This repository contains data of solar activity forecast.
Data are obtained using a method based on a combination of the numerical solution of the nonlinear mean-field dynamo equations and the artificial neural network.
The detail can be found in the paper (https://doi.org/10.1017/S0022377818000600, PDF-file in the doc/ folder).
We update our forecast of 13-months running averaged sunspot numbers at the beginning of each month.
The files predict_YY_MM.csvappear in the data/ folder.
Format: Comma Separated values
Contents:
- Column 1-2: Gregorian calendar Year-Month
- Column 3: predicted sunspot numbers
| Date period | Reference |
|---|---|
| 1997/01—2017/10 | tuning methodology |
| 2017/11—2021/09 | monthly forecast for testing |
| 2021/10— | monthly forecast during the project |
We compare our forecast with observational data provided by https://www.sidc.be/silso/datafiles
(SN_ms_tot_V2.0.csv is available in the doc/ folder).
- Actual forecast
- Latest monthly evaluated predictions and observations
R
- Comparison with the other methods of forecast
- Forecast data in table form. Each column is for the date of predicted 13-months running average sunspot number (-6 months from the date of the given forecast). Each raw is for the date when the solar activity was predicted
- All forecasts curves
-
Residual between the predicted and the observed sunspots number for different forecast depths (
m- months ahead)
-
Standart deviation of the residual between the predicted and the observed sunspots number for different forecast depths (
m- months ahead)
The project is supported by the Russian Science Foundation (grant No.21-72-20067).