Assessment on the Relationship of Spectral Indices With Land Surface Temperature Using Google Earth Engine: A Case Study of Chitwan District, Nepal

Land surface temperature (LST) is a significant characteristic that influences the Earth's climate system and the flow of energy between land and atmosphere. Land cover variations, as measured by spectral indices, have a substantial influence on LST. This study uses Google Earth Engine (GEE) to analyze the link between LST and several spectral indices in Nepal's Chitwan area. Chitwan was chosen for its diversified scenery, which includes thick forest within Chitwan National Park, permanent water bodies, urban areas, and continuous urbanization with rising temperatures. Landsat 8 data were used to generate the Normalized Difference Built-up Index (NDBI), Normalized Difference Water Index (NDWI), Enhanced Vegetation Index (EVI), Normalized Difference Bareness Index (NDBaI), and LST. All data collecting, processing, and extraction were done using the GEE platform. The spatial analysis discovered considerable changes in LST across the district, ranging from 18.86°C in vegetated regions to 40.69°C in urban centers and bare fields. The distribution of spectral indices revealed additional information: high EVI values indicated healthy vegetation, whereas high NDBI values linked to built-up regions. The NDWI successfully mapped water bodies, whereas NDBaI found regions with little vegetative cover. Pearson's correlation analysis supported the predicted relationships: lower LST with vegetation and water (EVI, NDWI: negative correlation), and higher LST with built-up areas (NDBI: positive correlation). This research emphasizes the use of remote sensing for analyzing land cover-LST interactions, as well as the role of plant cover in moderating urban heat island impacts.