This page has daily 12Z model output discussion as it comes in. Discussion updated Monday-Friday by 1:15p.
HEADLINE: GFS VERY WET AT MID-DAY, BUT PATTERN SUSPECT BY LATE WEEK 1…
North America: The 12z GFS is notably wetter in the Midwest and Ohio Valley than the 00z or 06z runs, both due to unrealistic convective feedback issues in the first five days and a wetter pattern continuing across the Midwest in early week 2. This is due to a much weaker and further east ridge on the 12z run than previous runs. As this is much weaker with the ridging than basically all other modeling, this run’s QPF of 4″+ through day 10 across much of the corn belt is solidly an outlier.
South America: The 12z run is a little wetter in central Argentina and a little drier in southern Brazil, but there are no major changes overall.