Classification of Local Climate Zones Using ASTER and Landsat Data for High-Density Cities

Abstract

The local climate zone (LCZ) scheme provides a standard method to conduct urban heat island studies, in which urban landscapes are classified into different LCZs according to urban structures, land cover, and construction materials. Based on the LCZ classification scheme, the World Urban Database and Access Portal Tools (WUDAPT) is a new initiative to generate LCZ maps of cities worldwide with the use of freely available Landsat data. This paper aims to evaluate the performance of the original WUDAPT method in LCZ mapping for high-density cities. To further improve LCZ mapping accuracy for high-density cities, we investigate the usage of both freely available Landsat and advanced spaceborne thermal emission and reflection radiometer (ASTER) satellite data to generate better LCZ mapping results. Experiments on two high-density Chinese cities, Guangzhou and Wuhan, showed that combining Landsat and ASTER data can improve the overall performance of LCZ mapping results, especially for urban areas. This finding indicates that further applications of the WUDAPT method for high-density cities can include both ASTER and Landsat data.