The Science
Solar energetic particle (SEP) events remain one of the least predictable components of space weather. CLEAR's science program targets the specific physical processes that make SEP prediction difficult — and the forecast products that operations still lack.
Multi-point observations reveal that SEPs can spread over more than 200° of longitude — at times filling the inner heliosphere from a single eruption. Identifying the acceleration and transport mechanisms that drive these wide-spread events is one of the central open problems in heliophysics.
SEP intensities and spectral shapes vary by many orders of magnitude between events. Disentangling the roles of seed populations, shock geometry and efficiency, and magnetic connectivity is a prerequisite for forecasting flux and spectrum — not just event probability.
Predicting the full onset–peak–decay profile requires coupling particle acceleration at CME-driven shocks with transport through a turbulent, structured heliosphere. The timing and duration of SEP events remain difficult to forecast with existing tools.
Fundamental physics
Each CLEAR tool is built to answer a specific physics question — and the answers feed directly back into better prediction.
How solar eruptions accelerate particles from thermal to relativistic energies across the heliosphere.
How interplanetary turbulence, shocks, and magnetic connectivity shape SEP propagation.
How suprathermal seed populations modulate the severity of the radiation hazard.
Architecture
The building blocks of the integrated framework — physics-based simulations, empirical models, and a machine-learning pipeline that share inputs and feed each other's forecasts.
Space Weather Modeling Framework
Seamlessly couples domain models from the solar corona to the low terrestrial atmosphere — sun-to-mud. The Geospace configuration runs in operations at NOAA/SWPC.
Alfvén Wave Solar-atmosphere Model
Cutting-edge model for the time-dependent background corona and heliosphere through which particles propagate. AWSoM-R achieves faster-than-real-time performance.
Eruptive Event Generator with Gibson-Low flux rope
First magnetically-driven solar-eruption model available for community use. Simulates CME evolution initialized with observed active-region magnetic fields.
Multiple Field Line Advection Model for Particle Acceleration
Calculates particle acceleration at shocks and propagation along a multitude of Lagrangian magnetic field lines.
Particle acceleration and transport model
Complements M-FLAMPA in modeling SEP acceleration at CME-driven shocks and propagation through the heliosphere.
Stream-Aligned MHD
Eliminates the "V-shaped" field lines that plague ordinary MHD near the heliospheric current sheet, giving correct magnetic connectivity for every SEP model.
SEP prediction derived from STEReo observations
Empirical models relating peak proton intensity and spectra (10–130 MeV) to CME speed and direction relative to the spacecraft's magnetic footpoint.
Relativistic Electron Alert System for Exploration
Uses promptly-arriving near-relativistic electrons at L1 as an early indicator of an energetic proton event — including backside events others cannot see.
Machine-learning SEP forecast pipeline
Conditional ML models trained on SDO/HMI, SOHO/MDI, GONG magnetograms, GOES flares and LASCO/SECCHI CME parameters — enabling pre-eruption probabilistic forecasts.
Benchmarking & Validation Framework
Spanning more than four decades of observations from GOES, SDO, SOHO, STEREO, ACE, Parker Solar Probe, and Solar Orbiter, the CLEAR benchmark establishes community-standard event definitions, cross-mission calibration, and reproducible validation metrics — the foundation for rigorous model development, transparent intercomparison, and a bridge between scientific research and operational forecasting. The operational benchmark is openly available through GitHub and NASA CCMC.
Community-standard event definitions and cross-mission calibration unify a scattered archive into a reproducible data foundation.
A transparent, metric-driven basis for intercomparing SEP forecast models across methods, research groups, and operational agencies.
Machine-learning-ready training and validation sets curated for SEP forecasting — spectra, time-intensity profiles, and event metadata.
Missions contributing
Continuous, overlapping records from six flagship missions — the basis for cross-instrument calibration and decades-long event coverage.
GOES
Operational proton-flux and solar X-ray monitoring
SDO
HMI magnetograms, AIA EUV imagery, and SDO/HMI SHARP active-region patches
SOHO
Coronagraph, in-situ particles at L1, and SOHO/MDI SMARP active-region patches
STEREO
Multi-point heliospheric imaging and proton-flux measurements
ACE
In-situ solar-wind and particle measurements
Parker Solar Probe
Closest-ever solar observations
Solar Orbiter
Remote + in-situ from novel vantage points
