Detailed Review Of The Mega Snowstorm Of 2011

Folks, I want to apologize right off the bat to most of you. This blog post is very technical and lengthy. I put together this article for our use in the weather center to better understand and document the storm. However, I thought that there may be a few of you that would understand and enjoy the article. If you are looking for a more basic review of the storm you can check out the following link

Mega Snowstorm 2011


It was a very bittersweet early morning. A nightmare and a dream were coming true at the very same moment. The 4th snowstorm within 8 days was starting to move into northeast Oklahoma and northwest Arkansas. When my day started around midnight on February 9th, I knew it was going to be a busy and snowy day but had no idea that I was just a few hours away from seeing a historic storm that would easily rival the worst snowstorms in Arkansas history. I have seen some impressive upslope snowstorms (mountain snows), lake-effect snow, and classic Midwest blizzards, but this was much different.

 Science Of The Storm

The storm system from a synoptic point of view was relatively weak, especially compared to the storm that moved in just a week prior. The storm was also driven mostly by mesobands of snowfall that were not forecasted well or at all be numerical computer models. There was only a moderate amount of moisture to work with compared to previous snowstorms. The mixing ratios were at 2g/kg in the 650 to 750 mb layer. However, there was an incredible amount of lift with this storm and the lift was driven even more by not one but two sharp short waves and strong isentropic lift that moved into the region. The computer models had disagreed on the placement and strength of the 500mb wave several days before the storm had arrived. It became more obvious in the 24 hours leading up to the event that the wave would be much stronger in the southern part of the area, but there was a second wave developing over Kansas and moving to the southeast as well. By late evening on the 8th (00z to 6z on the 9th) the models finally began to give the northern wave more credit and increased its strength. As the wave was heading southeast the southern wave was heading straight east. Right before the two waves merged, positive vorticity advection maximized in between the waves leading to maximized lift right over northeast Oklahoma and Northwest Arkansas. This along with strong isentropic lift led to an intense mesoband of snowfall over the area. Snowfall rates reached 2-4” per hour along that band. The biggest problem was that the band didn’t move for about 6 hours leading to extremely high snow totals in the range of 18-24”. Finally the two short waves merged into one wave and pushed east/southeast later in the day.


Most of the models continued to show the highest QPF in the southern part of the area, most likely in response to the best lift on the models which was out ahead of the southern short wave. There was also a huge difference in model output anywhere from 0.10” on the GFS to 1.2” on the NAM. In the 36 hours leading up to the event. The GFS increased QPF to 0.2” while the NAM lowered totals to around 0.5”. The Canadian and European kept showing around 0.4” consistently during the event. While, an average of 0.4” did fall in the river valley and Ouachitas the NAM was on to something. While it still failed to forecast part of the mesoband of snowfall it did accurately predict some of the other types of lift involved with the storm in the last 12 hours before the storm hit.

Our forecast started at 2-4” in northwest Arkansas 24 hours before the storm hit with 4-7” in the river valley. 12 hours before the storm hit the forecast was upgraded to 3-5” in northwest Arkansas and 5-7”+ in the river valley. At midnight on the 9th we changed the forecast to 4-8” area wide. This forecast made it to the blog but would not make it on-air. Due to detecting and observing the mesoband, a change was made to the northwest Arkansas forecast again to compensate for the intense, narrow band of snowfall. The forecast was changed to 8-12”+, which was changed less than an hour later to 10-16”+. The final forecast was made around 7am which included all of NWA in a 10-18” band with a very narrow swath included in an 18-24” band. The river valley’s forecast changed from the 4-8” to 5-10” during that time.


Delaware, Benton, Washington, Madison, and Newton Counties were hit the hardest. All of these areas picked up on 12 to 25” of snowfall. There was a line roughly along US 412 that had the highest totals. This band stretched from Jay, OK to Huntsville, AR where 20 to 25” fell. The highest elevations in northern Crawford, Franklin, and Johnson counties had 8-12” while the river valley generally had 4-8”. The snow ratio varied quite a bit across the area from as low as 10 to 1 to as high as 18 to 1 in most spots. The average snow ratio was much lower than expected considering the cold conditions and model QPF. The average snow ratio across the area came out to 13.8 to 1. This means that within the mesoband actual precipitation amounts were over 1.5” in spots and that has been documented. At the station in Rogers, AR 1.55” (17 to 18” of snow) was recorded. At my house in Springdale, AR 3 core samples revealed an average of 1.60” of precipitation (20” of snow).

 Storm Investigation

As mentioned above, the key to this storm was the exceptional lift. There were several types of lift that led to significant snowfall. First, there was great upper-level divergence area wide due the placement of the jet streaks. Second, there was strong isentropic lift. The strong isentropic lift combined with very high levels of positive vorticity advection out ahead of the southern short wave. This led to a frotogenetic band (mesoband) of heavy snowfall between 8z and 11z (2am to 5am).

 Snow rates were 2-4” per hour in this band. A second shortwave approaching from the northwest moved close enough to the area to provide additional positive vorticity advection around 12z (6am).

