The Increase in Storm Forecast Accuracy

One of at least 20 similar comments. They were
part of a thread that began a week ago, today. 

You might have seen the controversy that began a week ago today on Twitter and other social media pertaining to weather forecast accuracy.

I understand that some forecasts bust. And, if you have an outdoor event and the forecast is wrong, it can be extremely frustrating. I get it.

But, the idea that weather forecasts are usually wrong is incorrect. And, the progress weather science has made in warning of storms is stunning.

Because of that on-line discussion, I thought the forecasts of flash flooding and the violent, long-track tornado of this past Sunday in the South would make a good example of the progress weather science has made (hat tip, Greg Carbin).

Rainfall Amounts
Below are NWS forecast rainfall amounts from 7am Sunday through 7am Monday. The forecast was published at 3:41am Sunday morning. Because of the already wet ground, the forecast rainfall, up to 4" in Alabama, was forecast to cause major flash flooding.

Here is a map of the actual rainfall amounts.

The correlation between the two is amazing. And, yes, serious flash flooding occurred.

Violent Tornado in Southern Mississippi
A second example: There was a violent, long-track (52 miles) tornado Sunday evening in southern Mississippi. It was more than a mile wide in places and it was rated EF-4 intensity. It passed just south of the city limits of Hattiesburg.

Note the tornado first touched down at 7:10pm.

At noon, I tweeted this forecast of the most likely area for strong tornadoes.

Updates for this geographic area were sent later in the day. The violent Mississippi tornado occurred entirely within the rectangle.

At 5:40, I sent my first tweet calling attention to this storm (90 minutes before the violent tornado touched down). There is no "hook echo" or other typical signatures of a tornado, let alone a violent storm in progress. But, because of the progress we have made in weather science, I was able to outline the area of concern as well as to say the atmosphere was in the immediate area was "favorable for tornado development. If you live in this area, PLEASE monitor for possible warnings."

This was at 6:23. The rotation was strengthening and moving east northeast.

After multiple updates explaining the likelihood of a tornado developing, here is a specific tweet warning of the tornado three minutes before it touched down. Given the advance notices to prepare, three minutes was plenty of time to dash into the closet, basement, or bathroom.

Five minutes after the tornado touched down, thanks to the NWS's investment in dual-polarization radar, I was able to inform readers the tornado was both on the ground and violent even though the sky was dark (per photographs, not shown).

There were detailed path forecasts.

Weather science would not have been able to do all of this even a decade ago.

I'm using my forecasts and warnings not to boast but because they are easy for me to obtain. The NWS did a fine job warning of this tornado, also. Even though an EF-4 is in the most violent 2% of all tornadoes, there was just a single fatality. Amazing.

The next time someone claims the forecasts are wrong for "three weeks straight" or there were three days of sunny and warm when it was supposed to be cold and rainy, please send them a link to this essay.

(c) Copyright 2020, Mike Smith Enterprises, LLC