Extension of Convolutional Neural Network along Temporal and Vertical Directions for Precipitation Downscaling

TL;DR

This paper extends 2D CNN models for precipitation downscaling by incorporating temporal and vertical dimensions, demonstrating improved accuracy in precipitation estimation using 3D CNN variants.

Contribution

It introduces 3D CNN models along temporal and vertical axes for climate data downscaling, showing enhanced performance over traditional 2D CNNs.

Findings

3D-CNN-Vert achieved the lowest RMSE and highest NSE.

Both 3D-CNN-Time and 3D-CNN-Vert outperformed 2D CNN.

Vertical extension yielded the best precipitation estimates.

Abstract

Deep learning has been utilized for the statistical downscaling of climate data. Specifically, a two-dimensional (2D) convolutional neural network (CNN) has been successfully applied to precipitation estimation. This study implements a three-dimensional (3D) CNN to estimate watershed-scale daily precipitation from 3D atmospheric data and compares the results with those for a 2D CNN. The 2D CNN is extended along the time direction (3D-CNN-Time) and the vertical direction (3D-CNN-Vert). The precipitation estimates of these extended CNNs are compared with those of the 2D CNN in terms of the root-mean-square error (RMSE), Nash-Sutcliffe efficiency (NSE), and 99th percentile RMSE. It is found that both 3D-CNN-Time and 3D-CNN-Vert improve the model accuracy for precipitation estimation compared to the 2D CNN. 3D-CNN-Vert provided the best estimates during the training and test periods in terms of RMSE and NSE.