CURRENT OVERVIEW OF THE USE OF ARTIFICIAL INTELLIGENCE IN NATURAL DISASTERS: A LITERATURE REVIEW

Abstract

Natural disasters have increased in recent years due to climate change, resulting in a significant number of victims and a growing need for efficient monitoring and classification systems. This study presents an overview of the main artificial intelligence (AI) technologies and automated monitoring applied to the risk assessment phase of disaster management. It describes the main types of natural disasters, such as floods, forest fires, earthquakes, landslides, tropical cyclones, drought, hurricanes, tsunamis, deforestation, and avalanches. It also examines the technological tools and AI algorithms used for their analysis, including Synthetic Aperture Radar (SAR), Geographic Information Systems (GIS), computer vision, decision support systems (DSS), and LiDAR. Furthermore, it identifies the most commonly used AI models in research, such as support vector machines (SVMs), random forests (RF), LSTM neural networks, and decision trees (DTs). Understanding the types of disasters and the technologies used to predict and detect them will allow for the identification of new research areas. Likewise, familiarity with the models and technologies described in the scientific literature guides the development of intelligent risk management systems and improves the effectiveness of emergency response teams.