Bad forecast Monday 

Posted: 4:40 am Tuesday, June 26th, 2018

By Kirk Mellish

I often observed that some of my worst forecasts are on Mondays. Perhaps its the weekend “hangover” even if not the alcohol type. It’s just being a little less in touch with data from partial down-time on weekend, and of course in this case a Monday after a week of vacation.

Yesterday was definitely a case in point. It was a busted forecast on my part, I didn’t anticipate an organized squall line to form based on the overnight data available at the start of the day. Storms were in my forecast, but only of the hit and miss variety.

An area of enhanced lift from low pressure (surface and aloft) over Kentucky and Tennessee was anticipated to move east mainly impacting KY, TN, far North and East Georgia and the Carolinas, leaving Atlanta with just the usual pop-up storms.

See the morning visible GOES-16 Satellite with the cloud swirl in Kentucky and Tennessee where the low is along a stationary front:

Screen Shot 2018-06-25 at 6.26.25 PM


However, what happened in reality is that the weather observation network and models failed to detect that there was an area of vorticity energy and lift extending farther to the Southwest into Northern MS/AL that would head our way.

It became increasingly clear late in the afternoon that such a feature was there once Satellite imagery showed a hint of a line trying to organize to our NW:

Screen Shot 2018-06-25 at 6.09.40 PM

Late in the afternoon the weather net and models “found” the features missed the prior 24 hours…

You can see the absolute vorticity in the 500mb jet stream flow (dashed red lines with vorticity max indicated by pink X and numerical values) from KY trailing Southwest to northern Louisiana:

Screen Shot 2018-06-25 at 6.40.10 PM

And the resulting strong upward vertical motion field associated ahead of it (700-500mb Upward Vertical Velocities):

Screen Shot 2018-06-25 at 6.42.48 PM

And as a result of those previously unknown features, BOOM goes the atmospheric dynamite:

Screen Shot 2018-06-25 at 6.13.35 PM

It may have been aided and abetted by a smaller convectively induced mesoscale vorticity (MCV) and outflow boundaries (kind of mini-cold fronts) from earlier storms in the lower Midwest the night before:

Screen Shot 2018-06-25 at 6.48.57 PM

Had those features been captured by the weather grid a stormy forecast would have been more straight forward to make.

Lots of lightning associated with the storms along with isolated damaging wind gusts and some hail with the heavy downpours as it developed into a QLCS/MCS.

(WxBell graphics):




This is why you have seen my tweets and blogs many times in the past explaining how thunderstorm forecasts are DIFFERENT from other types of forecasts and thus sometimes require updates every few hours instead of every 12. Monday was a clear case in point.

Had the weather balloons better detected the vorticity and lift that was actually heading our way then the models would have provided better guidance and the morning forecast would have reflected the greater threat.

It’s interesting to note, that some of the best national aviation forecasters and severe weather forecasters never issued updates reflecting the situation even as it was ongoing. That’s highly unusual and reflects the rapid and unexpected evolution of the Monday weather situation.

Anyway, mea culpa.

Follow me on Twitter @MellishMeterWSB.

To further underscore the thunderstorm forecasting dilemma, witness for yourself some data projections from early this morning Tuesday June 26th…

This is one Hi-resolution CA model (convection-allowing) it’s run once an hour.

Well just one hour part it shows today being very quiet thunderstorm-wise (convection) OR being rather active at the end of the day!

O N E hour apart. smh


Easy-smeasy. Yeah right.