This code accompanies the paper Post-processing Probabilistic Forecasts of the Solar Wind by Data Mining Similar Scenarios, to be published in the AGU Journal Space Weather in 2026. This code generates probabilistic forecasts of the solar wind speed using a 11-year historical record of ADAPT-WSA forecasts and ACE observations at L1.
The methodology used here builds on k-NN methods, analog ensembles, and the skew normal distribution.
The solar wind predictions from WSA for use with this code, as well as the processed data made using this code, can be downloaded from Zenodo. The full WSA runs (including FITS output) are available on HuggingFace.
The author can be reached at daniel.e.dasilva@nasa.gov or mail@danieldasilva.org.