Sunday, June 15, 2008

June 16

Classic Day 2 outlook for June 2. Not this year. This year the upper low over the midwest set up shop in late June with a block forecast for the next 184 hours. This should put a hold on severe weather somewhere south of IA, though you can never be too certain of that (the dryline, weak disturbances, diurnal destabilization, etc).

In fact some May25-June15 daily composite geopotential height maps (2006, 2003, 2002, 2001) all indicate a ridge at 500 hPa over the Plains and these were slow years for severe storms in the Plains. The other years were biased toward SW flow with a trough hanging off the S Pac coast (2008) or a more zonal pattern across the Plains (back to 1998). We already know there are favorable patterns for convective storms but there are a lot factors that contribute to the favorable synoptic setup. One thing is certain: this year was favorable for severe storms including the "floods o
f 08".

I created a figure using the oceanographers temperature salinity diagram [I used the log of the counts rather than the counts to highlight the lower numbers]. The figure shows day of the year along the X axis, and number of tornadoes in the Y axis. Its a 2d histogram basically (using a 7 day bin and a 4 tornado count) for 1979-2007 over the CONUS. Anyway, it shows the double frequency maximum of outbreaks centered near day 130 and day 300. It also shows (via the yellow frequency band) when tornadoes are typically observed (from day 100 to 230 or April to late July). Why show this figure? To highlight the rarity of the number of days where tornado occurrence exceeds around 20. So what happened in 2008?

2008 tornado days in May above 20[4]; 30 [1]; 40[1]; 50+[5].
June 17th [3] [1] [0] [3]

These numbers will come down, but the number of days is impressive. 33 percent of the months May and June had what can be considered outbreak type days. That is obscene. No doubt the tornado counts only spell part of the story. The predominant convective mode was supercellular and thus along with tornados came heavy rain. Seems pretty obvious, but this is one year where a regional climate model should be compared to observations (not for the flood) but for detection of the cape-shear environment.