Yeah, you know, I find it very interesting to compare weather modeling with climate modeling, because weather modeling really is amazing. We had a February blizzard several years ago that was predicted about 4 days in advance, before the storm even existed! It was predicted to form over Texas somewhere, swing up the Mississippi valley, and land right on Chicago, which it did. 2 feet of snow.
It seems to me that climate is average weather, but I don’t think you can scientifically model average conditions. I’ve wondered what happens if you just run a weather model for weeks or months. Does it flatten out, or will it keep going with some sort of reasonable looking weather, which is simply increasingly inaccurate? I suspect the former, but I don’t know if this question has been pursued.
I would note, though, that on general considerations we might readily conclude that the extension of weather modeling accuracy is logarithmic in the detail and precision of the compuation. That is, maybe it has to be 100 times as precise to get an extra day, so that getting a month of predictions is not even on the horizon.
I suppose there’s an answer to this, but I think it’s a thought provoking contrast.