A Scalable Remote Sensing and Machine Learning Framework for High-Resolution Land Surface Temperature Modeling

Understanding urban heat trends, biolodiversity losss, and effects of climate change on ecosystems depends on knowledge of land surface temperature. This work integrates multisensor satellite images with socioenvironmental and vegetation characteristics using machine learning to provide a scalable framework for high-resolution LST estimation and further explore its effects on vegetation, water, population, and adaptations. We combined thermal data from the moderate-resolution imaging spectroradiometer (MODIS), optical imaging from Sentinel-2, and auxiliary information, such as precipitation, altitude, and population density using the Google Earth Engine platform into a single geographic model. A random forest regression approach is used to forecast LST, designed on 4305 samples and tested on 1076 randomly dispersed sites at 100-m resolution. The model shows good predictive ability with a root-mean-squared error of 0.624 °C and a mean absolute error of 0.484 °C. Significant correlations are found among surface temperature, topography, rainfall, vegetation types, the hydrological cycle and human activity through spatial analysis. The generated fine-scale LST maps provide practical information for reducing urban heat, adapting to climate change, and spatial planning in highly populated coastal zones. This study highlights the potential of cloud-based geospatial platforms and machine learning for precise, effective, and scalable environmental monitoring in complex coastal systems.