Methods for Extracting Urban Construction Land Using Night-Light Data: Assessment and Application

Abstract

As a new data source, night light remote sensing data are more suitable for the background of rapid urbanization in China. Therefore, in recent years, many scholars have used the night light remote sensing data to extract and study the expansion of urban construction land. However, researchers are still trying to find better methods to avoid the internal defects of night light data. This study assessed the advantages and disadvantages of existing methods for extracting urban construction land using night light data through actual applications. The results indicated that the areas of urban construction lands extracted using threshold methods were much closer to the authoritative data, and the shape coincidence degrees were also better, while the shape similarity degrees of urban construction lands extracted using the neighborhood analysis method were much better. Further analyses revealed that we may find a better method for extracting urban construction land by considering both the intensity of night light data and the natural boundaries of cities and towns.