spsw/EL-PASO-PAD-Model-Data

EL-PASO PAD Model Data

Coefficient files of the empirical electron pitch angle distribution (PAD) model of Smirnov et al. (2022), mirrored from GFZ Data Services for use by EL-PASO. The mirror exists so that EL-PASO can download the coefficients reliably, independent of the availability of the original host.

The files are an unmodified copy of the dataset published at https://doi.org/10.5880/GFZ.2.7.2022.001, restricted to the PAD_model folder plus the accompanying documentation.

Model

The model describes normalized electron PADs in the outer radiation belt, derived from Van Allen Probes data, as a sum of odd sine harmonics:

g(alpha) = A1 * sin(alpha) + A3 * sin(3 * alpha) + A5 * sin(5 * alpha)

Each coefficient A1, A3 and A5 is a 3D polynomial in cos(MLT), sin(MLT) and L, whose terms depend quadratically on the solar wind dynamic pressure Pdyn. See equation (4) of the paper and the model description PDF for details.

• Valid range: 3.05 < L < 5.95.
• Energy channels (keV): 37, 56, 78, 106, 142, 179, 220, 246, 342, 453, 588, 735, 871, 1080, 1650.
• For the 37 keV channel, the model is better used in the outer zone, because in the inner zone the shapes are steeper and the fits may not be ideal.

Files

File	Description
Pijk_<E>_keV.dat	Coefficients for energy channel <E> and Pdyn below 5.5 nPa (polynomials of degree 7)
Pijk_<E>_keV_HIGH.dat	Coefficients for energy channel <E> during high activity, Pdyn above 5.5 nPa (polynomials of degree 5)
readme.txt	Description of the coefficient file format (from the original PAD_model folder)
read_coefs.py	Reference implementation for reading and evaluating the coefficients
2022-001_Smirnov-et-al_readme.txt	Top-level readme of the original GFZ dataset (license and citation)
2022-001_Smirnov-et-al_Model-Description.pdf	Model description
2022-001_Smirnov-et-al_List-of-files.pdf	List of files of the original GFZ dataset

File format

Each Pijk_*.dat file is a plain text table with 12 columns:

Columns	Content
1, 2, 3	Powers (i, j, k) of the polynomial term cos(MLT)^i * sin(MLT)^j * L^k
4, 5, 6	c0, c1, c2 for the A1 coefficient
7, 8, 9	c0, c1, c2 for the A3 coefficient
10, 11, 12	c0, c1, c2 for the A5 coefficient

For each coefficient, the value is c0 * Pdyn^2 + c1 * Pdyn + c2, summed over the polynomial terms in the rows. For Pdyn above 5.5 nPa, use the _HIGH files, in which the coefficients no longer depend on Pdyn.

Usage

from huggingface_hub import hf_hub_download
import numpy as np

path = hf_hub_download(
    repo_id="spsw/EL-PASO-PAD-Model-Data",
    filename="Pijk_246_keV.dat",
    repo_type="dataset",
)
coefs = np.loadtxt(path)

In EL-PASO, the model is available through el_paso.processing.models.get_pa_distribution_smirnov_et_al_2022 and the files are downloaded automatically on first use.

License

The data is licensed under CC BY 4.0, as in the original dataset. Please cite both the data and the paper when using it.

Citation

Data:

Smirnov, Artem; Shprits, Yuri; Allison, Hayley; Aseev, Nikita; Drozdov, Alexander; Kollmann, Peter; Wang, Dedong; Saikin, Anthony (2022): An empirical model of electron pitch angle distributions in the outer radiation belt based on Van Allen Probes data. V. 1.1. GFZ Data Services. https://doi.org/10.5880/GFZ.2.7.2022.001

Paper:

Smirnov et al. (2022), Space Weather. https://doi.org/10.1029/2022SW003053

For questions about the model, the original data contact is Artem Smirnov (arsmirnov95@gmail.com).