NF2 solar extrapolation MOC
NF2 takes vector magnetogram boundary data and trains a neural field for coronal magnetic-field extrapolation.
Use this note as the index for the NF2 cluster.
Code and data path
- NF2 codebase pipeline
- JSOC DRMS segments
- SHARP data product
- HARP active-region patch ID
- CEA projection
- HMI SHARP CEA vector components
- HMI SHARP field inclination azimuth disambig
- NF2 FITS loader component conventions
- NF2 segment choice for evaluations
- SST solar telescope data
- CEA vectors vs polar magnetic-field representation
Neural model and training
- NF2 and physics-informed neural coronal extrapolation
- SIREN neural field model in NF2
- NF2 training loop and loss scheduling
- Potential-field loss scaling in NF2
- Residuals in NF2 and PINNs
Physics and modelling
- Coronal magnetic-field extrapolation as an inverse problem
- Potential field extrapolation
- Force-free and NLFFF coronal fields
- From MHD to the force-free equations
- Photosphere versus corona force-free assumption
- Preprocessing vector magnetograms for NLFFF
- Solenoidal magnetic fields and no magnetic monopoles
- Magnetic energy and free energy in active regions
- Magnetic topology and field-line tracing
Classical methods and evaluation
- Classical NLFFF extrapolation methods
- Classical optimisation and the NF2 loss
- NLFFF quality metrics
- NF2 diagnostics for physical credibility
- NOAA AR 11158 as an NF2 benchmark
- Low and Lou analytical NLFFF solution
Practical default
For normal HMI active-region NF2 work, use:
SHARP CEA + HARP number + Br/Bp/Bt(+ errors)For AR 11158, that means SHARP data product, CEA projection, HARP 377, and HMI SHARP CEA vector components.