Artificial intelligence and geospatial techniques for mapping fire-prone areas: A review

Wildfires have emerged as a significant global challenge, driven by rising temperatures, prolonged droughts, shifting precipitation patterns due to climate change, and human-induced land-use changes such as deforestation, urban expansion, and agricultural intensification. Their increasing frequency and severity threaten ecosystems, human health, infrastructure, and economic resources, necessitating effective prediction, detection, and management strategies. This review provides a comprehensive assessment of geospatial techniques and their integration with Artificial Intelligence (AI) and Machine Learning (ML) approaches for wildfire monitoring, mapping, and prediction. Various platforms, including ground-based observations, satellite remote sensing, and unoccupied aerial vehicles (UAVs), are evaluated for their capabilities and limitations. Traditional methods, such as visual observations and ground-based sensors, are discussed, highlighting constraints in spatial coverage, timeliness, and scalability. Satellite-based remote sensing, with its wide area coverage and multi-temporal data acquisition, enables the derivation of key fire-related parameters, including vegetation indices, land surface temperature, and burned area mapping. UAV-based systems complement satellites by providing high-resolution imagery and facilitating access to remote or hazardous locations, enhancing early detection and monitoring. Furthermore, the integration of AI/ML algorithms with multi-source datasets has revolutionized wildfire modeling, enabling more accurate fire prediction and risk assessment. The review also explores emerging challenges, including data heterogeneity, real-time processing requirements, and the integration of diverse data sources, while emphasizing the need for robust computational infrastructure and collaborative frameworks. Unlike previous reviews that focus on individual aspects, this study uniquely synthesizes multi-source geospatial datasets, advanced AI/ML techniques, and bibliometric co-occurrence analysis to identify research hotspots, knowledge gaps, and future directions. By providing a holistic framework for next-generation wildfire monitoring and risk assessment, this review contributes to the development of effective fire management strategies, promoting environmental sustainability, community resilience, and informed decision-making.