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

This study presents the first statistical assessment of using STEREO-A as a sub-L1 monitor for geomagnetic storm forecasting, demonstrating that while radial separations up to 0.05 au do not guarantee lead time advantages, a new time-shifting methodology can successfully identify 26 of 47 storms, particularly intense events, despite systematic biases in predicting the timing and strength of geomagnetic minima.

The Gist

The Big Picture: Why We Need a "Weather Watchdog" Closer to the Sun

Imagine Earth is a ship sailing through the ocean of space. Sometimes, the Sun sends out massive, invisible tsunamis of magnetic energy and charged particles called Coronal Mass Ejections (CMEs). When these hit Earth, they cause geomagnetic storms that can knock out satellites, disrupt GPS, and even crash power grids.

Right now, we have a "weather station" (satellites like DSCOVR and ACE) parked at a spot called L1, which is about 1 million miles (0.01 AU) in front of Earth. This station gives us a 10-to-60-minute warning before a storm hits.

The Problem: 10 to 60 minutes is like getting a warning that a hurricane is coming after you've already heard the first thunderclap. You don't have enough time to shut down the power grid or put satellites in safe mode.

The Solution: Scientists want to build a new "weather station" even closer to the Sun (a sub-L1 mission). This would give us hours of warning instead of minutes. Two upcoming European missions, HENON and SHIELD, plan to do exactly this.

The Experiment: Using a "Time Traveler" as a Test Pilot

We haven't launched HENON or SHIELD yet, but we had a lucky accident. The STEREO-A spacecraft, which was originally sent to orbit the Sun from a different angle, drifted right in front of Earth between late 2022 and mid-2024.

For a brief window, STEREO-A acted as a perfect test subject for these future missions. It was:

• Closer to the Sun: About 5 times closer than our current L1 station (roughly 0.01 to 0.05 AU ahead of Earth).
• Aligned: It passed directly in front of Earth, just like the future missions plan to do.

The authors of this paper used STEREO-A's data to answer two big questions:

1. Does being closer actually give us more warning time?
2. Can we use that data to accurately predict how bad the storm will be when it hits Earth?

Key Findings (The "Weather Report")

1. The "Sweet Spot" for Distance

The team found that simply being "ahead" isn't always enough.

• The Analogy: Imagine two runners on a track. If the front runner (STEREO-A) is too far ahead, they might be running on a different lane than the back runner (L1). If the storm (the runner) curves or changes lanes, the front runner might miss it entirely, or the back runner might see it first because the storm is curving toward them.
• The Result: They discovered that if a future mission is placed farther than 0.95 AU from the Sun (too far back), it sometimes sees the storm after the L1 station does. To guarantee a warning, the new station must be closer than 0.95 AU.
• The Good News: The planned HENON (0.90 AU) and SHIELD (0.86 AU) missions are close enough to the Sun to ensure they will always see the storms first.

2. The "East vs. West" Twist

The paper found a weird quirk: It matters where the station is relative to Earth's path.

• The Analogy: Think of the solar wind like a river with a strong current that twists (the Parker Spiral). If your weather station is on the "East" side of the river, the storm hits it earlier. If it's on the "West" side, the storm might hit the L1 station first, even if the weather station is closer to the source.
• The Result: Being on the East side of the Sun-Earth line generally gives a bigger time advantage. Future missions need to be smart about their orbital paths to maximize this benefit.

3. Predicting the Storm's Strength

Knowing when a storm hits is great, but knowing how strong it will be is crucial.

• The Method: The team built a computer model that took the data from STEREO-A, "shifted" it forward in time to simulate it arriving at Earth, and then ran it through a prediction engine (the TL Model) to guess the SYM-H index (a score that measures how hard the storm hits Earth's magnetic field).
• The Result:
  • Success Rate: They correctly identified 26 out of 47 storms.
  • The Heavy Hitters: They were excellent at spotting the intense, dangerous storms (82% success rate). This is the most important part for protecting our technology.
  • The Flaw: The model tended to predict the storms would hit later and be stronger than they actually were.
  • The Analogy: It's like a weather app that says, "A tornado is coming in 2 hours and it will be a Category 5!" when it actually arrives in 1.5 hours and is only a Category 3. It's better to be safe than sorry, but it means we need to tweak the math to be more precise.

What This Means for the Future

This study is a "dress rehearsal" for the HENON and SHIELD missions. It proves that:

1. It works: Putting a satellite closer to the Sun does give us more time to prepare.
2. The location matters: We need to place these satellites closer than 0.95 AU and keep them within a specific angle (less than 15 degrees) of the Sun-Earth line to ensure they see the same storms as Earth.
3. The tech is ready: We have a method to turn that early data into a forecast, even if we need to fine-tune the "strength" predictions.

In short: We are no longer guessing if a "closer-to-the-Sun" weather station is a good idea. We have the proof. The next step is to launch HENON and SHIELD to turn those extra hours of warning into real-world protection for our power grids and satellites.

Technical Summary

1. Problem Statement

Space weather forecasting faces a persistent challenge known as the "Bz problem": the inability to reliably predict the southward component of the interplanetary magnetic field ($B_z$) more than one hour in advance. Current monitoring relies on spacecraft at the Sun-Earth Lagrangian point 1 (L1), approximately 0.01 AU upstream of Earth. While L1 provides a lead time of 10–60 minutes, this is often insufficient for critical infrastructure protection.

To extend lead times, future missions (e.g., ESA's HENON and SHIELD) propose placing spacecraft closer to the Sun (sub-L1) on distant retrograde orbits (DROs). However, the optimal orbital distance and longitudinal separation required to ensure consistent lead time gains and accurate geoeffectiveness predictions remain undetermined due to a lack of multi-spacecraft observations at medium separations (1°–20°).

