Document Type: Original Article
Department of Geomatics, Computer Science and Mathematics, Hochschule für Technik Stuttgart, Germany
Various building indices to identify and extract sealed surfaces have been developed and implemented by various authors. Previous research has shown that building indices are easy to implement since they do not use complex algorithms and therefore can be used as quick methods for monitoring impervious surfaces. The aim of this study is to assess the ability of selected indices to identify sealed surfaces. Also, previously, authors have posted results that building indices face difficulty in distinguishing between sealed surfaces and bare lands owing to the spectral similarity between these two land covers. Additionally, it has been concluded by some researches that the performance of building indices also depends on the time of image capture i.e. dry and wet seasons. In this study, we implement 6 selected indices using sentinel 2 data covering Nürtingen city in Stuttgart, investigate and compare their performance in different times of the year. Google earth engine is used to conduct these investigations.