Without looking at objective verification stats, I'd argue the worst subjective flaw in medium-to-S2S range modeling this year was the month-long stretch from early April to early May when there was always a classic Plains severe weather pattern in the offing at days 10-20. Of course, it never materialized in any form or fashion. (Yes, there was a stretch of chase opportunities from about May 15-25, but they came admist a pattern that was far from textbook and bore little resemblance to aggressive medium range predictions earlier in the season). Since the second week of May or so, NWP depictions of the big picture for the 1-2 week range have seemed to settle down and have performed reasonably well. In fact, nightly GEFS runs even nailed the nauseating, detestable June that's now all but locked in.It would be interesting to go back and evaluate long-range model accuracy for this odd season. Glancing through the posts in this thread, several of the long range predictions were in error. I'm still waiting for the "amazing" season Accu-wx predicted. Maybe they were talking about the monsoon? lol
From my post 30 days ago:
Bear any resemblance to the 500 mb analysis the past week, or the 500 mb forecast for any snapshot from any ensemble run you can pick over the next two?The 00z GEFS now integrates out to 840 hours (35 days, for those keeping count at home), and essentially every run asymptotes toward an upper high over Chihuahua and trough over Quebec as the primary NA features at 500 mb throughout weeks 3-5.
The fact that seasonal modeling from 6+ months out, along with the GEFS 4-5 weeks out, had such strong signals for this kind of outcome is something I can't get off my mind. With the usual disclaimer that I have no formal background in climate or even S2S modeling, it stands to reason there's something in the boundary conditions (meaning SSTs, land/soil state, polar sea ice, etc.) that made this predictable, even as certain teleconnections at times argued for classically active periods this spring. The burning question for all of us, of course, is how persistent the cause(s) may or may not be as we move years into the future.