Abstract
Rapid urbanization and changes in land use land cover (LULC) have affected the thermal character of cities across the globe. Researchers found that Remote Sensing (RS) based urban biophysical composition-vegetation coverage, built-up area coverage, impervious surface coverage, soil fractions can significantly explain land surface temperature (LST) variation both at a local and regional scale. Taking Dhaka City, Bangladesh, as an example, this study aimed at exploring the effect of biophysical composition on LST in the year 2020 and find out Urban Heat Island (UHI) clusters with different LST setting using spatial autocorrelation technique. We found that LST presented a significant positive correlation with Normalized Difference Built-up Index (NDBI), road percentage, and negative correlation with Normalized Difference Vegetation Index (NDVI), Normalised Difference Water Index (NDWI), Normalised Difference Bareness Index (NDBaI), porosity, water percentage, open space percentage. Further, the study conducted local Moran's I and Hot Spots (Getis-Ord-Gi statistics) analysis to find out the location of UHI clusters. Findings showed that 34 % of neighborhoods fall within one of the four spatial patterns of UHI clusters. Study findings would help urban planners and policymakers to have cogent understating on the relationship between urban thermal behavior, and its surface biophysical composition at fine spatial scales and hopes to provide a valuable reference for community planning, resource allocation, and sustainable development.