Eclipse Weather: Too Early For A Forecast

August 8, 2017


9am Tuesday

Have you seen a weather “forecast” or two for the eclipse?  There’s NO ONE that knows what it’s going to do on eclipse day yet.  If you’ve seen one of these, keep in mind it’s just a grab for clicks (online clickbait) or viewers (TV eyeballs) and has no basis in meteorology.

We won’t have a general idea of possible rain that day until around 7-9 days out.  As for cloud cover, which is far tougher to predict, we’ll start taking a stab at it this coming Sunday which will just be a general guess, and then hone in on a detailed forecast during the week leading up to the event itself on Monday the 21st.  Trust me, forecasting low marine clouds west of the Cascades is something we can easily get wrong just 3 days ahead of time.

As mentioned in the previous post, we DO have models that go out two full weeks like the GFS and ECMWF often referenced here.  But the “operational” versions of these models can flip around wildly with each run that comes in.  For example the 0z (evening run) GFS model looked like this on eclipse day:


That’s a cold upper-level trough just like we see in May or June.  That’s a showery and cool setup, but even in this pattern you can get large breaks in the cloud cover.  But wait!  6 hours later the overnight run of the GFS came in like this:

gfs_namer_324_500_vort_ht (1)

Much better, patchy morning clouds MAYBE west of the Cascades, otherwise all sunny and a very settled summer pattern with highs at least in the 80s.  In this model the trough is way up in SE Alaska.  You see the problem and why it would be ridiculous to make a forecast more than a week in advance?

That said, just as I mentioned in the last post, at least part of NEXT WEEK DOES FEATURE  A RARE COOL AND SHOWERY WEATHER PATTERN FOR MID AUGUST.  36 of 51 ECMWF model ensemble members (over 50%) project at least a tenth of an inch of rain over the metro area between now and eclipse day.  Some give us a half-inch or more.


Yet once again there is no need to panic…for now.  Take a look at 24hr rain totals from the same ensemble members:


You see a good percentage of the “rain action” happens Monday-Thursday next week.  There is a secondary smaller spike around the 21st/22nd (eclipse day), but only 7 of 51 ensemble members show more than .05″.  And this says very little about cloud cover, which of course is all we care about at 10:18am Monday the 21st.  Again, we won’t be able to start working on that one until about a week out.

To wrap it up, there is no need to panic/worry at this point.  We’ve got you covered here at FOX12; ignore any other “forecasts” you see in the next 5-6 days.  I’m right with all the rest of you watching the cloud forecast VERY carefully starting this coming Sunday/Monday!

Chief Meteorologist Mark Nelsen