Past Speakers

Abstract: We present an innovative multi-fidelity machine learning (ML) method for forecasting extreme space weather events. Multi-fidelity technique is a powerful ensemble tool to improve the accuracy of the predictions, and in fact it has been widely used for physical model predictions. One of the major difficulties for implementing this technique on ML methods is that typically ML models do not provide the uncertainty associated with a prediction. Hence, an ACCRUE (ACCurate and Reliable Uncertainty Estimate) based Gated Recurrent Unit (GRU) method is developed, to forecast the uncertainty for the model predictions simultaneously. We have implemented this method on two space weather applications, i.e., 1) a one-to-six-hour lead-time model that predicts the value of Disturbance storm time (Dst) using solar wind (SW) data; and 2) an geoelectric field (E) model with 1-hour leading time using SW and SuperMag data. The first model can forecast Dst 6 hours ahead with a root-mean-square-error (RMSE) of 13.54 nT during more than 50 strong storm events in recent two solar cycles. This significantly outperforms physics-based and prior empirical models. The geoelectric field model 1-hr prediction also agrees well with ground truth data generated from SuperMag data and Magnetotellurics (MT) surveys.

Bio: Dr. Andong Hu is currently a postdoc researcher in Cooperative Institute for Research in Environmental Sciences at the University of Colorado Boulder, with a NASA project Ensemble Learning for Accurate and Reliable Uncertainty Quantification. Before that, he was working in Multiscale Dynamic (MD) Group, Centrum Wiskunde & Informatica(CWI), the Dutch national research institute of mathematics and computer science, in Amsterdam, with Jannis Teunissen and Enrico Camporeale. He achieved his Geoscience doctoral degree (2020) in Satellite Positioning for Atmosphere, Climate and Environment (SPACE), Royal Melbourne Institute of Technology (RMIT) University, supervised by Kefei Zhang and Brett Carter. His major interests cover Global Navigation Satellite System (GNSS), Machine Learning and Space Weather.