Spatiotemporal assessment of the correlation between surface thermal field variance and human thermal discomfort in Dhaka

Rapid urbanization in tropical megacities has significantly altered local energy balances, intensifying severe microclimatic degradation. While traditional urban climatology depends on Land Surface Temperature (LST) and the Urban Heat Island (UHI) effect, these physical metrics often fail to detect the true human thermal discomfort experienced by residents. To address this critical gap, this study evaluates the spatiotemporal dynamics of human-centric urban thermal discomfort in Dhaka over a two-decade period from 2005 to 2025. Utilizing the Google Earth Engine (GEE) Python API, this study combines Landsat 5 and 8 optical-thermal data with ERA5-HEAT atmospheric reanalysis. This study uniquely quantifies localized structural degradation through the Urban Thermal Field Variance Index (UTFVI). It applies bilinear spatial interpolation to downscale coarse atmospheric data into high-resolution (30m) Universal Thermal Climate Index (UTCI) gradients, effectively translating surface anomalies into actual heat-health exposure. The findings reveal a rapid deterioration of Dhaka's urban thermal environment. Densely built-up areas saw a dramatic shift in peak LSTs over two decades, climbing from 32°C to over 40°C. At the same time, heat stress for people intensified. By 2025, average temperatures (UTCI) will hit 37.8°C to 38.1°C, putting residents in the 'Strong Heat Stress' category. Data from 500 random locations shows a strong, steady link between extreme environmental heat (UTFVI > 0.15) and the highest levels of physical discomfort. These findings confirm that the localized depletion of permeable surfaces directly forces severe atmospheric heat stress, providing a replicable, dual-index framework to guide targeted microclimatic interventions and climate-resilient urban planning in the Global South.