Urban Morphology as a Driver of Surface Thermal Patterns: A Multi-Sensor Remote Sensing Assessment (2014–2024)

Urban morphology—the spatial configuration of built structures, vegetation, and water bodies—plays a decisive role in shaping surface thermal behavior in rapidly expanding tropical cities. Dhaka, one of the world’s most densely populated megacities, illustrates how physical form influences thermal exposure. This study examines the relationship between urban morphology, vegetation dynamics, and Land Surface Temperature (LST) using multi-sensor remote sensing data from 2014–2024. Landsat 8 and Landsat 9 Collection 2 Level-2 thermal and surface reflectance products were processed to generate annual composites representing built-up intensity, vegetation cover, and water presence. Morphological characteristics were quantified at the ward level using the NB_BUDENS building-density layer from the EO4SD Urban Morphology dataset as a proxy for structural compactness and impervious surface dominance. Although NB_BUDENS is not a thermal variable, it effectively captures how dense built-form configurations restrict ventilation, reduce shading, and limit ecological cooling capacity. High-density wards consistently exhibit elevated LSTs—typically 1–2 °C warmer than adjacent vegetated or water-rich areas. Vegetation provides measurable cooling effects, while water bodies act as localized thermal buffers. NDVI–LST analysis reveals a modest but consistent negative correlation (r = –0.20), with recent years clustering toward lower NDVI values and higher LSTs, indicating increasingly heat-prone conditions. Dual-axis trend analysis shows stable NDVI values alongside a gradual upward LST trend, underscoring the insufficiency of existing green infrastructure. Together, these findings demonstrate how remote sensing can quantify morphology-driven thermal exposure and support climate-responsive urban planning in tropical megacities.