Forecasting the Ambient Solar Wind With a Physics-Informed Learned Inner Boundary: Part I-Differentiable Inverse Reconstruction From Solar Images
Abstract
The solar-wind speed at the inner boundary of heliospheric models is not observed; near-Earth monitors record the wind only after nonlinear propagation. We pose the reconstruction of an effective boundary speed at 21.5 R⊙ as an inverse problem and train a map from solar observations to that boundary by differentiating through a reduced heliospheric transport solver from 21.5R⊙ to measurement locations near the L1 Lagrangian point. Screening 42 candidate features from SDO/AIA 193~Å images and SDO/HMI magnetograms identifies coronal-hole area, its proximity to the ecliptic and its darkness as the most informative; the reference model adds their recent history, the expansion factor of GONG-based potential-field models and the speed observed one rotation earlier. In five-fold chronological cross-validation over solar cycle 24, the reconstruction outperforms 27-day recurrence and a linear regression on the screened features at a fixed lead (ROC area for wind above 600~km~s−1: 0.879 versus 0.802 and 0.850) with a predicted-to-observed standard-deviation ratio of 0.81, at the cost of a +26~km~s−1 bias from its tail-weighted loss. A factorial comparison shows that the skill comes from training through a speed-dependent propagation: a fixed lag degrades the score by 8.6~km~s−1, stream interaction adds little, and a linear map, or a model with one coefficient per physical quantity, matches the neural network. About 57\% of the variance of the hourly speed is not determined by these inputs, and the remaining error is common to all model classes: better reconstructions need more informative inputs rather than more flexible models. Forecasting and probabilistic prediction are addressed in Parts II and III.
Related
- Gaussian processes on ray-guided transformed uniform grids for fast, flexible, and auto-differentiable adaptive source reconstruction in lens modelling
- Cosmo-PINN: A Physics-Informed Neural Network for Cosmological Reconstruction
- Physics-Informed Neural Networks and Data-Driven Models for GRB X-ray Light-Curve Gap Reconstruction
- A differentiable and optimizable 3D model for interpretation of observed spectral data cubes
- A Differentiable Neural Surrogate for Photon Propagation in Neutrino Telescopes
BibTeX
@misc{camporeale2026forecasting,
title = {{Forecasting the Ambient Solar Wind With a Physics-Informed Learned Inner Boundary: Part I-Differentiable Inverse Reconstruction From Solar Images}},
author = {Enrico Camporeale and Andong Hu and Prateek Mayank and Thomas E. Berger},
howpublished = {arXiv},
year = {2026},
month = nov,
doi = {10.48550/arxiv.2611.01310},
eprint = {2611.01310},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2611.01310},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - Forecasting the Ambient Solar Wind With a Physics-Informed Learned Inner Boundary: Part I-Differentiable Inverse Reconstruction From Solar Images
AU - Enrico Camporeale
AU - Andong Hu
AU - Prateek Mayank
AU - Thomas E. Berger
PY - 2026
DA - 2026/11/01
PB - arXiv
DO - 10.48550/arxiv.2611.01310
UR - https://arxiv.org/abs/2611.01310
AB - The solar-wind speed at the inner boundary of heliospheric models is not observed; near-Earth monitors record the wind only after nonlinear propagation. We pose the reconstruction of an effective boundary speed at 21.5 R_⊙ as an inverse problem and train a map from solar observations to that boundary by differentiating through a reduced heliospheric transport solver from 21.5R_⊙ to measurement locations near the L1 Lagrangian point. Screening 42 candidate features from SDO/AIA 193~Å images and SDO/HMI magnetograms identifies coronal-hole area, its proximity to the ecliptic and its darkness as the most informative; the reference model adds their recent history, the expansion factor of GONG-based potential-field models and the speed observed one rotation earlier. In five-fold chronological cross-validation over solar cycle 24, the reconstruction outperforms 27-day recurrence and a linear regression on the screened features at a fixed lead (ROC area for wind above 600~km~s^−1: 0.879 versus 0.802 and 0.850) with a predicted-to-observed standard-deviation ratio of 0.81, at the cost of a +26~km~s^−1 bias from its tail-weighted loss. A factorial comparison shows that the skill comes from training through a speed-dependent propagation: a fixed lag degrades the score by 8.6~km~s^−1, stream interaction adds little, and a linear map, or a model with one coefficient per physical quantity, matches the neural network. About 57\% of the variance of the hourly speed is not determined by these inputs, and the remaining error is common to all model classes: better reconstructions need more informative inputs rather than more flexible models. Forecasting and probabilistic prediction are addressed in Parts II and III.
ER -