AI in Climate Science: Leveraging XGBoost and Random Forest for Ocean Temperature Forecasting

By examining meteorological and oceanographic variables such as wind patterns, humidity, and air temperature, this study investigates the application of machine learning approaches, Random Forest Regressor and XGBoost Regressor, to forecast sea surface temperature (SST). Using an El Niño dataset, the study applies rigorous preprocessing to integrate spatial-temporal variability and address data discrepancies. Comparative research indicates that XGBoost outperforms Random Forest in terms of prediction accuracy, as seen by higher R2 scores and lower RMSE. The findings provide valuable insights into the ways that oceanic systems are being impacted by climate change and show how advanced machine learning algorithms can capture nonlinear interactions.