Assessment of impervious surface in border area using multi-index from landsat 8 and fuzzy c means

Abstract

This paper describes the process to generate an impervious surface map from Landsat 8 multi-index, namely urban index (UI), normalized difference built-up index (NDBI), normalized difference vegetation index (NDVI), and land surface temperature (LST). These multi-index and LST were input into fuzzy c means (FCM) clustering algorithm to create four fuzzy raster maps, namely high UI, high NDBI, low NDVI, and high LST. The output from FCM served as input for the fuzzy overlay method (FOM) resulted in impervious surface and non-impervious surface classes in the form of a raster map with 30 m x 30 m spatial resolution. The western border between Bogor Municipality and Bogor Regency is chosen as a study area because of data imbalance in both municipal authorities to delineate urban regions in conventional methods. Therefore FCM and FOM were applied to tackle the uncertainty might come from data imbalance and serve as time-saver methods to map impervious surface. The final impervious surface map yielded from the aforementioned methods, shows 82% overall accuracy. Results show that the impervious surface area in Bogor Municipality western border side is around 11.33 km2 (66.45%), while the impervious surface area in Bogor Regency border side is around 14.35 km2 (62.2%).