Wednesday, November 30, 2011

How Predictable Are Downslope Winds?

The warning flags have been raised and the National Weather Service has issued a high wind watch for late tonight through Thursday evening for the Salt Lake and Tooele Valleys and northern Wasatch Front for strong and potentially damaging downslope winds.

As we have discussed in earlier posts, there is good reason for issuing such a warning.  The key large-scale ingredients for a downslope windstorm have been anticipated by earlier computer model forecasts and are evident in the latest forecast produced by the NAM model.  These key ingredients are strong easterly to northeasterly flow, a stable layer near crest level, and a level aloft with reverse shear (i.e., winds weakening with height).

NAM model forecast valid 8 AM MST Thu 1 Dec 2011
However, although our computer models today can predict the environment in which a downslope windstorm is likely, they provide us with little useful information with regards to just how strong the peak winds will be.

The National Weather Service Salt Lake City Forecast Office now runs a high-resolution (4-km grid spacing) computer modeling system that better resolves the Wasatch Mountains and can produce downslope windstorms, but it produces a single forecast.  The latest forecast produced by this modeling system shows strong (>50 knot) winds in the Farmington area tomorrow morning.

SLCFO WRF-model forecast valid 9 AM MST Thu 1 Dec 2011
Unfortunately, we can't hang our hats on that 50 knot wind speed.  There are several reasons for this.  First, even at 4-km grid spacing, the details of the Wasatch Mountains and the processes that contribute to downslope windstorm development are not fully resolved.

Second, recent work by Reinecke and Durran (2009) suggests that the magnitude of the winds produced in computer model forecasts are very senstive to small changes in the initial conditions.  Specifically, they ran a 70 member forecast ensemble to examine the sensitivity of two downslope windstorms in the Sierra Nevada to slightly different initial conditions.   In one event, the difference in downslope wind speeds produced by the strongest 10 and weakest 10 members of the ensemble was 28 m/s (56 kt) for a 6-hour forecast.  In a second case, the difference was 22 m/s (44 kt) for a 12-h forecast.

Therefore, in the event anticipated for tonight and tomorrow, high, potentially damaging winds are likely, but the peak magnitude of the event remains uncertain.  Hopefully in the future, when we have high resolution forecast ensembles with many members, we'll be able to provide better guidance with regards to the likelihood of winds that exceed critical thresholds.

Update @ 1:40 PM MST 30 Nov: Prof. John Horel will be leading a "storm chasing" team into the storm late tonight through tomorrow morning.  They will be collecting observations with our portable mesonet stations, mobile truck-mounted mesonet stations, and two portable upper-air sounding systems.    I'd also like to note that the Reinecke of the Reinecke and Durran (2009) paper above, is Alex Reinecke, a University of Utah alum who worked in the Atmsopheric Sciences department and is currently at the Naval Research Lab.  Nice work Alex!

No comments:

Post a Comment