Where the cloud forecast for the solar eclipse shined and stumbled

Monday’s total solar eclipse put on an unforgettable show for sky gazers along a roughly 115-mile-wide swath from Texas to Maine. That is, provided clouds didn’t get in the way.

In the one to two weeks leading up to the eclipse, the cloud forecast loomed large. The predictions helped guide where many people decided to watch the eclipse. So how did they perform?

We provided daily forecast updates beginning March 29 on our cloud tracker page — by displaying an average of cloud simulations from a large group of computer models known as the National Blend of Models. (On the day before and day of the eclipse, we instead showed the cloud cover forecast from the National Weather Service — which relies heavily on this model blend but is tweaked by actual meteorologists.)

Overall, the forecasts presented did a reasonably good job simulating the cloud cover — even up to about 10 days in advance. They predicted sunnier conditions in New England and cloudier weather in Texas. However, there is definitely room for improvement as the predictions were overly pessimistic in some areas and did shift around a bit.

An animation of the progression of the forecasts reveals several things:

Advertisement

• Models consistently simulated clouds in areas near the path of totality because of a predicted front between the eastern Great Lakes and Texas. The location of the front and clouds wavered a bit between seven and 10 days out but then more or less locked in place.
• Deeper colors take over as the eclipse date approached, indicative of the fact that the simulations were coming into stronger agreement on weather features.

In addition to displaying the model forecasts on our tracker page, we analyzed how they were shifting and attempted to provide some insight on the implications of the overall pattern. As one example, our first updates noted the potential for thunderstorms after the eclipse near Dallas and the possibility of at least partial sunshine before the storms popped up. Indeed, storms developed there Monday evening, as did breaks in the clouds prior.

Where the forecast was best

Historically, northern New England had the lowest chance of clear skies in the early April along the eclipse’s path. However, models correctly simulated an unusually strong high pressure zone in the eastern United States would produce abundant sunshine in that region — and did so up to 10 days into the future.

On the edge of high clouds and the wall-to-wall sunshine, forecasts for Burlington, Vt., didn’t waver much and the only our very first forecast indicated more than a 50 percent chance of clouds.

Advertisement

Defying historical norms, forecasts in the central and southern Texas started off rather cloudy and generally stayed that way. The forecast for Dallas proved tricky despite not being too far off — as it was near the edge of where cloudier weather started and remained low confidence from start to finish.

The forecasts for low cloud cover around Indianapolis were also good. The forecast for mostly sunny skies was pretty much locked in five to six days in advance despite the possibility of high clouds.

Generally, the best forecasts were in northern New England, central Texas, Indiana and Ohio.

Where the forecast was worst

Places in the immediate vicinity of the front had the trickiest cloud cover forecasts, especially in Arkansas and the eastern Great Lakes.

Even though forecasts for Little Rock were initially about as cloudy as they were for Dallas, the city ended up being one of the sunniest places for the eclipse. It wasn’t until two days before the eclipse that it became apparent there was a strong chance of abundant sunshine.

Advertisement

Buffalo and Cleveland started off with sunny forecasts and only Cleveland ended up with clear skies, and its forecast was a moving target. The forecast for Buffalo took a cloudy turn after the initial forecasts for sunshine. The same was true for northwest Pennsylvania and portions of northeast Ohio.

While forecasts up to 10 days or so into the future provided a general sense of what would ultimately transpire, forecasts made further into the future mostly did not.

A notable example was AccuWeather’s 30-day forecast, which called for “poor” viewing conditions in northern New England and “good” viewing in southern Texas — the exact opposite of what transpired. However, its prediction for fair to good viewing toward the middle of the path turned out well.

Our own “way too early” look at the cloud cover forecast — two weeks before the eclipse — also wasn’t great as it presented an outlook for cloudier conditions in northern New England and sunnier conditions in Texas; however, there were models available at that time that were starting to sense what would actually happen. We wrote, “There are some that hint at higher pressure and sunnier skies for eastern or northeastern parts of the path of totality.”

Overall, sunshine won the day

As April is a notoriously cloudy time of year, the weather ended up about as good one can expect. A significant majority of people in the path of totality witnessed at least intermittent moments of celestial magic — western New York being a notable exception.

Especially in the middle of the eclipse path, the forecast trended sunnier with time as the front wasn’t a particularly strong one. Dallas caught a solid view of the eclipse, despite partial cloud cover.

ncG1vNJzZmivp6x7uK3SoaCnn6Sku7G70q1lnKedZMSmrdOhnKtnYmV%2FdXuPbWZqaV%2BasK21z6ycZpucpMKlecWoqZ6bkajBbr7Er6Cer18%3D