Another NWS Tornado Forecast and Warning Failure

Kari King via KFOR TV

This is a two-section posting pertaining to the warning system issues in central Oklahoma this morning: The first part is for everyone and the second part is for meteorologists, storm chasers and emergency managers. 

 

For the Public

 

While your clocks were automatically resetting this morning, tornadoes were raking central Oklahoma. And, while – thank God – there were no fatalities, these tornadoes brought terrible damage. They also bring the National Weather Service’s (NWS) issues with tornado forecasting and tornado warning into stark relief. 

 

Friday, I apologized and wrote about Wednesday's tornado forecasts where no tornadoes occurred. 

This morning was the opposite, and more dangerous, situation: no forecast yet strong tornadoes occurred. Imagine being awakened by the sound of your home flying apart in the middle of night with no warning whatsoever. This was the circumstance of dozens of people in the Sooner State this morning. 

The Weather Channel

This event was not a more or less obvious spring supercell tornado situation and, perhaps, a little latitude can be allowed. But, at least one forecaster correctly discerned that this was a “high risk” situation! More on that shortly.

 

I wish I could inform you of some politicians who are on top of the NWS, FEMA, Red Cross and other disaster forecast and recovery agencies' issues so you could factor it into your voting decision. I can’t find any. If you know of some, please let me know in the comments and I will get into this and will publish their name(s) before election day. 

 

After a failure like this, the instinct of most in meteorology is to rally around the NWS when it has a problem. For example, this is what was posted earlier today on Twitter/X by a past chair and current member of the American Meteorological Society’s Board of Private Sector Meteorology:

I disagree that this NWS office is “understaffed” -- this was their Norman, OK office which is generally considered the gold standard of Tornado Alley offices. I have no opinion as to their pay, except that salary does not necessarily reflect competence. Regardless, the storm warning responsibility is a team effort. Pay and staffing have nothing to do with why these warnings were substandard. We have documented on this blog example, after example, after example of substandard warnings that have, too often, led to a loss of life.

 

It is long past time for the meteorological profession to stop its denial that the NWS has serious issues -- especially regarding tornado forecasts and warnings. We must create a National Disaster Review Board, modeled after the hugely successful National Transportation Safety Board which had made flying so incredibly safe. 

For Meteorologists – The Forecast

Let’s begin with the forecast.

In these outlooks, the odds refer to the chance of a tornado occurring within 25 miles of a given point. Red dots are tornado locations as of noon today. More dots will be added by the NWS.

 

The National Weather Service 3pm CDT Saturday forecast, above, the final tornado forecast of the day, called for almost no chance of tornadoes in Oklahoma and central Oklahoma in particular. Five percent (brown) is the significant threshold. The green is sort of “a tornado is not impossible” demarcation. No one stays up past midnight because of a 2% risk, especially if there is no tornado watch in effect when they retire. This was, obviously, a poor forecast. 

 

But take a look at this private sector forecast issued at 1pm Saturday.

It called for a “high” risk (on my four-point scale) of strong (!) tornadoes (hatching = strong).

 

NOAA/NWS does have a tool designed to highlight, with higher precision, areas where tornadoes are likely. That tool was not turned on last night, so there was no chance it could have given notice to human forecasters!

This is a screen capture of part of the warn on forecast tool. Note the most recent date of use at upper left: last Wednesday.

 

Nadocast, and other private sector forecasts, aren’t perfect. Nadocast, like me, over-forecast tornadoes on Wednesday (30th). But, there are a surprising number of situations where Nadocast is superior. 

While the NWS had one of its CYA “severe thunderstorm watch that also forecasts tornadoes” out, the official tornado watch didn’t come until 1:42am CST after the EF-3 Sooner Road Tornado occurred and just minutes before the ≥ EF-3 tornado that struck Harrah, Oklahoma.

For Meteorologists – The Lack of Warning 

 

This event was complicated to decipher due to some of the storms occurring during the Daylight Saving Time to standard time transition. For ease of reading, I’m listing the times of all of these events in standard time. 

 

“Sooner Road” Tornado

 

According to the NWS radar’s “tornado debris signature” (a sure indication the tornado is on the ground), the Sooner Road Tornado was on the ground no later than 12:18am. A tornado warning was issued at 12:21am (available at 1:23am). That means, at best, the tornado warning’s lead-time was minus 3 minutes (NOAA/NWS’s stated goal is +13  minutes). Below is the Oklahoma City Terminal Doppler Weather Radar at 1:17am CDT (12:17am CST).

Why is a tornado warning clearly indicated?

·      Hook echo

·      Strong rotation over a small area.

·      Individual supercells ahead of a line, strong “pattern recognition” indicator

·      Significant Tornado Parameter was +1 

 

Harrah Tornado

This tornado is rated by the National Weather Service as "at least" EF-3 intensity. 

You can see the flattened homes. KOCO-TV

In this case, the tornado warning should have been issued by 1:49pm.

The rotation was rapidly increasing at the circle. Harrah is near the top of the image to the north northeast of the developing tornado. Unfortunately, the tornado warning wasn't issued until 1:58am (available at 2am CST, below) by which time lofted debris are occurring 

Radar showing lofted debris, meaning the tornado was on the ground by 1:58 am CST.

The lead-time in this case was, at best, zero minutes against the published goal of 13 minutes. 

Summary

Over and over since the mid-2010’s, we have documented obvious tornado situations with inadequate warnings. We are extremely fortunate there were only 11 non-life-threatening injuries this morning. We cannot keep hoping our luck holds.