Use the links below to follow through a guided analysis of the storm from start to finish. This is a graphics intensive review and some Internet connection systems will require more time to download the images.
The blizzard of 1998 is extraordinary in two areas; the duration of the event and the total amount of snow that fell in the northern Black Hills. From a meteorological perspective, it is the role that the terrain played in production of the heavy snow, and the location of the snow that is most interesting. Click on upslope for a detailed discussion of terrain features and their relationship to precipitation in the Black Hills.
This analysis begins with a synoptic discussion as seen by the eyes of the forecasters that were working the front line during the event. I hope you find this review an interesting exploration of terrain interaction and winter precipitation.
Synoptic Review and Discussion
The overall northern hemisphere pattern for the 1997/98 winter season featured a split polar jet. The arctic front and subsequent airmasses remained locked in northern Canada by a strong ridge anchored over the Rocky mountains. At the same time, pacific storm systems plunged into southern California periodically undercutting the ridge. One such low geopotential slid under the ridge beginning on February 23rd and, by the 25th, began intensifying over the northern plains.
This review begins with a look at the initial model objective analysis at 980225 00z 500mb. The date-time format is YYMMDD HHZ. A strong short wave trough extended across the Rocky mountains, undercutting the mean ridge over the northern Rockies. Thirty-six hours later this wave had become a closed low over the northern plains. See the following 980226 12Z products: 500mb, 850mb A & B, 900-700mb Orographic Omega.
Between 12 UTC on the 25th and 12 UTC on the 26th, explosive cyclogenesis occurred over the plains from northeastern Colorado to south central South Dakota. A set of surface maps (980225 12z, 980225 21z, 980226 00z, 980226 06z, 980226 18z, 980227 06z, 980227 12z, 980228 00z, 980228 15z, 980301 00z, 980301 15z, 980302 00z) reveal the evolution of the initial low pressure center. This low tracked northeastward across South Dakota during the night on February 25-26. The spiral bands show up well on the Rapid City Doppler radar (980225 1930z, 980225 2105z, 980225 2210z). The snow combined with northwest winds up to 50 mph and caused blizzard conditions with near zero visibility. A tracked rescue vehicle was used by Pennington County Search and Rescue to bring stranded motorists just east of New Underwood on Interstate 90 to safety. Rapid City resident Gerry Broer returning from Sioux Falls was among those that spent the night at New Underwood. "It was pretty bad driving for awhile. And then a semi jack-knifed causing the interstate to be blocked, that is when they came to get us."
The spiral bands moved into the northern Black Hills to cause the initial bursts of heavy snow and upslope. By 980225 2210z, radar revealed that these same bands were dissipating downstream of the hills and along the eastern slopes. Rapid City received an initial burst of snow as these bands passed through. Then they dissipated as the downsloping winds overrode the dynamic and convective mechanisms responsible for their inception.
The upslope and downslope pattern set the pace for the next four days as the upper level system remained closed and locked in place over the northern plains. This is revealed in the following set of images showing the 500 mb low and its lack of eastward movement. The 850 mb maps indicate the strong northerly flow and areas of moisture that wrapped around the system to inundate the northern Black Hills. And then the atmospheric profile viewed by the soundings offers an evaluation for the Froude Number.
February 26th to March 2nd Radar Images
The following images were taken from the Rapid City Doppler radar. They begin on February 26 and go through March 2nd. The first three were taken in precipitation mode VCP 21, while the last three are in clear air mode VCP 32.
Base reflectivity for February 26, 03:03 UTC. Notice the absence of returns along the eastern portions of the Black Hills, from Rapid City to Hot springs, and then another increase further south. This is due to a combination of upslope, then downslope, and then lee side convergence and wake turbulence.
Base reflectivity February 27 at 14:39 UTC. Again, notice upslope and lee side convergence over pine ridge.
Base reflectivity for February 28, 00:19 UTC. Not much change, except for the appearance of wake rotors to the east of the black hills.
Radar in clear air mode, VCP 32. Base reflectivity March 1, 09:31 UTC. Enhancement over the plains due to wrap around from the northwest. This added another 18 inches in 6 hours to the northern Black Hills.
Base reflectivity March 2, 03:54 UTC. Returns have decreased to a few flurries and it looks like it is over.
Northern Black Hills gets one last round, March 2, 08:09 UTC.
The graphic above shows the total snowfall received from the storm from February 25th to March 2nd.
The above link leads to a pictorial storm survey of the Black Hills taken on March 1, 1998. The map begins 5 miles west of Rapid City where very little snow was evident. Just 20 miles further into the hills and a dramatic change has occurred with over 20 inches of snow in "Down Town" Rochford. The survey turns north into Lead where the official storm total was 103 inches for the 5 day period. Then we turn through Spearfish Canyon into the town of Spearfish. The upslope mechanisms are at work and are visible in picture 13, viewing south towards the northern Black Hills. The trip ends in Sturgis where Harley enthusiasts are bound and determined to just hibernate.
Summary and Conclusion
Model forecasts and products were dependable during the blizzard and gave accurate predictions of the upcoming and prolonged event. As a result, winter storm watches were posted over the region 36 hours before it began. These were refined into the appropriate warnings describing in detail the threat to life and property.
This blizzard event had weather conditions that warranted an extended period of winter weather warnings and advisories. The duration of the blizzard conditions over northwestern South Dakota and northeastern Wyoming prevented many residents from restocking critical supplies. In the northern Black Hills, search and rescue teams used snowcats to reach stranded residents with water, food, clothing and heating fuel. Over the open plains, the prolonged blizzard conditions became an issue for ranchers and travelers as roads were blocked and impassible. Wind Chills throughout the event were also of concern with readings ranging from 20 to 50 degrees F. below zero for 5 days.
A review of the forecast products and warning statements revealed an outstanding performance was accomplished by the Rapid City staff in keeping the public informed and up to date with the latest conditions. When prolonged blizzard events occur, it becomes most challenging to keep forecasts and warnings consistent between the issuance times. It was decided early on that it would be easier to leave warnings in place regardless of brief improvements in conditions...therefore minimizing the public's sense of confusion and presenting overall consistency along with accuracy.
From a meteorological perspective, the interaction of terrain features with the synoptic weather patterns became a predominant feature in precipitation formation and enhancement. This highlights one of the main forecast dilemmas that most operational forecasters are challenged with routinely, and one of numerical weather prediction's greatest hurdles.