptive TsHARP method showed optimal performance at the 240 m resolution with a correlation coefficient of R2 = 0.85 and RMSE 1.83 °C. The model enhances the capability of the adapted TsHARP in unmixing temperature with decreasing resolution. It concluded that applying such improved technique to ASTER satellite data is an urgent and cost-effective application, which is indeed very timely to input for urban/cities landscape planning for metropolitan regions around the world.