I get a lot of questions about the weather on April 8, the day of the total solar eclipse in the eastern half of the United States.
Our operational forecasting models achieve that period and only last a few days. But given the uncertainty in such a long prediction, I have refrained from commenting until now.
Model solutions are starting to stabilize and I am increasingly confident that I can provide useful information. So let me do the first of these event-related weather updates.
The total acreage would extend from Texas to Ohio to northern Maine (see below).
The total eclipse will occur around 1800 UTC (11 a.m. PDT) most of the way (early southwest, later northeast). In Seattle, only about 20% of the sun is covered (see below). You will hardly notice this, even when the sky is clear.
Forecast situation
Below is the forecast relative humidity of 700-300 hPa (10,000-30,000 ft), a good measure of cloud cover, at 1800 UTC April 8 based on the US GFS model.
Parts of southwest Texas may have good visibility, but I’m concerned about visibility conditions in much of the Midwest. Some openings in Ohio and northern Maine.
But at this predicted time (270 hours), There remains considerable uncertainty. Only if we receive within about 5 days (120 hours) will the confidence in the forecast be relatively high.
To illustrate the uncertainty, let me show you a series of forecasts for the humidity field above for four different forecast forecasts, but verified at the same time (1800 UTC April 8). Specifically, you are viewing 288, 282, 276, and 270-hour forecast animations of the sea level pressure (solid line) and humidity field.
They all have some kind of low pressure system and a cloud over the middle of the country and it looks like the solutions are converging. But there are significant variations between forecasts.
For Seattle, the difference is quite large.
.gif)
An animation of the valid rainfall forecast for the same eclipse period is shown below. There was enough rain to worry eclipse watchers in the Midwest. And not promising for western Washington.
Another approach to such long-term forecasting is to consider the forecast of multiple forecasts.
Below is the North American Ensemble Forecast System (NAEFS) prediction of clouds over Dallas/Fort Worth Texas over time. I put a red arrow at the time of the eclipse. Remember: aggregate forecasts show the results of many different forecasts, all of which are valid at a given time.
10 represents total cloud cover. The horizontal line shows the median cloud cover for that location, and the yellow box around the medium shows the 50% coverage of the composite forecast components. (remember that the median is in the middle of the distribution of multiple forecasts, above as well as below).
A lot of clouds and a lot of uncertainty in that position.
Either way, if you’re going to see the eclipse, don’t expect perfect conditions and be prepared to move, depending on closer forecasts.
