Replies

As a long time warning forecaster (over 30 years now), I have experienced the dramatic change in NWS capability as technology improved. I have taught many seminars on warning philosophy, based on both my experiences and those of other warning forecasters. I study warning failures in great detail. Much of the problem with warning verification comes from lack of detection, or ground truth. I always advise warning forecasters to separate reports from events. In other words, just because no one saw the tornado doesn't mean in didn't occur. I have personally worked events where spotters insisted there was no tornado, when in fact it was right in front of them in broad daylight with good visibility. How can that happen? I didn't fit the perceived profile. It didn't “look” like a tornado. Not all tornadoes are visible. Should we wait until a tornado is actually sighted before issuing the warning? What if the tornado spins up in the middle of a populated area? It happens. Who wants to be the mine canary? What about at night when visibility is low and few are awake to see?

Personally, I have always been conservative with warnings (I'm a long-time Freeper. What would you expect?), and tend to set the threshold rather high. Normally, warning forecasters want to see more than just rotation in a thunderstorm. We prefer to see vertical and temporal consistency. That means the rotation occurs over a vertical depth, rather than one level...and over a period of time, rather than just once. The threshold will vary depending on the situation. If it is a wild day with a high threat, the threshold will be lower. You can't take a lot of time to analyze one storm when there are 10 others going on at the same time. You have to decide, and move on to another storm. If they are moving at 60 mph, you also don't have time to dither. Warning decisions are usually made in a matter of seconds.

I always preach that the toughest decision is the one NOT to warn. This is because if you are considering a warning, then you have evidence of a threat. You must conclude that the evidence is not strong enough to go with a no warn decision. Now you are betting the public’s life on things we don't fully understand. The easy approach is to just issue the warning and “be safe”. Of course, this just waters down the value of warnings.

Bottom line, it is a tough job. There is so much uncertainty, and often conflicting information. You are gambling with real lives. I have had numerous fatalities during my watch, but fortunately they all occurred with warnings in place. I have been lucky. I have watched other warning forecasters not so lucky. They will carry the scars of the wrong decision to the grave.

mesoman7, rest assured that your efforts are appreciated, and take comfort that your warnings have saved countless lives. Thank you for your effort.

I remember when a local TV station installed their new Doppler radar back in the early 80s. One Sunday morning their tracking and warnings surely saved many lives - mine included - as they tracked a storm that hit (among other places) Mannford, OK. Though lives were lost (as I recall), I'm certain that more lives would have been lost were it not for the efforts of Gary Shore in Tulsa.