Mapping forest fire risks in northern Vietnam's tropical forests

Forest fires in tropical regions are increasingly influenced by climate change, posing serious threats to biodiversity, livelihoods, and ecosystem stability. This study applies the Maximum Entropy (MaxEnt) model to predict forest fire occurrence in northern Vietnam under current and future climate scenarios (Representative Concentration Pathways (RCP) 4.5 and RCP 8.5 for 2050 and 2070). Using nine environmental predictors and 20 years of active fire data from Fire Information for Resource Management System (FIRMS), we identified temperature seasonality, elevation, and slope as key drivers of fire risk. The model achieved a mean Area Under the Curve (AUC) of 0.71, with temperature seasonality contributing 54.1%, elevation 22.2%, and slope 18.1% to the predictive performance. The model revealed that areas with moderate to high fire probability are concentrated in the northwestern region and Quang Ninh province, covering approximately 13,000 km² and 9,000 km² under current conditions, respectively. Under future climate conditions, the area of high fire probability is projected to expand to 14,489 km² by 2050 for RCP 8.5, especially under high-emission scenarios. While the model demonstrates strong spatial predictive capacity, limitations include the absence of fuel and anthropogenic variables. Importantly, the findings provide actionable insights for forest managers and policymakers to prioritize high-risk zones, enhance early warning systems, and develop climate-adaptive fire mitigation strategies. This approach offers a scalable framework for fire risk assessment in tropical forests and can be extended to other regions in Vietnam and globally.