The most significant winter storm of the 2006-2007 season for Southeast Arizona so far delivered some impressive snowfall, even to lower elevations. The large scale pattern featured a highly amplified ridge across the Eastern Pacific with an amplified trough and a small, yet strong upper level low pressure across the great basin into the Southwest. This ~540dm 500mb upper level low moved from Southern Utah at 12Z on the 21st across west central Arizona during the day to Western Pima County overnight. This inside entirely overland track is typically cold for us with very little moisture to work with, even to the east of the upper low track. What made this storm unique is it followed a rather significant storm two days prior that moistened up the lower levels. In addition, as the upper level low tracked SSW across Arizona, the low level flow likely backed just enough to pull in considerable moisture from the Gulf of California.

This system had extremely impressive dynamics to work with including strong frontogenetic forcing, vorticity advection and upper level divergence as it approached southern Arizona the afternoon of the 21st. Rainfall which broke out across the metro Phoenix area during the morning of the 21st was a sign of things to come across Southeast Arizona later in the day. A solid area of precipitation moved into the Tucson metro area during the early-mid afternoon which initially helped cool the atmosphere, even across the lower elevations into the middle-upper 30’s. Snow levels were initially above 3000 feet. However, as the precipitation continued through the rest of the afternoon and into the early evening, snow levels dropped to the valley floors including the city of Tucson. IR satellite imagery showed cloud tops of -40C during the period of precipitation from 23Z-01Z. These cold cloud tops which coincided with the tropopause near 400mb on the 0Z KTUS sounding indicated the lift was very strong and the precipitation was convective in nature. It was during this period that snow mixed in across the city of Tucson even with temperatures still in the mid 30’s before sunset.

There was a short lag in precipitation before a second round of heavier precipitation organized across Southwest Tucson and moved across the Southern and Eastern Portions of the city between 130Z and 03Z. This second round was right along the interface between the dry slot wrapping in near the center of the upper level and the moist inflow. This portion of the storm was mostly snow and is what gave portions of the city measurable snowfall anywhere from a dusting to 2”. Although the cloud tops with this second round were a bit warmer (ranging from -30°C to -35°C), it was still convective in nature. The strong lift and dynamic cooling seemed to play a critical role in lowering snow levels down to the valley floors including the city of Tucson. As the snow fell in the city, the temperatures hovered near or just above the freezing mark slightly holding down accumulations. There is no official snowfall tally for the Tucson International Airport as the FAA takes limited observations which do not include snowfall.

After 03Z or so the precipitation came to an end in the Tucson metro area as the dry slot moved in and skies cleared. This is the portion of the event which became very tricky with regard to temperatures and public impact. With a new snow cover across much of the area, even with saturated lower levels, radiational cooling was kicking in as the temperatures dropped in tandem with the dewpoint. The 12Z January 22nd sounding did indeed show a low level radiation inversion.

As the temperatures and dewpoint both dropped freezing fog set in across the entire Tucson metro area overnight wreaking havoc on bridges across the city. Shortly after midnight there was a seven car pileup on 22nd St, along with numerous accidents on other bridges. This prompted the Tucson Police Department to take the unprecedented action of closing all the bridges across the city including those that cross the Rillito and Pantano Wash. Many Monday morning commuters were caught off guard as they were turned around at all bridges until the ice was scraped off with front end loaders, magnesium chloride was sprayed and sand (very rarely used) was laid down across city bridges.

Areas to the east of the Tucson metropolitan area including Cochise, Graham and Greenlee counties never made it into the dry slot and had on and off snow overnight and through Monday. Areas along the pivot point along the moist axis including Central Cochise county from Bisbee north across Highway 191 into Graham County and the City of Safford got a very impressive snowfall with amounts ranging from 8 to 10 inches.

How did the models do?

The models were quite late on getting a handle on this system. All of the guidance 3-5 days ahead of the event was showing the trough axis/upper level low staying to our east across New Mexico with little moisture across Arizona. The models started to change their tune on the 12Z run January 19th run which made the first trend to the west, However the moisture availability was rather limited. By the 0Z January 20th runs, the GFS and the UKMET made the first real aggressive change with a much moister/wetter solution while the NAM solution had the upper low track still over Western New Mexico. The next run from 12Z/20th showed all of the model solutions including the NAM/GFS/UKMET/ECMWF with the track over West/Central Arizona, However the timing was a little bit slow. From the 12Z Saturday run until the event Sunday afternoon and Sunday night the models had a pretty consistent handle on the moisture availability and the track of the system.

As of the 00Z/20th runs with the GFS and UKMET and the 12Z/20th runs of the NAM with the farther west solution showing a 500mb low in the 540dm range actually crossing Southern Arizona , all of the guidance was consistent with snow levels getting very low including the Tucson area. Before the dry slot moved into the Tucson area, the models were bringing 1000-500 thickness values in the mid 530dm range and 700mb temps of -10 to -12C into Tucson. 850 temps from the models were quite a bit more marginal for snow as the all of the models had temperatures dropping just below 0C across the Tucson metro area. It is unlikely the models were able to resolve the dynamic cooling which helped cool the boundary layer across the Tucson metro area enough to bring snowfall to the valley floors.

The QPF forecast is quite interesting in that the models showed quite a bit more precipitation for the event during the earlier runs from Saturday the 20th compared to the 12z run Sunday the 21st close to the time of the actual event. The 12Z/20th NAM run had about 0.2” for Tucson and the GFS had 0.3”. Sunday runs from both the NAM and GFS only had 0.1-0.2” across the Tucson area with 0.3-0.4” across the eastern zones which was underdone. The best explanation for the underdone QPF is the very strong lift and the ambient moisture already in place from the prior system resulting in a shorter time to saturate the atmosphere.

Once the models did get a handle on this system, even though the QPF was underdone overall they were not bad. It was clear there was going to be a sharp western cutoff from the precipitation, and indeed there was. In addition, the models kept hitting Graham, Cochise and Greenlee County hard in the moist axis on the eastern side of the upper level low, those locations certainly did get the brunt of snowfall.

The MOS guidance did struggle with low temperatures Monday morning the 22nd across Tucson. All 12z/21st MOS guidance had low temperatures at or above 32F for Tucson for early morning on the 22nd. Given that there was a light snow cover across the airport and skies had cleared out late in the evening, the temperatures dropped in tandem with the dewpoint into the upper 20’s.