Multiple‐Hour‐Ahead Forecast of the Dst Index Using a Combination of Long Short‐Term Memory Neural Network and Gaussian Process. Issue 11 (28th November 2018)

• Record Type: Journal Article
• Authors: Gruet, M. A.; Chandorkar, M.; Sicard, A.; Camporeale, E.
• Is Part Of: Space weather. Volume 16:Issue 11(2018)
• Page Start: 1882
• Page End: 1896
• Publication Date: 2018-11-28
• Subjects: neural networks -- Gaussian process -- multiple‐hour‐ahead forecasts -- geomagnetic index; Space environment -- Periodicals; 551.509992
• DOI: 10.1029/2018SW001898
• Languages: English
• ISSNs: 1542-7390

Abstract

Abstract: In this study, we present a method that combines a Long Short‐Term Memory (LSTM) recurrent neural network with a Gaussian process (GP) model to provide up to 6‐hr‐ahead probabilistic forecasts of the Dst geomagnetic index. The proposed approach brings together the sequence modeling capabilities of a recurrent neural network with the error bars and confidence bounds provided by a GP. Our model is trained using the hourly OMNI and Global Positioning System (GPS) databases, both of which are publicly available. We first develop a LSTM network to get a single‐point prediction of Dst. This model yields great accuracy in forecasting the Dst index from 1 to 6 hr ahead, with a correlation coefficient always higher than 0.873 and a root‐mean‐square error lower than 9.86. However, even if global metrics show excellent performance, it remains poor in predicting intense storms (Dst < −250 nT) 6 hr in advance. To improve it and to obtain probabilistic forecasts, we combine the LSTM model obtained with a GP and evaluate the hybrid predictor using the receiver operating characteristic curve and the reliability diagram. We conclude that this hybrid methodology provides improvements in the forecast of geomagnetic storms, from 1 to 6 hr ahead. Key Points: First use of a Long Short‐Term Memory network to provide single‐point prediction of the Dst index, up to 6 hr ahead Development of a method that combines neural network and Gaussian process to obtain a probabilistic forecast from one to 6 hr ahead Use of specific metrics to evaluate probabilistic forecast, like receiver operating characteristic curves and reliability diagram