Friday, March 24, 2017

Looking Back at Yesterday's Epic Deluge

Yesterday we were in clear outlier mode in Salt Lake City as we broke the all-time record for precipitation on a calendar day in March and experienced one of the wettest days ever.  I mangled some of the numbers in the previous post, so we'll take a little bit of time today to look back on the event and set the record straight.

The primary large-scale feature driving the event was a mid-level trough that moved very slowly across Salt Lake City and stalled very near the Wasatch Crest for most of the day.  This trough was a large-scale feature extending northeastward to the Canadian border, and served as the locus for precipitation development along much of its lengths.  The image below shows the scene at 2000 UTC (2 PM MDT).  Wind barbs are at 700 mb (10,000 ft).  Gaps in the radar echo coverage in portions of Wyoming reflect both poor radar coverage and some orographic effects.

Cloud and precipitation bands of this type are sometimes called "wrap-around" because the wrap around the backside of the surface or low-level trough.  Indeed that was the case yesterday, as shown below with the sea level pressure analysis, although I'm not entirely satisfied with that classification of this system for a variety of reasons I won't get into here for lack of time.

The National Weather Service reports that the daily precipitation was 1.97".  Note that this is for the midnight to midnight calendar day based on Mountain Standard Time.  We are currently on Mountain Daylight Time.  If the calendar day was defined based on daylight time, the daily total would have been an even larger 2.07 inches, based on surface airways observations provided by the airport.  This highlights an important aspect of meteorological records.  Calendar day records are not the same as 24-hour records, since the latter can use arbitrary start and end times.  Often, major precipitation events straddle calendar days and are broken into two smaller pieces.  If you are using statistics of 24-hour precipitation based on calendar day records to design your storm-runoff system, this is an issue to consider!  Nevertheless, many meteorological records are based on calendar days, I suppose for historical and practical reasons.  For example, many of our meteorological records are collected by volunteers who provide daily observations of maximum temperature, minimum temperature and precipitation amount.

Below is a summary of the 50 largest calendar day precipitation events at the Salt Lake City airport and, prior to the creation of the airport, downtown.  1.97" makes yesterday the 6th wettest calendar day on record and the wettest March day on record.

Source: NOAA Regional climate Centers
The precipitation, however, was not evenly distributed during the calendar day.  It was heaviest prior to about 6 AM MDT, when over an inch fell in the prior 6-hour period, and ceased at 5:40 PM MDT (at the airport, showers lingered on the east bench).  Radar imagery at about 4 AM MDT (0957 UTC) shows that the area around the Airport and portions of the I-15 corridor to the north experience very heavy precipitation with areas of radar reflectivity greater than 35 dBZ (yellow) and localized pockets above 40 dBZ.  Although I'm not showing a loop, these areas of heavy precipitation were quasistationary for much of the night and very important in pushing the precipitation amounts at Salt Lake City to record levels.  

Source: NCAR/RAL
As wet as it was at the airport, there were locations that were wetter.  Tooele recorded 2.49 inches of precipitation and the Rocky Basin Settlement SNOTEL in the Oquirrh Mountains recorded 2 inches of precipitation water equivalent and 22 inches of snow from midnight to midnight MST.  

The central Wasatch didn't do as well because the strongest precipitation along the band was just to the west, along with the northwesterly flow behind the mid-level trough.  As a result, the large-scale precipitation dynamics in the central Wasatch yesterday were weaker.  If one looks at the wind time series on top of Alta's Mt. Baldy, one can see the slow trough passage from 0000 MDT on the 23rd to 1800 MDT on the 23rd with the wind shifting very gradually from south to north-northwest.  Note, however, how weak the winds were for much of the period, especially from 0000-1200 MDT.  It wasn't until the trough had passed late in the day that the northwesterly flow intensified, but by then, the larger-scale precipitation dynamics were dying.  

Source: MesoWest
It wasn't a total disaster for the central Wasatch.  For the midnight to midnight MST period, Alta-Collins picked up about 9 inches of snow and 0.82" of water.  However, this is less than observed in the northern Wasatch and far less than observed in the Oquirrh Mountains.  

The spring pattern continues tonight and tomorrow with a cold front moving in and bringing mountain snow to the central Wasatch starting early tomorrow morning.  The NAM meteogram below tells the story pretty well.  This looks to be a quick hitting event, dumping several inches of snow in the morning, tapering off quickly to snow showers afternoon.  

Last night's NCAR ensemble produced anywhere from 0.6 to 1.6 inches of water for Alta-Collins.  Most members are between 0.6 and 1.1 inches of water.  

Probably 4-8 inch totals lie in the most likely range, but lets keep our fingers crossed we can do a little better.  


  1. One interesting aspect of the storm yesterday was how cold it got in some lower valley areas, with the airport turning to snow for a while even with 700-mb temp near -6 C and a warm surface to begin with. Many afternoon temps in the Salt Lake Valley were barely warmer than stations near 7,000 feet. My thought is this is due to a combination of some drier air getting ingested into the precip band at lower levels (evaporational cooling) and cooling due to so much falling precipitation in the melt zone. But I was surprised to see the snow level down to 4,000 feet given the overall situation.

  2. Are there any numbers out there for 24-hour precipitation records? Yesterday's 24 hour total looks like it was somewhere around 2.14".