2. Methodology

The authors utilized a unique observational window between November 2022 and June 2024, during which the STEREO-A spacecraft crossed the Sun-Earth line, acting as a natural sub-L1 monitor. STEREO-A was positioned 0.01–0.06 AU upstream of Earth and within ±15° longitudinal separation.

The study employed the following methodological steps:

• Data Selection: Identified 32 Coronal Mass Ejection (CME) events observed by both STEREO-A and L1 spacecraft (ACE/DSCOVR).
• Statistical Analysis: Analyzed differences in arrival times ($\Delta t_s$) and intrinsic physical properties (magnetic field strength $B_{tot}$, speed, density, and the dawn-to-dusk electric field $vB_z$) between the two locations.
• Time-Shifting Pipeline: Developed a baseline methodology to shift sub-L1 data to Earth.
  • Used the OMNI low-resolution time-shift equations (King & Papitashvili, 2005).
  • Optimized parameters (specifically the phase front orientation parameter $n$) based on the 32 CME events to minimize Mean Absolute Error (MAE).
  • Implemented an ensemble approach (2,500 members) to account for uncertainties in solar wind speed and phase front orientation, avoiding the need for real-time CME identification.
• Geomagnetic Modeling: Applied the Temerin and Li (TL) model to the time-shifted STEREO-A data to generate 1-minute resolution SYM-H indices (a high-resolution proxy for the Dst index).
• Validation: Compared modeled SYM-H indices against observed OMNI high-resolution data to assess detection rates, timing accuracy, and intensity prediction for geomagnetic storms (defined as SYM-H < -50 nT).

3. Key Contributions

• First Statistical Study: This is the first study to statistically evaluate geomagnetic storm forecasting using in-situ sub-L1 data from a spacecraft crossing the Sun-Earth line.
• Orbital Constraints: Established empirical constraints for future mission design regarding radial distance and longitudinal separation.
• Baseline Methodology: Created an empirically motivated, real-time capable pipeline for shifting sub-L1 data and modeling geomagnetic responses without requiring complex real-time CME detection algorithms.
• Physical Insights: Quantified how CME properties evolve over short radial distances (0.01–0.05 AU) and how longitudinal separation affects arrival times and geoeffective parameters.

4. Key Results

A. Lead Time and Orbital Requirements

• Radial Distance Limit: 25% of the 32 CME events showed a negative lead time (detected at L1 before STEREO-A). The study concludes that future sub-L1 monitors must be placed closer than 0.95 AU to the Sun to guarantee a lead time advantage.
• Longitudinal Asymmetry: Lead time gains are highly dependent on longitude. Events east of the Sun-Earth line showed greater lead time gains than those to the west.
• Optimal Separation: To ensure both spacecraft measure the same solar wind structure, the longitudinal separation should be < 15° (ideally < 12° for smaller structures).

B. Physical Evolution of CMEs

• Magnetic Field Variations: Mean magnetic field strength ($B_{tot}$) changed by ~16% between STEREO-A and L1. Contrary to simple global expansion laws, variations were dominated by longitudinal effects rather than radial expansion for separations < 0.05 AU.
• $vB_z$ Timing: The timing of the minimum $vB_z$ (a key driver of geomagnetic storms) did not always align with the shock arrival. In some cases, the geoeffective minimum was observed at L1 before STEREO-A, even when the shock arrived at STEREO-A first.

C. Forecasting Performance

• Detection Rate: Using the STEREO-A-based pipeline, 26 out of 47 (55%) observed geomagnetic storms were correctly identified.
• Intensity Bias: The model successfully detected 82% of intense storms (SYM-H < -100 nT). However, the pipeline tended to overestimate storm intensity (58% of cases) and predict the minimum SYM-H later (72% of cases) than observed.
  • Cause: Overestimation is attributed to biases in the TL model (trained on 1995–2002 data) and the time-shifting approach.
• False Alarms: False alarms were primarily moderate storms; intense false alarms were rare.

D. Target Region Analysis

For the specific "target region" (simulating future HENON/SHIELD orbits: >0.05 AU radial separation, -8° to +2° longitude):

• 5 storms were correctly identified (hits), 3 were missed, and 1 false alarm occurred.
• The detection rate for intense events remained high, validating the concept for future missions.

5. Significance and Recommendations

• Mission Design: The study provides critical design parameters for upcoming ESA missions:
  • HENON (2026): Will be at ~0.90 AU. This study confirms this distance is sufficient for lead time gains, though it will only periodically align with Earth.
  • SHIELD (mid-2030s): A multi-spacecraft mission at ~0.86 AU. The study supports the strategy of deploying multiple spacecraft to ensure at least one is always within ±10° to ±15° of the Sun-Earth line.
• Operational Feasibility: The results demonstrate that sub-L1 data can be effectively used for real-time forecasting, provided that:
  1. The spacecraft is closer than 0.95 AU.
  2. Longitudinal separation is minimized (<15°).
  3. Ensemble methods are used to account for physical evolution and time-shift uncertainties.
• Future Work: The authors recommend using future DRO missions (like HENON) to investigate the east-west asymmetry in lead times, as STEREO-A's trajectory was not symmetric. Additionally, improved time-shifting techniques (e.g., accounting for inelastic collisions or automated CME detection) are needed to reduce timing biases.

In conclusion, the paper validates the scientific and operational viability of sub-L1 missions for extending space weather warning times, while highlighting the trade-offs between lead time, longitudinal alignment, and forecast accuracy.