Evaluation of Urban Land Cover Indices Affecting Urban Heat Island Formation Using Advanced Geo-Spatial Techniques

Abstract

Abstract Evaluation of land surface temperature during the climate change process is important in urban planning. The urban thermal environment is closely related to land surface characteristics. The relationship between land surface properties and land surface temperature (LST) is among the current research topics. As a result of advances in geospatial and remote sensing fields, remote sensing-based spectral indices have been developed to investigate land use/land cover (LULC) effects on the urban thermal environment. The study aims to investigate the impact of remote sensing-based LULC indices on LST in Ankara metropolitan city. LST values, which are an important representation of the urban heat island, were calculated from Landsat 8 OLI/TIRS data for 2013, 2018, and 2023. Urban Thermal Field Variance Index (UTFVI) was used to define the urban heat island effect from a thermal perspective. Additionally, to define the urban heat island effect in terms of land cover characteristics, temporal-spatial changes of the LULC indices which are Normalized Difference Vegetation Index (NDVI), Normalized Difference Water Index (NDWI), Normalized Difference Built-Up Index (NDBI), Normalized Difference Impervious Index (NDISI) has been evaluated. Linear regression analysis was performed to determine the effect of LULC indices on LST. As a result of the analysis, it was determined that NDVI and NDWI had a negative correlation with LST, while NDBI and NDISI had a positive correlation. The highest correlation values belong to 2023. NDVI (R²=0.4944) and NDWI (R²=0.2666) affect the 2023 LST negatively, while NDBI (R²=0.3664) and NDISI (R²=0.6010) affect it positively. While the results show the importance of green vegetation and water surfaces in reducing LST, they show that NDISI, which is a representation of impervious surfaces, has the most impact on increasing LST. The results of the study also reveal the impact and importance of spatial patterns of LULC indices on LST.