Mile-wide tornado smashes Windsor, Colorado; plus, hurricane season commentary

A mile-wide tornado swept through Colorado between 11am and noon yesterday, ripping the roofs off buildings, tossing cars into the air, and killing at least one person in Weld County, northeast of Denver. Hail up to 2.75" in diameter accompanied the storm, which took an unusual north-northwesterly track , parallel to the Rocky Mountains north of Denver. Hardest hit was the town of Windsor, between Fort Collins and Greely, where damage appeared to be at least EF3. A local TV station took some impressive live video of the tornado, and a wunderground web cam video in Windsor, Colorado (Figure 1) caught the funnel as it passed east of the camera.


Figure 1.Webcam view looking east at 11:45am MDT in Windsor, Colorado as the tornado passed by. Note the golf ball-sized hail covering the ground. Image credit: windsorweather.com.

The Weather Underground's tornado expert, Rob Carver, had this explanation of yesterday's tornadoes:

Two low pressure systems over the Western U.S. were the cause of the severe weather outbreak of May 22. A strong upper-level low over the Great Basin formed a surface low in the lee of the Rockies, and brought a strong southerly mid-level jet over the Plains of Colorado and Kansas. As the surface low formed, it brought warm moist air northwards, forming a warm front. This warm air moved westward, rising with the terrain, causing thunderstorms to form. Once these storms formed, the juxtaposition of easterly flow at the surface with southerly flow aloft (i.e., wind shear) produced significant spin in the horizontal direction, which was tilted by the storms to vertical spin, which triggered the formation of tornadic thunderstorms. This movement of the warm front up the slope of the Rockies to help trigger tornadic thunderstorms is a rare occurrence.



Figure 2. Radar reflectivity image (top) of the Windsor, Colorado tornado at 11:45am MDT May 22, 2008. Note the classic hook-shaped echo associated with the tornado. Bottom: Doppler velocity image of the tornado, showing a small core of red and blue colors right next to each other, denoting strong winds towards and away from the radar, the classic signature of a tornado vortex. For those interested, we've saved an animation of the reflectivity and Doppler velocity.

Severe weather forecast
Severe weather is expected today over Kansas, Nebraska, and surrounding states. NOAA's Storm Prediction Center has placed portions of this area under its "Moderate Risk" category for severe weather, one step below its highest level of concern, "High Risk". Yesterday was also a "Moderate Risk" day, which SPC later upgraded to "High Risk" once the tornadoes started pounding Colorado. More severe weather is expected Saturday and Sunday over the Midwest as the upper-level low pressure system responsible moves slowly eastward. The Weather Underground Severe Weather page and Tornado page and WebCam page are good places to go to follow the severe weather. Also, tune in to the chase accounts and awesome storm photos from Wunderblogger Mike Theiss, who was in Kansas yesterday, and has posted many spectacular photos of yesterday's storms. According to Mike's blog:

Today Cloud 9 Tours saw 4 tornadoes and ONE developing almost overhead. We experienced winds over 100mph from the circulation of the meso that passed overhead as a cone tornado developed. Stay Tuned....

Possible development in the Eastern Pacific late next week
The past four days, the ECMWF model has been predicting the formation of a tropical storm in the the Eastern Pacific, just off the coast of Guatemala, around May 29. The GFS model has also been predicting something might develop, but in the Western Caribbean near Mexico's Yucatan Peninsula. The GFS has been rather inconsistent with its handling of this potential storm, and I am inclined to discount its forecast--especially since last night's long range runs of the NOGAPS, Canadian, and UKMET models all show development in the Eastern Pacific, not the Caribbean. All five models predict a northward shift in the jet stream and substantial relaxation in wind shear over the Eastern Pacific and Western Caribbean next week. It is common to see May tropical storms in the Eastern Pacific, and it would not be a surprise to see something develop there. I'd put the odds of something popping up in the Western Caribbean or Gulf of Mexico at less than 10%, though.

NOAA's seasonal hurricane forecast
NOAA issued its annual seasonal hurricane forecast yesterday, which I discussed in detail in yesterday's blog entry. Yesterday's forecast came out with a little more uncertainty attached to it than previous forecasts, which is a good thing. The media attention and fanfare that accompanies these forecasts is rather excessive, given the low skill they have. In fact, we don't even know if the NOAA forecasts have ANY mathematical skill, because they've never released a verification study of their forecasts. I doubt that the skill is very high--as Eric Berger of the Houston Chronicle pointed out in a blog yesterday, NOAA has blown its forecast of Accumulated Cyclone Energy (ACE) each of the last six years. NOAA held its usual press conference to announce the forecast; this year, it was held at the home of the NOAA Hurricane Hunter aircraft, MacDill AFB in Tampa. The NOAA dignitaries present said all the right things, preaching the need for preparedness regardless of the forecast for the upcoming hurricane season. Still, I wonder if NOAA might be hurting themselves by making such a public spectacle over the release of a forecast that no one knows the accuracy of, and has performed poorly by some measures in recent years. I do like the fact they are issuing public hurricane forecasts, as I expect their accuracy and value will improve in coming years, but they're definitely not worth the attention they're getting at present.

Jeff Masters