zzh wrote:CFS is a large scale climate model. It's not good for individual TCs at all.
I've been using it since 2014 and it's been fairly reliable as far as giving more detail about the level of activity during given time periods and a rough, blurry number of when, where, and what the levels of activity will be (though it has on occasion with stronger storms been accurate a mouth or so out, but obviously no way to know when it is without hindsight).
It's an additional tool that gives more insight beyond looking at a map with general anomalies--which when used on their own tell us far less than we'd like to believe.
I mentioned the hiccup run earlier mostly because it was somewhat amusing that it shows the entire season with nothing but one slow-moving Andrew-type track.