The Moab Flood

Brian Schott has an excellent article discussing the Moab flood and some of the problems with the use of phrases like "100 year event" that was published yesterday in the Salt Lake Tribune.  In it, Moab's city engineer, Chuck Williams, does a good job describing some of the concerns about that phrasing and also how prior rainfall set the stage for the event by saturating soils (where they exist in that red rock country).  

I'll talk a little here about the precipitation piece of the puzzle.  Below is the  National Severe Storms Laboratory's Multi-Radar Multi-Sensor (MRMS) system 1-hour accumulated precipitation estimate for the period ending at 1:50 UTC 21 August, or 7:50 PM MDT 20 August.  Based on my visual inspection, this was the 1-hour period with the greatest coverage and intensity of precipitation in the Moab area.  Precipitation is heaviest along US-191 and in the Behind the Rocks area to the west, with pixel-level values maximizing at about 1.4 inches.  To the east, in the Mill Creek basin, amounts vary from about .6 to 1.2 inches depending on location.

Source: https://mrms.nssl.noaa.gov/qvs/product_viewer/

I also took a look at shorter-time-scale precipitation estimates.  Radar-estimated precipitation at one location peaked at about 0.7 inches for a 15 minute period (note: this is one pixel, not the entire area).  Based on the Salt Lake Tribune article, the NWS received a report of nearly an inch in 20 minutes, so these are fairly close. 

Now if we were to look at the so-called recurrence interval for such precip amounts in the Moab area, 0.7 inches in 15 minutes has a recurrence interval of about 50 years and 1.4 inches in an hour about 100 years.  That sounds like a rare event.  However, these recurrence intervals have many problems, some discussed in the Tribune article linked above, but one not mentioned is that for precipitation, they are for a point.  During the monsoon we basically have thunderstorms dropping heavy precipitation in highly localized areas.  The odds that a specific location gets 0.7 inches of rain in 15 minutes is pretty low.  The odds that it happens somewhere in southern Utah is, however, much higher, and this is especially true during an active monsoon season like this one.  

I suspect that if I put some effort into it, I could find multiple 100 year precipitation events at various locations in southern Utah this summer (and probably northern Utah too).  Another way to think about that, and this is a bit of an oversimplification (but so is recurrence interval), is that about 1 in 100th of southern Utah will see a 100 year storm in any given summer.  The Moab event was rare for Moab, but it may not be rare for southern Utah in the monsoon.  

I often like to say that we are not prepared for the climate of the 20th century, let alone the climate of the 21st century.  The frequency or recurrence intervals of extreme events in our part of the world my contain estimate uncertainties simply because they are based on a limited time period at a limited number of sites.  This is especially true for precipitation accumulations in short time intervals like an hour or 15 minutes.  Additionally, for precipitation, these are point estimates and what would be really helpful is information on the characteristics of precipitation within individual catchment basins.  Radar can provide a path forward, although there are many areas of Utah that remain undersampled.   Finally, in a warming world, the statistics of the past may not represent the statistics of the future and storm intensity is expected to increase.  It will be essential that we learn from events like this one and build resiliency to future extreme weather and climate events.