This repository is part of the SatelliteToolbox.jl ecosystem.

SatelliteToolboxSpaceIndexSets

This repository contains scripts to generate space index sets for the SatelliteToolbox.jl ecosystem. The generated files are stored in the files directory and are updated automatically every day by a GitHub action.

Generated Files

File	Description
f107_observed_prediction_coefficients.csv	Coefficients to predict the observed F10.7 solar flux index
f107_adjusted_prediction_coefficients.csv	Coefficients to predict the adjusted F10.7 solar flux index

F10.7 Prediction Coefficients

The script fit_f107.jl generates the coefficients to predict the F10.7 solar flux index for satellite decay analysis. It fits either the observed or the adjusted data obtained from Celestrak, from 1957-10-02 up to the current day, considering the following harmonic model:

$$\bar{F}_{10.7} = F_0 + \sum_{i = 1}^{6} a_i \sin\left(2\pi i \frac{t - t_0}{P}\right) + b_i \cos\left(2\pi i \frac{t - t_0}{P}\right),$$

where $t_0$ is the reference day (1957-10-02) represented in Julian day. The fitting coefficients are:

• $F_0$: the mean value of the F10.7 index [sfu];
• $a_i$: the sine coefficients for the i-th harmonic [sfu];
• $b_i$: the cosine coefficients for the i-th harmonic [sfu]; and
• $P$: the period of the solar cycle [days].

This model was based on the information available in [1].

Output Format

The output is stored in the CSV files files/f107_observed_prediction_coefficients.csv (observed index) and files/f107_adjusted_prediction_coefficients.csv (adjusted index) with the following columns:

t₀, F₀, P, a₁, a₂, a₃, a₄, a₅, a₆, b₁, b₂, b₃, b₄, b₅, b₆

where t₀ is expressed in Julian day, F₀ and the harmonic coefficients in sfu, and P in days.

Running Locally

The script requires Julia and its project environment can be instantiated with:

julia --project=scripts/f107_prediction -e 'using Pkg; Pkg.instantiate()'

Afterward, run the fitting with:

julia --project=scripts/f107_prediction scripts/f107_prediction/fit_f107.jl

The output file path can be passed as the first command-line argument and the F10.7 index set (observed or adjusted) as the second one. If omitted, the script fits the observed index and writes the file next to itself:

julia --project=scripts/f107_prediction scripts/f107_prediction/fit_f107.jl files/f107_observed_prediction_coefficients.csv observed
julia --project=scripts/f107_prediction scripts/f107_prediction/fit_f107.jl files/f107_adjusted_prediction_coefficients.csv adjusted

The script prints a report of each step, including the convergence status, the fitted parameters, and the root-mean-square error of the residuals.

Automation

The GitHub action f107_prediction.yml runs the fitting script every day at 21:00 UTC for both the observed and the adjusted indices, writes the outputs to the files directory, and commits the result if the coefficients changed. It can also be triggered manually from the Actions tab using the workflow_dispatch event.

References

• [1] D. Whitlock (2006). Modeling the Effect of High Solar Activity on the Orbital Debris Environment. Orbital Debris Quarterly News, vol. 10, n. 2, April, 2006.