A review of artificial intelligence expert systems for environmental surveillance and disaster management

Natural disasters are occurring more frequently and with greater intensity worldwide due to the accelerating effects of climate change. Conventional disaster management strategies often fail to address these complex, rapidly evolving hazards. This review aims to fill the gap by systematically analyzing studies published between 2021 and 2025, identifying technological trends, examining the strengths and shortcomings of current systems, and outlining future directions for research and implementation. This review, covering the period between 2021 and 2025 included 26 primary study papers that were investigated out of 7,462 search articles. The analysis shows a clear progression from traditional rule-based models to more flexible hybrid architectures that combine expert reasoning with machine-learning capabilities. These systems have demonstrated valuable contributions to early-warning systems, real-time monitoring, and hazard prediction, particularly in data-limited or uncertainty-prone environments. Despite their potential, several limitations persist, including data inconsistency, limited scalability, uneven integration into national emergency frameworks, and challenges related to transparency and ethical governance. By synthesizing existing trends and identifying gaps in current practice, this review highlights future research opportunities, such as explainable hybrid models, edge-based deployment, collaborative global data infrastructures, and stronger policy frameworks. These insights contribute to a clearer understanding of how expert systems can support more resilient and adaptive disaster-management strategies in the future.