Interesting. It’s not clear from the article debunking the link what the statistical significance they found between nao and solar, and the research paper they linked to is paywalled. The other link that tries to find a link for February doesn’t really make a convincing case for it, imho.
I’ll see if I can derive the p-value between those datasets broken down by month, just to satisfy my curiosity. If its well over 0.05 for all DJF months, that’s productive because I know I wouldn’t waste any more time on solar when predicting whether there will be blocking in the nao domain or not.