Severe weather: When forecasts go wrong, and why we weren't right lately

This is more of an opinion and post-mortem piece on today’s severe weather, as well as recent events. As of this writing, there have been about 10 tornado reports and scattered reports of wind damage and hail. While that is still a noteworthy event – especially if you are directly impacted – there is sure to be a lot of criticism directed at meteorologists, and especially the Storm Prediction Center, regarding a sense of “crying wolf” after today and two other recent events – last Thursday (March 30) and Sunday (April 2) that both appeared to underperform. I will try to speak my mind here and make quite a few points.

1) The models today have performed quite poorly, which was even mentioned in one of the mesoscale discussions this morning (reference MD 440) and even the best in the business have had a hard time analyzing the setup. The HRRR – normally the best short range model – has been downright garbage today. Without good model data, we have essentially gone back in time 20 years. Back then, watch boxes were much less confident, risk areas were enormous (I’ve seen high risk areas covering multiple whole states that busted back in those days) and we are basically playing on the fly.

2) Three weather radar sites were down today in critical locations. They were in Maxwell AFB, AL (near Montgomery – technical issues), Fort Rucker, AL (near Dothan – lightning struck it last night) and Moody AFB, GA (near Valdosta – technical issues). Even though they are very high tech equipment, they are not foolproof. The Doppler system has plenty of overlap to ensure that adjacent stations can cover when one is under maintenance or has technical issues. However, it was extremely difficult to cover for three adjacent sites all being down at once. To make matters worse, they were situated in the broad warm sector which had very critical observations (which were missing) that would have been inserted into weather models. Given the missing or inaccurate data, they may have greatly worsened the model accuracy. Thus, efforts to provide backup radar sources – such as mobile Doppler radars – during critical situations should be considered.

3) Similarly, a high risk was also issued for Sunday (activity was relatively limited) and there was a Day 2 moderate risk last Thursday (pretty much nothing happened in that area). The more active days lately have been slight or enhanced risk for the most part. What we must all remember is that meteorology is a VERY difficult job. It’s no different than forecasting a winter storm where small changes can dramatically tilt snowfall totals. In this case, a change in the wind direction of 10 degrees or adding just a few degrees to the dew point can make a dramatic difference. Also, if the atmosphere is TOO unstable, convection could just develop to the point of choking themselves off, and that same convection could block moisture return elsewhere. We've seen a coastal MCS cut off moisture (last Thursday), slightly less shear than analyzed resulting in more linear storms (Sunday) and poor data analysis resulting in model issues on moisture and instability (today). There are many, many factors involved in forecasting severe weather, and we can’t just hug one model and expect results.

4) The Governor of Alabama declared a State of Emergency yesterday, and many schools and businesses closed from Alabama to North Carolina today. Some school districts also closed last Thursday for what turned out to be a non-event. It begs the question, are we too cautious? I would say, in the face of a potential major tornado outbreak which was supported by most models at the time, they made the best of a tough situation. The challenge is interpreting the message so that next time it doesn’t get ignored. We have seen tragedies at schools or workplaces in recent years and litigation is a big fear that districts and businesses have. I know that many people live in places that can’t easily withstand tornadic winds and issues over child care often exist too. Really, it is a tough call. But every life saved is important – and local tornado shelters can solve the issue as well. It’s part of my upcoming thesis and helpful on the road to a Weather-Ready Nation.

5) It is also important to remember that, even though you may not be affected and most areas may have not seen severe weather, some people did. 2 people died on Sunday morning in a mobile home from a relatively weak tornado in southern Louisiana. I’m sure their family is not calling these events a bust. Straight line winds and flash flooding have also resulted in damage and communities having to pick up pieces. Just because we don’t see the intense tornadoes which grab the attention of cable news doesn’t mean that it didn’t result in trouble.

6) Finally, TV stations have also taken heat for interrupting shows for weather coverage. While local TV stations pride themselves on entertaining the audience, the better stations focus on news first and foremost. Those stations should get a lot of respect for breaking away from a TV show that can be easily viewed online or later via DVR to serve the community. They are under no obligation – corporate or government – to do such, and there are a handful of stations that don’t care (I have complained before). I know Melanie Christopher at WJTV in Jackson, MS produced an excellent PSA stating the truth, and they are by no means not the only station that believes in community over entertainment. You may also think that your station doesn't need to break because the storms are not hitting your community. Remember, most TV markets cover many counties, and they are life threatening situations somewhere else in the area. Don't be selfish!

It makes me angry to know what some people think these days. I even see it among the elites of society. But we need to all know that science is never perfect and always challenging. That is why meteorologists have degrees in tough disciplines through extensive science and math courses. They are all out for everyone in their community and the nation as a whole. We need to understand the challenges and move forward.

Forecaster Craig Ceecee