Looking at Forecasts: A deeper dive

The purpose of this post is to help you understand how to look at weather forecasts: specifically, those dealing with cloud coverage.

One important thing to start with is that "Weatherman Willie on Channel 5" and the weather app on your phone should not be your only resources. They can be resources to use; just not the only ones.

The key thing to note is that there are multiple weather models, and when we are 5 days or more away from the day of interest, these models can be in significant disagreement.

Once you are within 2-3 days, you may start to see some general agreement between models (although you will certainly still see minor to moderate differences.)

Of course, realize that it's all just forecasting.

I'll direct your attention to a site that I use: https://weather.us/ 

Here's how to get to the cloud cover forecast. Click on "Forecast"- a menu will drop down. Then click on GFS. This is one of the weather models.

Under "Parameter Selection"- click "All". Then click "Clouds, Sunshine, Short Wave Radiation"

You can click on "cloud coverage". That shows the percentage of the sky predicted to be covered by clouds at the time. It defaults to the next hour, but you can change the date to show the cloud coverage prediction for the time of the eclipse (or as close to it as possible)

(For Texas, totality is between 1 and 2 p.m. local time on 4/8) 

There is a magnifying glass icon at the top of the map; click in to zoom in on your state. You can also change that by selecting "change map selection".

You'll also notice that there are blue rectangles at the top with white letters such as "ECMWF" "NAM-C" and so on. These are the different weather models. Please note that, this early on, some of them may not yet appear. The NAM-C and HRRR models are high resolution models that don't appear until closer to the eclipse. You can also switch between the weather models by selecting "Change model run" in the menu.

You'll also notice that there are options for low clouds, middle clouds, and high clouds. Play around with these. Note that some weather models do not give this option. 

Generally speaking: low clouds can be thick, and will likely obscure the eclipse. There are 2 silver linings, though: 1. If the low clouds are of the "popcorn" cumulus variety (ragged, small), some of them may dissipate partially or completely during the temperature drop leading into totality. 2. There can be days with lots of low level cumulus clouds that have plenty of breaks. There is hope that one of them may be in front of the sun during at least part of the totality. Hopefully there are no stratus clouds: these can be rather widespread.

High clouds do not respond to the cooling associated with the eclipse. They can be tricky: if these cirrus clouds are very wispy, they may not significantly interfere with the corona. Perhaps the outer streamers may not stretch as far, and photography may be affected, but the visual experience is still excellent. It's when the cirrus becomes thicker that the view becomes more affected. Unfortunately, it's hard to use weather models to predict how thick the high cirrus clouds will be. This is better seen on satellite photos and loops, as well as visually.

As for mid level clouds, they will not respond to the temperature drop, and they can be thick enough to obscure the eclipse, so it's best to move from them (if you can) or try to find a clear patch.

An important note about sky coverage. Note that you can use the scale at the bottom of the map. Also remember that you can see the predicted percentage of coverage by high level, mid level, and low level clouds.

Let's use 50% as an example. Well, 50% cloud coverage could mean several things. It could mean that 1/2 the sky is predicted to be clear, and 1/2 the sky completely cloudy (a distinct line), or it could mean that there are clouds all over the sky in all directions and overhead, but there are enough blue patches that 50% of the sky is covered by clouds. Or it could mean something in between.

How can you tell? Look at the map. If in one direction the predicted cloud coverage percentage drops off very quickly going in one direction, and increases very quickly going in another, then this would indicate a sharp clearing. If the predicted cloud coverage percentage is 40-60% within several hundred miles, then that would indicate scattered clouds in all directions with blue patches in between. 

Don't forget to select the high, middle and low cloud options and compare the models! Also, pay attention to the direction of the clouds. If a prediction calls for 40% sky coverage, that sounds pretty good, but if that 40% is in the direction of the sun at totality time, that's not a good thing!

Note that high level clouds can be seen from farther away: they may be up at 18-20,000 feet.

The key thing is this: if you look at the various models as of now (April 2 as I write this), you will see that they are all over the place. Once we get closer, look for consistencies between the models. Also, don't focus on the small details. If your home or the area that you are wanting to watch the eclipse is in an area that is predicted to have 40% cloud coverage, and another area fairly close by is predicted to have 65%, then your location may end up having the 65% coverage, and the other area may have the 40%.

My next post will show you how to look at satellite photos and loops on the day of the eclipse, and how to determine the types of clouds that may be coming in the direction of your observing site.

Final note: Some people may not like the format of www.weather.us . If you want to use a different site to compare weather models, see meteorologist (and eclipse chaser) Jay Anderson's site at 

https://eclipsophile.com/eclipse-day-weather/ - scroll to "Numerical models". The websites listed there will allow you to compare cloud coverage predictions from the various models. 

I hope you have learned something!