Around 13z (7am) the two short waves phased over the area leading to intense snow rates over 2 to 3” per hour in the same vicinity where the initial frotogenetic band developed earlier.

Finally the phased waves began to shift southeast around 14z (8am), however this is when the slightly weaker, but still substantial deformation zone developed over northeast Oklahoma.

Unlike, the persistent mesoband earlier in the morning, the deformation zone was moving on the back side of the wave progressing to the southeast through the morning. At 17z (11am) the deformation zone shifted to the Arkansas river valley and was already showing signs of weakening.

Upper Levels

 (11z or 5am image of the 1.5 degree velocity showing the low-level convergence and an isentropic wind pattern)

(16z or 10am image of the 1.5 degree velocity showing the change of wind direction and location of the deformation zone)

How the Models Handled the Storm

To sum this up in one word…Bad! However, looking closer at the storm there were aspects where the models did very well and other areas where they did very badly. The computer models as a whole did very poorly more than 24 hours away from the storm. The models continued to struggle with the upper-level wave driving the storm system. However within 12 hours the models, especially the 6z NAM, did much better with the storm. They still only had about 1/3 of the total precipitation/snowfall forecasted compared to what fell within the mesoband, but they did model certain aspects of the storm correctly.

First off the 6z NAM did forecast the upper-level divergence very well and most of the models did accurately forecast that the QPF would be near 0.30-0.50” (which occurred outside of the mesoband). The problem was that not a single model accurately depicted the mesoband from 8z (2am) through 12z (6am). This band formed due to a combination of high positive vorticity advection and strong isentropic lift. This band helped to produce in upwards of an additional inch of precipitation (12”+ of snowfall) during that time.

 Lessons Learned

This is a humbling reminder that there are not enough observations making their way into the models, especially upper-air observations. These errors in combination with model resolution (which is still a little too broad to resolve fine meteorological features such as a frotogenetic snowband) were a big part of why the models failed to pick up on these features. Given the data at hand, there was no way a forecaster could have had any confidence that two feet of snowfall would fall in northwest Arkansas 24 hours before the storm moved into the area. Better data and finer resolution models would certainly help in the future.

On the other hand, one feature could prove to be a warning sign in the future that intense mesobands of snowfall could develop despite the computer model’s forecast. Areas of high positive vorticity advection moving into an area of high isotropic lift could lead to an increased potential for frotogenetic snowbands. This seemed to cause a lot of extreme snowfall yet the models didn’t make this connection. The two short waves phasing together also helped contribute to significant snowfall, but the models did at least try and respond to this process. The NAM at least did forecast the deformation zone on the backside of the storm and upper-level divergence very well.


13 Responses

  1. So basically what you are saying is this was really a secret project of the federal government…….

  2. could we possibly see another snow storm this year?

    • Dave,

      It is possible. The long range pattern will be very active through the 1st week of March.

      Ross Ellet

  3. I heard that the almanac is calling for one more snow in March. What are your thoughts on this possibility?

    • Heather,

      Another snowstorm would not surprise me. We will be in an active pattern through the first week of March.

      Ross Ellet

  4. This was certainly a snow storm for the books! I haven’t seen this much snow since we moved down here from NW Indiana!!! Which is famous of course for it’s “lake effect” snow!(Well those of us who lived not too far from the Indiana Illinois boarder!
    Thank you for all your effort in keeping us informed!!! It was certainly appreciated!!! Keep up the great work guys!!! We’ll keep watching 40/29 weather/news!!!

    • Patti,

      Thanks for the kind words! You are right, even though it wasn’t lake-effect snow, it sure had a lot of similar characteristics.

      Ross Ellet

  5. Weather geek-out. I love it, thanks. Very interesting. My wife is making fun of me for even reading it.

  6. I heard that the Farmer’s Almanac is calling for one more snow storm in March. What are your feelings on this?

  7. There was a lot of this information that was so far over my head but I still get the gist. The models failed to pick up the combination of lift and vorticity advection. What an interesting job you have! Thanks for sharing.

  8. Nice job on this report. Hopefully this sort of post-storm analysis will help the improve modeling equations to the extent possible, given the data limitations.

  9. Hi Ross,
    I completely understood that there was no way to predict what happened with this storm given the tools at your disposal. That being said, I’m very impressed with the amount of study you’ve devoted to what went wrong. I don’t know for sure, but I suspect a lot of meteorologists would just throw their hands up and say, “Well, we missed that one!” and go on about their business as usual. We are fortunate to have professionals like you in our corner who would take the time to break down in minute detail what happened after the fact in an effort to provide better forecasts in the future. I’m suprised more readers of the blog have not picked up on this. Maybe everybody is just “blogged out”, so to speak, after all the storms we’ve had, but I certainly appreciate your diligence, and I’m sure others do as well. Keep up the great work!
    Best regards,

    • Neil,

      Thanks for the comment! We are very passionate about understanding “why” the atmosphere is working the way that it is. I can’t think of many meteorological events that are more fascinating than this one for so many reasons. I am just glad that I had the opportunity to observe and study it. Take care!

      Ross Ellet

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: