Solar Orbiter Extends Geomagnetic Storm Warnings by Hours

CME-driven geomagnetic storms are most damaging when their southward magnetic field arrives with little warning. Monitors at L1 provide direct solar-wind measurements only shortly before Earth impact, leaving little time for operational response. The authors test whether Solar Orbiter, positioned well upstream of Earth during two March 2024 CMEs, can turn its real-time magnetometer stream into an earlier prediction of arrival time, magnetic structure, and the 1-minute Dst index.

Core Contribution

The paper’s contribution is an end-to-end, real-time procedure rather than a new physical model. It combines remote CME observations, an ensemble propagation forecast, Solar Orbiter in-situ magnetic measurements, and a geomagnetic-index model. The crucial step is to use a spacecraft that samples the CME 0.53 or 0.60 au before L1 as an upstream monitor, then propagate and scale its magnetic profile to Earth.

This is a demanding test of an appealing operational premise: a monitor need not sit exactly on the Sun–Earth line to provide useful warning. Solar Orbiter was sufficiently aligned for the two events, including one case with a 10°-scale longitudinal offset, but the evidence is confined to those favorable geometries.

Technical Approach

Figure 1 lays out the pipeline. Coronagraph and source-region information provide CME kinematics to the ELEvo drag-based ensemble model. The initial model predicts arrival at Solar Orbiter and L1; once the shock is identified at Solar Orbiter, that observed arrival constrains a second ELEvo ensemble for the L1 forecast. This update separates a conventional remote-sensing estimate from one informed by an actual upstream crossing.

For the magnetic forecast, the authors transform Solar Orbiter MAG observations into GSM coordinates, time-shift the profile to the predicted L1 interval, and apply radial scaling. For the first event, Figure 4 shows the predicted L1 field alongside real-time NOAA solar-wind data, with the modeled shock, leading edge, and trailing edge carried through the magnetic and Dst panels. The scaling is deliberately simple: the central prediction uses a magnetic-field exponent of -1.64, while -2 and -1.2 define an uncertainty range. The resulting field, together with prescribed plasma profiles, drives the Temerin & Li model to produce a 1-minute Dst forecast.

That design makes the paper practically useful but also exposes its assumptions. It treats radial evolution as more important than longitudinal evolution over the tested separations, and it has to approximate plasma inputs rather than measure the full upstream solar-wind state continuously.

Results and Analysis

For the 17 March event, the procedure produced its geomagnetic-index prediction 15.3 hours before the CME shock reached L1 and 33.9 hours before peak storm time. For the 23 March event, the corresponding lead times were 4.3 and 10.3 hours. Those are meaningful extensions beyond L1 nowcasting, especially for the first event, because the forecast contains an evolving field profile rather than only an arrival alert.

The upstream observation also improved the ELEvo arrival forecast in both cases relative to the prediction based only on DONKI kinematics, although several-hour arrival-time errors remained. The paper therefore supports a narrower conclusion than “upstream data solve CME arrival prediction”: a simple model can be materially corrected after a CME is sampled upstream, while residual propagation uncertainty still limits exact timing.

Figure 8 compares the inputs and Dst traces from the Solar-Orbiter procedure with L1 real-time solar-wind data and Wind science data across both storm intervals. It is an important diagnostic because it separates errors inherited from the magnetic and plasma inputs from errors in the geomagnetic response model. The predicted profiles reproduce enough of the event structure to yield realistic storm evolution, but the comparison also makes clear that density, speed, and magnetic-field discrepancies can move or alter the modeled Dst response.

The most persuasive aspect is that the system was exercised in real time on two actual storms, not reconstructed solely from final science products. Still, it is a two-event demonstration using unusually suitable Solar Orbiter geometry. Its operational case is strongest as validation for future continuous upstream monitors carrying both field and plasma instruments, rather than as proof that the same scaling will transfer unchanged across CME classes or strongly disturbed ambient wind.