Friday, August 9, 2019

Downpours, Debris Flows, and Road Closures, Oh My!

Yesterday evening's storms came through with a vengeance for some areas of northern Utah, generating heavy precipitation and debris flows that forced the closure of SR-210 in Little Cottonwood Canyon, SR-6 in Spanish Fork Canyon, and the Amtrak line through Spanish Fork Canyon, as well as the evacuation of residents near Loafer Canyon.

It was only a couple of weeks ago that the Tour de France was forced to alter stages due to severe thunderstorms and debris flows in the French Alps.  At that time, I wrote that organizers of the Tour of Utah should take the opportunity to learn from that experience since they be dealing with a similar scenario (see Weather Hazards and Bike Racing).  Although I recognized this as a possibility in Utah, I wasn't anticipating that we would see such a severe event this year.  Nevertheless it has happened, and the prologue scheduled for Little Cottonwood Canyon is only three days away (Monday, August 12).  Of course, when it comes to incidents like this, there are more important issues at play than a bike race.  Thankfully, I haven't heard any reports of fatalities or injuries.

Let's take a look at what happened in Little Cottonwood late yesterday.  To begin this discussion, we should first note the conditions leading up to the event.  On August 1st, the Alta Coop Site reported 0.40 inches of precipitation.  On the 3rd, the it reported 2.11 inches of precipitation.  That event produce runoff in the upper canyon, with shallow layers of mud and rocks covering the road in the upper canyon in places (see Saturday's Cottonwoods Deluge).  The next four days leading up to yesterday's event produced 0.01, 024, 0.01, and 0.02 inches of precipitation.  Thus, the water content of soils in the canyon was likely high, limiting their ability to absorb and retain precipitation.

I suspect it rained lightly yesterday morning (although automated gauges at Collins and Alta Base show no measurable precipitation).  As that precipitation band exited, skies cleared and band of precipitation with embedded thunderstorms began to move into northern Utah.  See the tweet below for radar image at 2222 UTC/1622 MDT).  The stage was set for late day thunderstorms.

Indeed, those storms produced, with embedded cells generating locally heavy precipitation.  The radar loop below covers the period from 0102-0220 UTC/1902–2020 MDT with a black line indicating the ridgeline between Little and Big Cottonwood Canyons.  It shows a very intense cell moving across the southern Salt Lake Valley and into the central Wasatch.  Radar reflectivities above 35 dBZ and frequently above 50 dBZ, indicative of heavy precipitation, linger on the ridge for a period of about 40 minutes.


Radar estimates of precipitation have their issues and so one needs to recognize their can be large uncertainties, but the one I have access too pegs the maximum accumulated precipitation for the 1 hour period ending at 0220 UTC/2020 MDT at about 3 inches very near the Little Cottonwood-Big Cottonwood Ridgeline.  That pixel appears to be just north of Lone Peak in the upper Broads Fork/Lake Blanche area.  Although that is north of the divide, there is some uncertainty in the exact location of where precipitation falls out due to transport by the wind and other factors in and below the radar sampling volume.


The bottom line is that heavy precipitation fell in a short period of time in an area where soils were already somewhat compromised by prior precipitation.

Given that the storm came in at dusk, UDOT is just starting to send out videos of the damage.  Here's one from their Public Information Officer John Gleason.

Inevitably, I expect that we're going to hear that this was a "one in whatever year" event.  Based on the NOAA Precipitation Frequency Data Server, a three inch accumulation in 60 min has an average recurrance interval at 10,500 ft in the central Wasatch of about 500 years.  Or, better put, the odds of such an accumulation happening at any point in that area in any given year is about 500 to 1.

However, some caution is needed here.  First, I'm not sure how confident to be in the radar precipitation estimates.  That is going to require some digging.  Sadly, I don't think there is a precipitation gauge very near the area of maximum precipitation.  Alta-Collins, Alta-Base, and the Snowbird SNOTEL reported .45", .59", .7", respectively, in the 1-hour period ending at 0200 UTC/2000 MDT, but they were outside the core of heaviest precipitation.

Second, as you can infer from the 1-hour accumulation image above, thunderstorm accumulations can be very localized.  Remember that 500 to 1 is a point probability, not an area probability.  The odds of such an accumulation happening somewhere in the Wasatch Range would be greater.

Nevertheless, this was certainly an exceptional event with remarkable impacts.  It serves as a reminder of the vulnerability of mountain communities and highways, especially SR-210, to severe and hazardous weather.

Blogger's Note: The recurrence interval figure has been updated from the original post to include the legend. 

No comments:

Post a Comment