This popped up on SEML today - good overview of multiple models:
Stephen
4 Likes
I was just going to share a link to that myself!
I have done a moderate dive into the models on that site and other forecasting sites, now that we are within 10 days. Overall it looks like largely cloudy weather from mid-Texas to mid-Indiana, with possible rain in northeast Texas to southwest (Arkansas (one of my locations). With areas of more open/broken clouds in western Ohio and New York. Mexico has some clouds but not thick. New England and Canada are back to a greater chance of clouds.
But this is so early that very little stock should be put into it. Only after we are a week away do the models start refining and coming together.
Dave
2 Likes
So I shouldn’t just kill myself now?
Seriously, any daily forecast updates you guys care to share will be appreciated. I will be in Mazatlan and watching for myself, but any insight into the numbers would be nice. Thanks!
1 Like
Here are the sites I am using. Please let me know if I should add any. Some good ones (like Astropheric) are not here because they do not cover Mexico.
You might want to include SpotWx, which has access to a number of models for any given location.
Dave
It’s there!! Added it when you mentioned it last week
I’m planning (as of now) to be in Texas, near Georgetown. The weather forecast is looking pretty ugly. That pivotalweather site predicts over 90% chance of cloud cover. We booked our trip (flights, hotels, car) so we can cancel by Thursday and get 100% refunds. So the big question: Will the prediction be solid enough by Thursday to make a go/no-go call? Frankly I would chance it if the cloud cover forecast was, say, less than 60%. But at 90%, I can’t see the point in going. Anyone else thinking along these lines?
Ken
That’s a tough one. Generally you don’t put much stock in the models outside of a week ahead, and they vary within themselves and among the others. Within 7 days is when they begin to coalesce and be more consistent. In this case the models even starting at 10 days have been maddeningly consistent in predicting significant clouds for most of Texas and SW Arkansas. It is hard to not be discouraged by that.
But I’d hold out and see what develops over the next few days. If there hasn’t been any significant evolution by Thursday you can probably be confident it won’t improve significantly. But if there is some indication of change, especially if the low to the NW starts changing position, then watch closely, hold your breath, and make your best decision on Thursday.
I’m doing pretty much the same, through I have some choices. I have campground reservations in SW Arkansas and I assumed I would probably be heading there, based on general climatic data. I also have family in NW Ohio, and surprisingly, that is looking better than AR at the moment. I’ll also be waiting until Thursday to make the call.
Dave
1 Like
Weather geeks might be interested in checking out this site - it’s particularly good for understanding cloud structure.
This will give you a plot of relative humidity, cloud cover (inferred from RH) and geo-potential height for a bunch of different pressure levels.
It’s showing 100% cloud in Mexico, but only very high cloud, which may still be kinda OK. I tend to pull the data down into Excel to inspect it further. I’ll try to post a chart later.
I’m going to be keeping a close eye on this in the next few days!
2 Likes
So now we’re 5 days out, and my Texas plan is looking like a huge bust. Various models showing 60-100% cloud cover. I have to decide by tomorrow midnight to get refunds on travel arrangements. Southern Illinois is looking very interesting now, and it’s within driving distance from my home north of Atlanta.
FWIW, I’ve canceled my arrangements for Texas. My new plan is to drive Cape Girardeau, MO, near the southern tip of Illinois. According to Pivotal Weather, the odds of cloud cover fairly low. However, the open-meteo.com site doesn’t agree. I’ll be checking again tomorrow and Saturday since I have more flexibility. Good luck, fellow eclipse chasers!
2 Likes
I’ve learned more how to read the Meteo data. You have to look at the clouds by layer instead of just the %cloud cover overall. Cape Girardeau is predicted to have some very high clouds 26,000 feet and higher, but no low- or mid-level clouds. Hopefully they are the thin, wispy kind that you can photograph right through.
Yes - the Open Meteo API is good for this - you can get cloud estimates at 19 different pressure levels.
I posted a slightly simplified view of this on our website yesterday: Article: What if it's cloudy?
The ECMWF model is looking considerably more positive than GFS still for northern Mexico.
1 Like
