A Multimodal Encoder–Decoder Neural Network for Forecasting Solar Wind Speed at L1

Abstract The solar wind, accelerated within the solar corona, sculpts the heliosphere and continuously interacts with planetary atmospheres. On Earth, high-speed solar wind streams may lead to severe disruption of satellite operations and power grids. Accurate and reliable forecasting of the ambient solar wind speed is therefore highly desirable. This work presents an encoder–decoder neural network framework for simultaneously forecasting the daily averaged solar wind speed for the subsequent 4 days. The encoder–decoder framework is trained with the two different modes of solar observations. The history of solar wind observations from prior solar rotations and EUV coronal observations up to 4 days prior to the current time forms the input to two different encoders. The decoder is designed to output the daily averaged solar wind speed from 4 days prior to the current time to 4 days into the future. Our model outputs the solar wind speed with rms errors (RMSEs) of 55, 58, 58, and 58 km s −1 and Pearson correlations of 0.78, 0.66, 0.64, and 0.63 for 1 to 4 days in advance, respectively. While the model is trained and validated on observations between 2010 and 2018, we demonstrate its robustness via application on unseen test data between 2019 and 2023, yielding RMSEs of 53 km s −1 and Pearson correlations of 0.55 for a 4 day advance prediction. Our encoder–decoder model thus produces much improved RMSE values compared to previous works and paves the way for developing comprehensive multimodal deep learning models for operational solar wind forecasting.