Advances in Emergency Demand Forecasting: A Comprehensive Systematic Review

Accurate emergency demand forecasting models can significantly improve decision makers' ability to make data-driven decisions during emergencies. These models help responders determine the progress of an emergency in advance and predict the potential demand for supplies, personnel, equipment, services, and other types of resources. Developing sophisticated predictive models has become vital with emerging data sources, mainly social media data and Volunteered Geographic Information (VGI). To this end, this paper first reviewed multiple reviews on emergency demand forecasting published in the same time frame and database, identifying recent advances in current forecasting models and methods and shortcomings of existing studies. At the same time, these related review studies failed to fully exploit the potential of accurate demand forecasting while failing to fully integrate cutting-edge techniques or provide a detailed categorization of demand forecasting methods. In view of these challenges, this paper focused on emergency demand forecasting and aims to fill these research gaps. By classifying, analyzing, and comparing research findings from 2020 to 2024, a comprehensive evaluation of diverse emergency demand forecasting methods is presented, drawing on publications from major databases such as Web of Science, Ei Compendex, and Google Scholar. In addition, this paper provides systematic and comprehensive emergency demand forecasts for natural disasters, general disasters, and public health emergencies, among others. It thoroughly examines critical inputs and outputs, compares various forecasting models, and integrates fundamental and technical analyses to improve the understanding and accuracy of forecasting methods. Ultimately, emergency demand forecasting is crucial in environmental research because it helps improve preparedness to respond to emergencies and can mitigate potential ecological and social impacts.