Short Term Prediction of Rainfall in Columbia Using a Generative Modeling Approach

This study focused on the use of artificial intelligence to predict precipitation in Columbia. Accurate prediction of rainfall is a challenging problem. This is a serious issue due to the numerous deaths every year due to events caused directly by excess rainfall. While weather forecasting has advanced significantly throughout the years there is still significant room for improvement. One of the main factors detrimental to accurate weather prediction is the multitude of variables required to achieve accurate prediction. Our research seeks to use publicly available satellite imagery to achieve accurate prediction of rainfall with relatively little input data. We seek to achieve that goal by using a custom generative modelling approach to make predictions.