The impact of impervious surface, vegetation, and soil areas on land surface temperatures in a semi-arid region using Landsat satellite images enriched with Ndaisi method data.

Impervious surfaces are a significant issue of both urbanization and environmental assessment. However, it is a problem to classify impervious surface (IS) and soil areas as separate classes in land cover classification. The objectives of this study are to obtain impervious surface, vegetation, and soil areas clearly of an urban complex with a semi-arid climate and to better determine the relationships of IS, vegetation, and soil areas with land surface temperatures (LSTs). For this purpose, IS, vegetation, and soil areas in a semi-arid city of Turkey-Kayseri city were identified by using Normalized Difference Anthropogenic Impervious Surface Index (NDAISI) data and support vector machine (SVM) method together in the classification of different areas. Landsat 5, 7, and 8 satellite images of 1987, 2000, and 2013 were used, respectively, in this study. Afterward, the effects of these areas on LSTs were analyzed. Regression analysis was used to determine the relationships between land cover areas and surface temperatures. To better demonstrate these relationships, besides common pixel-based and classical regional-based approaches, a new density-based regional analysis approach was proposed. This study is an innovative one in terms of detecting IS and indicating relationships between land cover areas and surface temperatures in semi-arid regions. Another innovation of the study is related to the results produced. The results showed that decreasing LST values were observed with increasing IS and vegetation cover values and increasing LST values were observed with increasing soil areas. The present findings may provide significant contributions to the literature and will facilitate the development of urban planning strategies in semi-arid regions.