Machine Learning for Climate Precipitation Prediction Modeling over South America

Abstract

Many natural disasters in South America are linked to meteorological phenomena. Therefore, forecasting and monitoring climatic events are fundamental issues for society and various sectors of the economy. In the last decades, machine learning models have been developed to tackle different issues in society, but there is still a gap in applications to applied physics. Here, different machine learning models are evaluated for precipitation prediction over South America. Currently, numerical weather prediction models are unable to precisely reproduce the precipitation patterns in South America due to many factors such as the lack of region-specific parametrizations and data availability. The results are compared to the general circulation atmospheric model currently used operationally in the National Institute for Space Research (INPE: Instituto Nacional de Pesquisas Espaciais), Brazil. Machine learning models are able to produce predictions with errors under 2 mm in most of the continent in comparison to satellite-observed precipitation patterns for different climate seasons, and also outperform INPE’s model for some regions (e.g., reduction of errors from 8 to 2 mm in central South America in winter). Another advantage is the computational performance from machine learning models, running faster with much lower computer resources than models based on differential equations currently used in operational centers. Therefore, it is important to consider machine learning models for precipitation forecasts in operational centers as a way to improve forecast quality and to reduce computation costs.

A New Look into the South America Precipitation Regimes: Observation and Forecast

“…Given the topographic heterogeneity and diverse atmospheric systems acting on the continent, the systematic errors of the model may be associated with the misrepresentation of regional features, such as the Andean topography and the mesoscale/local scale phenomena as the formation of convective complexes, orographic rains, and breeze circulations. In this scenario, machine learning is a promising methodology, given that seasonal predictions of precipitation for SA with the neural networks technique have resulted in up to a 75% error reduction in estimates [240].…”

Multiple Statistical Model Ensemble Predictions of Residual Chlorine in Drinking Water: Applications of Various Deep Learning and Machine Learning Algorithms

“…Application of statistical models is on the increase in a number of areas such as prediction of precipitation ( [12,[38][39][40][41][42], river flow ( [43][44][45][46][47]), and temperature ( [48,49], and [50]). Some recent studies on modelling water quality using statistical methods include Wadkar and Kote [51], Li et al [52], García-Ávila et al [29], and De Santi et al [53]).…”

Machine learning and artificial intelligence models development in rainfall-induced landslide prediction

“…There were four ML and one AI algorithm employed in this study. They are distributed random forest (DRF) [30], generalized linear model (GLM) [31], extreme gradient boosting (XGBoost) [32], generalized boosting machine (GBM) [33], and deep learning (DL) [34]- [36]. These algorithms were trained as binomial classifiers as characterized by the training datasets of each method.…”