Eastside Thunderstorms and Lightning

Yesterday, residents of eastern Oregon and Washington was provided with a great show of lightning, thunder, heavy rain and strong winds. The cause?   A line of strong thundershowers that moved northward out of California and Nevada during the early afternoon, strengthening during the later afternoon and evening.  Here is a visible satellite picture around 5 PM...you can see the east-west line of thunderstorms;  each individual storm is associated with white oval shapes, which are the anvils of the convection.

The Pendleton, Oregon radar got a nice view of the passing storms....here are a few examples an hour apart.  The red and blue areas are either very heavy rain or hail. (click on images to enlarge)

These thunderstorm were associated with a log of lightning...here are some samples from the national lightning detection network at 7 PM and 9:30 PM.

Lets take a look at the rainfall totals for the storm, which ranged as high as 1.29 inches at Heppner, Oregon (see graphic).  Do you know what happened at Heppner?  It was the site of the most deadly convective flash flood in U.S. history in which 247 individuals lost their lives.

Here is a picture of the aftermath of the storm on June 14, 1903.   There is a

Heppner, Oregon in 1903

comprehensive book on the subject published by UW Press called Calamity (info here).
Heppner is on the slopes of the Blue Mountains and located on a normally minor creek...thunderstorms in the Blues turned it into a torrent.

Strong winds accompanied these powerful storms--take a look at the observations at Walla Walla.  Winds gusted to 47 knots (54 mph), with higher gusts at nearby Hermiston (60 mph).  Before the storm the temperature was near 80F with a dew point of 62F.  Folks,that is a really high dew point for these parts....more like the eastern U.S.

How good were the forecasts?   Well, I wish I could cite this as another success for the WRF and other models, but it wasn't.  The morning WRF model run (initialized at 5 AM Monday) showed very little, just a few light showers moving through (see graphic of the 24h hour precipitation ending 5 AM this morning)

24-h precipitation from UW WRF ending 5 AM

This failure was also evident in the NWS model (NAM) and the UW ensembles.  The interesting thing is that the models and soundings showed great instability (high CAPE--convective available potential energy);they just didn't have any forcing (front, upper level short wave) to trigger them off.   The NWS folks in Spokane discussed this situation during the afternoon and were not sure what to predict, so they just noted the potential for thunderstorms, with a lot of uncertainty.  Here is what they said:

"Tonight and Tuesday...Upper level ridge axis remains fixed over 
  the PacNW...with deep southerly flow continuing to pump moisture 
  and instability northward. Model soundings and LAPS analysis show 
  CAPE values exceeding 1000 j/kg over most of our forecast area 
  east of a line from Republic to the Tri-Cities. These values are 
  very impressive for April...let alone any time of year...but that`s 
  what happens when surface temps rise into the 80s...with dewpoints 
  in the mid 40s to mid 50s. Despite the potential 
  instability...tapping into it is no sure bet...especially given 
  the lack of a trigger or focusing mechanism. Looking at the latest 
  satellite imagery...there really is nothing downstream of us which 
  will trigger convection in the immediate future. Looks like better 
  chances will arrive as the convection over southern OR/ID move 
  northward overnight. There is a small chance some of this will 
  reach the WA/OR border toward sunset...and spread gradually 
  northward through the overnight hours." 

This was a very reasonable analysis.  The NWS staff in Pendleton were even more emphatic about the potential  for thunderstorm...and they were correct.    Shows you the importance of having human forecasters for such difficult events.

Clearly, there was some forcing feature\s that the models missed...it will be interesting to explore this case further.  The runs initialized at 5 PM were much better, but they weren't available until after the action got started.