From theory to practice: Integrated Multi-Scale Geomatic and Artificial Intelligence Modeling of Urban Heat Islands for Climate Adaptation in Latin American Cities

Latin America is among the most urbanized regions in the world, where rapid and often unplanned urban growth has intensified climate-related challenges, particularly the Urban Heat Island (UHI) effect. Increasing thermal stress in cities affects public health, energy consumption, and environmental sustainability, underscoring the need for integrated modeling approaches that support urban climate adaptation. In this context, the Latin American Society of Remote Sensing and Spatial Information Systems (SELPER), in collaboration with researchers from the International Society for Photogrammetry and Remote Sensing (ISPRS), promotes the use of Earth Observation (EO), remote sensing, and geospatial technologies to improve the understanding of climate-driven urban processes.

So far, the first collaborative stage has analyzed thousands of 30 m resolution Landsat 5 and Landsat 8 images covering 16 large Latin American megacities in six countries, home to approximately 73 million inhabitants. The results reveal common patterns among these cities that include: diffuse urban development models, spatially and temporally heterogeneous behavior, progressive degradation and fragmentation of forested green areas, which impacts blue-green infrastructures, marked variability in construction materials and cover, land use, and urban morphology that influence surface thermal responses, including the formation of heat islands or urban cooling islands. The findings highlight the limitations of analyses at single scales and underscore the need to improve analysis methodologies through integrative frameworks across multiple scales.

Based on this new regional knowledge, this study proposes an integrated modelling framework based on geomatics and artificial intelligence (AI) for urban climate adaptation. Geomatics, which integrates geographic information systems (GIS), remote sensing, and spatial analysis, provides a comprehensive approach to examining UHI dynamics at the spatial scale.

Our research is now going to take on two new branches. First, we must continue to demonstrate the applicability and importance of GeoAI intelligence and machine learning techniques to support the efficient processing and integration of EO into decision-making. By linking observation, analysis, and exploratory predictive modeling, the proposed framework improves understanding of urban heat dynamics. It supports evidence-based climate adaptation strategies, including blue-green infrastructure enhancement and climate-resilient urban planning in Latin American cities.