COMPARATIVE ASSESSMENT OF IMPERVIOUS AREAS EXTRACTION USING LANDSAT 8 DATASET

Abstract

Impervious surfaces are a good representative of urban and built-up areas. The present work is done to carried out validation of Normalized Difference Anthropogenic Impervious Surface Index (NDAISI) to extract and mapping of impervious surfaces in different environmental conditions. For this, existing NDAISI approach has been used in different study areas to extract impervious surfaces. Previously, this index was used for Dehradun area which is a valley region. The result which obtained for Dehradun area was quite good because this index separated bare soil from impervious surfaces more effectively and provide a very good result which extract impervious areas about 20% far better than extracted by other existing indices. In this study, two different regions have been selected. The first region belongs to Delhi and surrounding region which is a well urbanised area and the 2nd area is Aravalli range which is a hilly region. In the present work, Otsu’s method was used for the validation of Normalized Difference Anthropogenic Impervious Surface Index (NDAISI) for extraction and mapping of impervious surfaces. After the validation was carried out for both the study areas it is found that for Delhi region, due to presence of anthropogenic impervious surfaces in abundance, the validation of NDAISI approach gives a very good result but in Aravalli range, due to presence of natural impervious surfaces in abundance validation of NDAISI approach is not provide a better result. For Aravalli range, BUC index approach provide more accurate result. Thus, it can be concluded that NDAISI gives good result for anthropogenic impervious area, however does not give satisfactory results for natural impervious surfaces. In the present study, it is found that BUc gives better results for natural impervious surfaces