REMOTE SENSING UTILIZATION FOR THE MODELLING OF LAND SURFACE TEMPERATURE FOR SUSTAINABLE CITY DEVELOPMENT

Abstract

Land Surface Temperature (LST) is a critical input for climatological, hydrological, agricultural, and change detection models. In the current study, Baghdad LST was evaluated using remotely sensed thermal infrared data. This study examines the extraction of both land surface emissivity (LSE) and LST from images taken on October 2001, October 2013 and October 2021 across the study area, as well as the link between vegetation abundance. To calculate LST over the study region, the Plank equation was utilized and the variety of LSE was explored and retrieved using the NDVI threshold approach. The findings show that in 2001, 2013 and 2021, respectively, the Land Surface Temperature (LST) ranges from 52 to 11, 53 to 23 and 59 to 29 degrees Celsius. Barren area had the lowest LSE value, while the lush vegetation had the highest LSE value, which was equivalent to 0.99. Ground observation points and Google Earth imagery were used for accuracy assessment of LULC maps. Barren land and building areas are associated with the highest temperatures that range from 52, 53, and 59degrees Celsius, while forests and water bodies have lower temperatures that range from 11, 23, and 29 degrees Celsius respectively, in 2001, 2013 and 2021. Also, the research reveals that low temperatures have a negative correlation with NDVI and NDWI, conversely high temperatures have a positive correlation with BUI.