I think that’s what you would expect with a long range ensemble mean because the ENSO state is the only thing that has predictive value at that time range. The individual members are just a random mess of spaghetti. I wonder if they would even have positive skill scores if the ENSO signal is removed.

The 12z AI looks really promising. I think we could be in business by Jan 20 or even a little bit sooner. No doubt a big pattern change is coming though.

That could be, but we really haven't been in a Nina this winter until about now. Mostly it's just neutral.

Perhaps with long range, as in months out forecasts, but I've never seen evidence in cases like this. The fact is that though we saw some improvements in the mid-January period a week or so ago, we never saw the models latching onto the kind of consistent signal we are seeing now. This is the kind of signal we saw in 2023 and 2019.

I know others have mentioned this but i'm pretty convinced that any long range forecast is just pulling historical trends associated with ENSO and not really factoring in any meaningful present condition or trend.