CLARE: Classification-based Regression for Electron Temperature Prediction

TL;DR

CLARE is a novel classification-based machine learning model that predicts electron temperature in Earth's plasmasphere with higher accuracy and uncertainty estimation, using satellite data and geomagnetic indices.

Contribution

The paper introduces CLARE, a classification-based regression approach that improves electron temperature prediction accuracy and provides uncertainty estimates, applied to satellite data.

Findings

Achieves 69.67% accuracy within 10% of ground truth.

Outperforms traditional regression models by 6.46%.

Effective during geomagnetic storm periods.

Abstract

Electron temperature (Te) is an important parameter governing space weather in the upper atmosphere, but has historically been underexplored in the space weather machine learning literature. We present CLARE, a machine learning model for predicting electron temperature in the Earth's plasmasphere trained on AKEBONO (EXOS-D) satellite measurements as well as solar and geomagnetic indices. CLARE uses a classification-based regression architecture that transforms the continuous Te output space into 150 discrete classification intervals. Training the model on a classification task improves prediction accuracy by 6.46% relative compared to a traditional regression model while also outputting uncertainty estimation information on its predictions. On a held out test set from the AKEBONO data, the model's Te predictions achieve 69.67% accuracy within 10% of the ground truth and 46.17% on a known geomagnetic storm period from January 30th to February 7th, 1991. We show that machine learning can be used to produce high-accuracy Te models on publicly available data.