Analytical and AI-based approaches to weather events in business: a survey

Abstract

One of the most noticeable aspects of climate change has been the increase in extreme weather events, mostly during recent years. Environmental changes and disasters resulting from these events, including hurricanes, floods, and heat waves, have caused significant ripples across industries. Understanding and minimizing the impact of these weather phenomena on logistics processes have become essential as their frequency and intensity increase. There have been more instances of delayed product deliveries because of extreme weather conditions. Unexpected events can disrupt parts of the transportation network, damage transport infrastructure, and cause congestion. This survey, based on previous studies, found that changing weather patterns over the past few years negatively affected several sectors, including agriculture, logistics, and manufacturing. Advanced machine learning models are needed to forecast shipment delivery delays and to leverage AI-driven optimization techniques and analytical models to mitigate the impact of extreme weather disruptions across multiple operations. The previous studies showed that several techniques are applied for improving the prediction of weather due to climate change and solving the effect of climate change on multiple sectors using 79 studies. The applied techniques, including LSTM and CNNs are used to tackle the difficulty of data structure to improve weather prediction; regression analysis, equation modeling, and DID analysis are applied for calculate the economic losses. Finally, optimization and simulation, Monte Carlo, and mixed linear programming are applied to optimize supply chain, and the weather events may also cause delivery delays which cause economic losses for several sectors.