Comparative Analysis of Machine Learning Methods in Climate Data Extraction and Forecasting

Climate forecasting plays a critical role in understanding and mitigating the impacts of climate change. Advances in machine learning (ML) have significantly enhanced the accuracy of climate projections, particularly when dealing with large and complex datasets derived from regional climate models (RCMs) and global climate models (GCMs). This study investigates the application of various ML techniques for predicting precipitation based on temperature data extracted from climate datasets. We evaluate a range of regression models, including linear models and advanced ensemble methods such as Gradient Boosting and XGBoost, along with deep learning approaches like Convolutional Neural Networks (CNNs). Through a rigorous evaluation of these models, we focus on their performance as measured by Mean Squared Error (MSE) to identify the most effective techniques for capturing the intricate relationships within climate data. By analyzing existing studies and performance metrics, we aim to provide insights into the capabilities and limitations of these ML models in improving climate predictions. Furthermore, this research highlights the need for incorporating additional climate variables and optimizing model configurations to enhance predictive accuracy in climate forecasting.