Investigating potential benefits of future sub-L1 missions with STEREO-A

TL;DR

This study assesses the potential of future sub-L1 spacecraft missions for geomagnetic storm forecasting by analyzing STEREO-A data, developing a methodology for real-time modeling, and evaluating detection success and lead times.

Contribution

It introduces a baseline methodology for using sub-L1 in situ data to predict geomagnetic storms and evaluates its effectiveness with STEREO-A observations.

Findings

26 of 47 geomagnetic storms correctly identified

Intense storms detected with 82% accuracy

Most predictions are later and stronger than actual events

Abstract

We present the first statistical study of geomagnetic storm forecasting using in situ data from the STEREO-A spacecraft as a sub-L1 monitor. Between November 2022 and June 2024, STEREO-A crossed the Sun-Earth line, covering longitudinal and radial separations of +/-15{\deg} from the Sun-Earth line and 0.01-0.06 au from Earth. This passage provides a unique opportunity to assess future sub-L1 mission concepts by ESA, such as HENON and SHIELD. We identify 32 coronal mass ejections (CMEs) observed by both STEREO-A and L1 spacecraft. Eight of these 32 CME events are first detected at L1, indicating that radial spacecraft separations of up to ~0.05 au do not always yield lead time advantages. Furthermore, we find greater (smaller) gains in lead time when STEREO-A is east (west) of the Sun-Earth line. We develop a baseline methodology for the use of future sub-L1 in situ data to enable time-shifting and real-time modeling of the geomagnetic SYM-H index. This is run continuously over the entire time period, therefore modeling the geomagnetic response of all solar wind structures. Our methodology is empirically motivated and should be considered a first approach in addressing the use of sub-L1 data. Following this methodology, 26 of 47 observed geomagnetic storms are correctly identified from STEREO-A data. Intense events (82%, SYM-H<-100 nT) are well detected, most of which are also associated with an identified CME event. Most SYM-H minima are predicted later (72%) and stronger (58%) than those observed due to biases introduced by our methodology.