Prediction of Tropical Pacific Rain Rates with Over-parameterized Neural Networks

TL;DR

This paper demonstrates that over-parameterized neural networks can effectively predict tropical rain rate distributions, including heavy tails, outperforming other methods and providing insights into key atmospheric features influencing rainfall.

Contribution

It introduces the application of over-parameterized neural networks to climate data, specifically for predicting rain rate distributions and understanding tail behavior in tropical rainfall.

Findings

Over-parameterized neural networks accurately predict rain rate distributions.

They outperform other machine learning methods in this task.

Spatial patterns of rain types are well captured across the tropical Pacific.

Abstract

The prediction of tropical rain rates from atmospheric profiles poses significant challenges, mainly due to the heavy-tailed distribution exhibited by tropical rainfall. This study introduces over-parameterized neural networks not only to forecast tropical rain rates, but also to explain their heavy-tailed distribution. The prediction is separately conducted for three rain types (stratiform, deep convective, and shallow convective) observed by the Global Precipitation Measurement satellite radar over the West and East Pacific regions. Atmospheric profiles of humidity, temperature, and zonal and meridional winds from the MERRA-2 reanalysis are considered as features. Although over-parameterized neural networks are well-known for their ``double descent phenomenon," little has been explored about their applicability to climate data and capability of capturing the tail behavior of data. In our results, over-parameterized neural networks accurately predict the rain rate distributions and outperform other machine learning methods. Spatial maps show that over-parameterized neural networks also successfully describe spatial patterns of each rain type across the tropical Pacific. In addition, we assess the feature importance for each over-parameterized neural network to provide insight into the key factors driving the predictions, with low-level humidity and temperature variables being the overall most important. These findings highlight the capability of over-parameterized neural networks in predicting the distribution of the rain rate and explaining extreme values.