A coherence-first framework for reading Earth systems
Resonance does not start with outputs and work backward. It starts with structure, geometry, and field configuration, then asks what those patterns make more likely. That is the platform worldview.
Measure the shape, not only the aftermath
Space weather has traditionally been measured by its outputs: how disturbed the field is, how high Kp reached, how deep Dst dropped. RFS-X measures something different: the geometric configuration of the field system before and during events.
The shape of how energy approaches, peaks, and departs carries information about what kind of event is occurring and where it is in its lifecycle. This is the difference between measuring how loud a sound is and understanding the shape of the waveform.
Different event types have characteristic shapes. Fast impulsive events rise sharply and recover quickly. Sustained events build gradually and decay slowly. Events embedded in longer activity periods have different structural signatures than isolated events.
Multiple layers, one coherence picture
RFS-X draws from multiple instrument families so the platform can see the field system across heliospheric forcing, orbital response, and ground response together.
The combination of these sources enables the detection of coherent pre-event patterns that individual instruments miss.
Solar wind speed and density, interplanetary magnetic field orientation (Bz), solar energetic particles, and X-ray flux from NOAA SWPC and NASA CDAWeb. These measure what the Sun is sending toward Earth.
Kp and Hp30 from GFZ Potsdam, and Dst from the Kyoto World Data Center. These measure how Earth's field is responding to heliospheric forcing.
GOES-16, GOES-18, and GOES-19 satellite magnetometers at geosynchronous orbit. These measure the field at 6.6 Earth radii, the orbital layer between the solar wind and the ground.
Five USGS observatories: Boulder CO, Fredericksburg VA, Honolulu HI, Sitka AK, and Tucson AZ. These measure the field at Earth's surface, the final layer in the cascade.
The NMDB composite from four stations. Cosmic ray suppression (Forbush decrease) follows CME arrival and is measurable hours before conventional storm indices peak.
A research-grade pattern detector
CPS monitors three simultaneous signals that historical analysis has found to precede confirmed geomagnetic storm events: field geometry elevation, orbital magnetometer deviation, and ground network dip/rebound.
When all three signals are simultaneously active, CPS registers a full pattern. Historical validation across confirmed storm events from 2018 to 2026 shows that this configuration preceded confirmed geomagnetic activity in the majority of examined cases.
CPS is a research instrument undergoing ongoing validation. It is not a forecast. It is an observation that a specific pattern is present or absent.
Every coherence event has a shape
The approach shape describes how the pre-event signals developed: V_SHARP events rise steeply in the final hours before onset. GRADUAL events build slowly across days. COMPLEX events show multiple sub-peaks or irregular approaches. PLATEAU events sustain an elevated state before a final acute phase.
The departure symmetry describes the relationship between the approach and recovery. When recovery is much slower than approach, something sustained the energy injection after the peak. When recovery matches the approach rate, the event was impulsive and self-contained.
These geometric characteristics cluster into families. Events in the same family tend to produce similar downstream outcomes in similar timeframes. This is the foundation of the pattern-recognition work that sits above raw detection.
Deliberate limits are part of the contract
RFS-X does not forecast specific values at specific times. It does not predict exact storm arrival times. It does not replace NOAA model output. It does not produce deterministic warnings or alerts.
What RFS-X does is measure the geometric configuration of the heliospheric and near-Earth field system and identify patterns that historical analysis associates with elevated space weather activity.
The constraint-based language used throughout this platform is intentional. The system is honest about what it knows and what it does not.
The public proof path stays visible
RFS-X detection outputs have been validated against 13 confirmed geomagnetic storm events from 2018 to 2026, using verified timestamps from NOAA SWPC records.
Across the 12 modern-era events within the instrument's coverage window, a coherence signal was detected at some tier in all 12 cases. A discrete precursor signal was detected before storm arrival in 8 of 12 events. The longest confirmed lead time was 87 hours before a G4 storm in October 2024.
This validation work is ongoing. Results are updated as new events occur and as the historical analysis deepens.