South America Seasonal Precipitation Prediction by Gradient-Boosting Machine-Learning Approach

Machine learning has experienced great success in many applications. Precipitation is a hard meteorological variable to predict, but it strong impact on society. Here, machine-learning technique—a formulation of gradient-boosted trees—is applied climate seasonal precipitation prediction over South America. The Optuna framework, based Bayesian optimization, was employed determine the optimal hyperparameters for gradient-boosting scheme. A comparison between forecasting among numerical atmospheric models used by National Institute Space Research (INPE, Brazil) as an operational procedure weather/climate forecasting, gradient boosting, and deep-learning techniques made regarding observation, with some showing better performance boosting