This is a part of study described in the dedicated post.
I am going to perform data clean up and feature extraction for Solar wind model fitting. The major predictor of the solar wind is considered to be coronal holes characteristics (e.g. see this paper)
I’ve got two CSV data sets that contain quantitative features extracted from the Sun images with computer vision algorithms.
One file is “green” (193nm) spectrum portion originated features, another one is “red” (211nm) spectrum portion originated features
After carrying out the data clean up and transformation I got 9 possible solar wind predictors out of initial 52 data columns to be used for model fitting.
Six of them are paired (3 predictors extracted from red image and the same 3 predictors extracted from green image).
Another three solar wind predictors which seem consistent in time are acquired by combination of red and green image features.
Initially these characteristics are calculated for red and green images separately. Averaging between them should lower the uncertainty of resulting quantities. As these quantities are acquired independently from different information channels (two images) but actually they describe the same characteristics: the properties of coronal holes.
Two variants of coronal holes area evaluations and evaluation of coronal holes count.
The next step is cleaning up the solar wind observations.
Having solar wind observations there will be possible to proceed to model composition and fitting.