The October annular eclipse is coming up, and the most important requirement will be clear skies. Here’s how to forecast the weather.

annular eclipse sequence May 2012
Imelda Joson and Edwin Aguirre took this beautiful sequence of an annular eclipse in May 2012.

With the October annular eclipse now less than two weeks away, it’s time to begin planning for the weather.

The climate statistics on my webpage reveal the probability of seeing sunny skies on October 14th, but you are now rolling the dice with Mother Nature as we move from climatology to reality. Past cloud probabilities are out of date. Planning must now rely on what we call deterministic forecasts — the output from computer models — at least for the time being.

There are a horde of computer model forecasts available online that you can use to look ahead, but if you try them all, you won’t find yourself any wiser than if you confine your look-ahead to just a few. Up to around October 9th, you should use the model outputs for mental “what-if” planning, but I wouldn’t begin canceling hotel reservations before then. After the 9th, the various models should begin to agree on at least the large-scale patterns of cloudiness and other weather elements. Those agreements will gradually coalesce as we approach the eclipse.

If you are an amateur astronomer or a storm chaser, you probably already have a favorite site to go for your weather advice. If not, a search or social media will turn up many model sources, so you might want to explore some of them to find one you like. There are quick links to four of them on my website. The major commercial weather providers that you see on TV also use computer outputs in their forecasts, but they stick to one model, either their own or a government one, and don’t give you a selection from which you can evaluate the reliability of eclipse-day predictions. By going to the internet, you are seeking a second (or even third) opinion before making a change in your plans.

Other eclipse travelers have recommended Windy to me for long-range eclipse planning and though it shows pretty much what any other internet site does, I like it because I can switch quickly between models and make very quick comparisons of individual cloud forecasts. Windy offers three long-range models: the ECMWF, the GFS, and ICON; these have an international flavor, as the models are European, American, and German in origin.

Meteologix is another popular eclipse traveler site that offers 32 models for contemplation, including the Canadian GDPS; it also allows a quick comparison of outputs. Generally speaking, the ECMWF is regarded as the most accurate of all of the models in common use and one that you should put on your list to evaluate.

The mainstream long-range models go out 10 to 15 days, so eclipse day is already in view. However, many sites limit the extended view to only the first 10 days or require a subscription to go farther. To get the full 15-day output, you can go to the College of DuPage site, a favorite of storm chasers.

Cloud cover forecast
The 11-day cloud-cover forecast for 10 a.m. CDT on eclipse day, produced by the American GFS model on October 3rd. Overcast areas are white, clear areas are black. Note that the path of the annular eclipse passes from Oregon through Texas in the U.S. Until about October 9th, numerical forecasts of the October 14th cloud pattern should be viewed with considerable skepticism.
College of DuPage / NOAA

As eclipse day approaches, a number of short-range models become available. These models typically have higher resolution than the long-range versions, but only make forecasts out to about 84 hours. Two of these, the American NAM and the Canadian RDPS, are readily available online. It’s probably best to avoid the highest-resolution short-range models such as the HRRR and the RAP, as these are designed to help you find your way around thunderstorms rather than find an eclipse.

Whatever model and provider you decide to use, be skeptical until models begin to agree — and until eclipse day is imminent. And on October 13th, be prepared to switch over to satellite photos for your planning — stay tuned for more information on that.


Image of Andrew James

Andrew James

October 4, 2023 at 6:45 pm

My experience is that predicting the weather during a solar eclipse is near impossible. There are other effects that don't apply to normal weather patterns. I think when the sun is obscured by the moon the drop in temperature has the biggest influence on how the cloud cover behaves. I have experienced firsthand eclipses go from 100% cloud cover about five minutes before totality, then perfectly clear skies, then when after the eclipse was over, five minutes later the cloud cover went back to 100%.
My other advice would be that observers need to also have the ability to be mobile. There is no point in standing waiting for some eclipse when you become aware that scattered cloud cover will exclude your view. It is easy to predict if this is going to happen because you can watch the speed and direction of the clouds passing over you in the sky, and then roughly judge if you'll be at the wrong place at the wrong time.
The last thing I would advise is that observers don't all gather closely together in the one spot. If the weather is bleak and everyone is clouded out, there is nothing worse than a bunch of amateur astronomers being all disappointed.
In the end, unfortunately, we are subject to the weather gods!

You must be logged in to post a comment.

Image of mesoscale


October 5, 2023 at 12:06 pm

I strongly recommend the Canadian model. It tends to be very accurate in my experience and has features specifically for astronomers.

This has been my go-to for star parties and astrophotography prep. One thing to keep in mind is that any cloud forecast, regardless of source, should be taken with a grain of salt if the date is 5 or more days in the future. Thankfully both the annular and total eclipses occur in arid regions which should increase the chances of success.Just be prepared to drive long distances and move quickly if the weather doesn't co-operate.

You must be logged in to post a comment.

Image of Ernie Ostuno

Ernie Ostuno

October 5, 2023 at 2:37 pm

Very good info here, coming from a meteorologist. I use climatology to plan well in advance, but usually pick a spot to stay the day before that is near a network of highways that will allow last minute long distance travel along the path. For instance, during the 21 August 2017 total eclipse I made reservations several months in advance to stay at Effingham, IL. This was north of the path, but was near access to east/west highways in case clouds were going to be a problem in Illinois.

One other weather resource I would recommend looking at are model forecast soundings:

Once you learn how to read the soundings (the red line and green lines far apart means a dry atmosphere and no clouds, the lines close together means the potential for clouds) then you can get an idea where the best chance for clear skies are in the vicinity of the path.

Clouds are difficult to predict, even in the short term, especially in mountainous areas where orographic effects can lead to "standing wave" clouds that can block part or all of the sky in one local area. With convective cumulus (typical summer thunderstorm clouds) the eclipse could actually help clear the sky by lessening solar heating. I have seen this happen during several eclipses including 2017.

You must be logged in to post a comment.

You must be logged in to post a comment.