Abstract Over the past four decades, climate change has emerged as a critical global concern, disrupting sea levels, temperatures, Arctic ice, biodiversity, and other ecological systems. Simultaneously, the rapid evolution of artificial intelligence (AI) and machine learning (ML) has driven their use as essential tools in forecasting climate transformations, aiding disaster preemption and environmental management. This paper examines current AI applications in predicting climate change phenomena and climate-induced disasters globally. It focuses on AI models used to forecast wildfires, floods, polar ice dynamics, and other key climate domains. Central to this study is the integration of advanced ML paradigms—such as neural networks, hybrid optimization methods, and deep learning—into forecasting systems. The accuracy of several ML models in predicting essential climate metrics is evaluated via global datasets and empirical benchmarks. Among the models assessed—the artificial neural network (ANN), ANN with particle swarm optimization (PSO), ANN with ant colony optimization (ACO), and ANN with bee colony optimization (BCO)—the BCO-optimized ANN demonstrated the best performance. The ACO-enhanced ANN achieved the best performance, with the lowest average root mean square error (RMSE), mean squared error (MSE), and mean absolute error (MAE) scores across the different variables tested—showing a 90.9% improvement in the RMSE and a 100.0% improvement in both the MSE and MAE values, along with a consistent R² value of 1 across all the parameters tested. This analysis underscores the significant potential of AI—particularly optimized neural architectures—in advancing climate forecasting and enabling more informed, proactive responses to climate-related challenges.
Mazumdar et al. (Thu,) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: