Prediction of Annual Snow Accumulation Using a Recurrent Graph Convolutional Approach

TL;DR

This paper introduces a graph attention network-based model to predict annual snow accumulation from radar data, demonstrating robustness with fewer input data points and larger datasets, advancing polar ice analysis.

Contribution

The study develops a novel graph attention network approach for snow accumulation prediction, improving data efficiency and scalability over previous methods.

Findings

Model maintains performance with fewer input data points.

Large datasets only slightly reduce prediction accuracy.

Graph attention networks outperform previous temporal graph convolutional networks.

Abstract

The precise tracking and prediction of polar ice layers can unveil historic trends in snow accumulation. In recent years, airborne radar sensors, such as the Snow Radar, have been shown to be able to measure these internal ice layers over large areas with a fine vertical resolution. In our previous work, we found that temporal graph convolutional networks perform reasonably well in predicting future snow accumulation when given temporal graphs containing deep ice layer thickness. In this work, we experiment with a graph attention network-based model and used it to predict more annual snow accumulation data points with fewer input data points on a larger dataset. We found that these large changes only very slightly negatively impacted performance.