Microclimate analysis in Basse Santa Su, The Gambia: modeling temperature, humidity, and heat stress in an extreme climate

Introduction: Heat exposure poses an increasing threat to human health, particularly in African low- and middle-income countries, where rapid urbanization, limited adaptation infrastructure, and climate change vulnerability merge. However, fine-scale meteorological data in this region are scarce, limiting heat exposure assessments. Moreover, compound humid heat remains largely unexplored in this context. This study aims to assess humid heat exposure in Basse Santa Su, The Gambia, a region highly vulnerable to humid heat, by generating spatial predictions informed by high-resolution microclimate measurements.

Methods: Over one year, a fixed network of low-cost measurement devices mounted across Basse Santa Su (approximately 11km²) collects time-resolved meteorological parameters (temperature, humidity, solar radiation, atmospheric pressure, wind speed and direction). Multilinear land-use regression (LUR) models will estimate spatial and temporal patterns of heat, humidity, and heat-stress distribution across the study area. Model predictors will include climate variables from ERA5-Land reanalysis and global high-resolution remote sensing data on relevant characteristics such as land-use, vegetation, topography and urban surface geometry.

Results: In 2025, the fixed measurement network mounted at 12 locations in Basse Santa Su recorded an average daily temperature of 29.6ºC with 38.6% relative humidity (RH) in the dry season (November–May), and an average daily temperature of 28.8ºC with 78.6% RH in the rainy season (June–October). The predictive models will estimate high-resolution, daily and hourly single weather variables (ambient temperature and relative humidity) and combined heat stress indices (e.g. Wet Bulb Globe Temperature (WBGT), physiological equivalent temperature (PET)) over the whole study area. These maps will assess the spatial and temporal variability in humid heat and identify high-risk neighbourhoods, contextual variables, and periods or seasons associated with higher or lower exposure.

Conclusion: This study evaluates the feasibility of combining a low-cost microclimate measurement network with a land-use regression modeling approach to characterize fine-scale spatial variability in temperature, humidity, and heat stress in a data-scarce, extreme climate setting. The resulting high-resolution humid heat exposure estimates will provide a critical foundation for further applications, such as heat-health impact assessments and targeted adaptation strategies and interventions in Basse Santa Su and comparable settings.