1Case Study of the March 29, 1998 Tornado Outbreak in Southern Minnesota: Early Onset KATHARINE J. HORST Department of Atmospheric and Oceanic Sciences, University of Wisconsin-Madison Manuscript in Final Form 11 May 2007 ABSTRACT On March 29, 1998 a series of tornadoes broke out over Southern Minnesota, resulting in catastrophic destruction in the small towns of Comfrey and St. Peter. The outbreak was only the seventh to ever occur in the month of March in Minnesota. The following case study is an attempt to understand the synoptic and mesoscale forcings that occurred that resulted in the development of supercells on this day. A brief discussion of Minnesota climatology is included, as well as a synoptic analysis that leans heavily toward the influence of a low pressure system in South Dakota and accompanying warm front, and concludes with a mesoscale analysis that focuses primarily on the influences of a lower level jet in convective instability as well as vertical shear in the development of the supercell structure. ________________________________ I. Introduction In the afternoon and early evening of March 29, 1998 a series of tornadoes broke out in southern Minnesota, devastating the communities of St. Peter and Comfrey. In just over three hours, 14 tornadoes were reported, including one F3 and one F4, resulting in roughly $230 million in damage. Two people lost their lives and at least 38 people were injured in one of the earliest outbreaks the state of Minnesota has ever seen. Debris from the F3 that struck St. Peter, causing significant damage to the Gustavus Adolphus campus, was recovered in the Twin Cities over 60 miles away. In a town of only around 10,000 residents, more than 500 homes were destroyed and another 1,700 were damaged. The following case study will be an attempt to understand the events that took place on this day that resulted in the production of such violent supercells. It will also look into possible reasons for the timing of these tornadoes; in the state of Minnesota prior to March 1998, only six tornadoes had ever occurred in that month, and the earliest known tornado to occur in the month of March took place on the 18th in 1968. Such significant storms this early in the season are not typical by any means, and an investigation into reasons for this early outbreak will be uncovered throughout this case study. It is hypothesized that all of the typical forcings, both on the synoptic and mesoscale levels, were in place for the development for severe weather and simply that unusually high temperatures and moisture availability on this particular day aided in the early onset of the severe weather season in this year. A description of the climatology of Minnesota will be included to gain an understanding of the2standard conditions that occur here on average. A section on supercell development will provide insight into the conditions necessary for such occurrences and the outbreak will be analyzed according to these conditions. Finally, the synoptic and mesoscale forcings will be discussed in an attempt to gain a full understanding of this case. II. Data The data used in this case was provided by the University of Wisconsin-Madison’s Department of Atmospheric and Oceanic Sciences. The data was analyzed using the GEMPAK and GARP interfaces, programs designed for the simple production of presentation quality visuals. The University of Wyoming’s Department of Engineering website provided the soundings and sounding data used in the production of the cross sections. Some of the conceptual models and diagrams were provided by papers and lectures notes and will be referenced accordingly throughout the case study. The data provided for this case included the 00Z data on the 29th as well as the 12Z data on the 29th, and the twelve hour forecast of the 00Z and the zero hour and 12 hour forecast of the 12Z data were primarily used in the analysis of this case. III. Synoptic Overview IIIa. Climatology of Minnesota Overall, the state of Minnesota is characterized by huge changes in temperature from season to season with temperatures capable of reaching as low as -60F (-51C) in the winter months and as high as 114F (45.5C) in the summer months. On average, temperatures range from around 13F to 73F throughout the year. In March in the southern portion of the state, temperatures average around 32F, increasing to about 47F by April. Southern Minnesota in particular is known for the high levels of humidity that accompany warm temperatures in the summer, with dew points capable of reaching as high as 70-80F (21-27C). In the spring when temperatures begin to increase, storm systems can develop as low pressures move in from the West and collide with warm, moist air that has advected from the Gulf of Mexico. In the summer months in fact, winds predominantly come from the south and southeast, providing the southern portion of the state with its high levels of moisture. Thunderstorm activity typical persists for 30-40 days out of the year, from the date of onset. Tornadoes in particular can occur anywhere from March to November, but typically occur in June, followed by July, May, and August. The southern portion of the state favors tornadic activity as it is in the northern corner of tornado alley. As will be highlighted in this analysis, temperatures at the end of March 1998 in southern Minnesota were much higher than the average, perhaps contributing to the early onset of the outbreak this particular season (MN DNR 2004). IIIb. Outbreak Overview In the 1950’s it was recognized by Fawbush and Miller that severe weather seemed to be consistently associated3with the location of mid to low level jet streaks, the location of convergence between moist and dry air and the extent to which moisture varied over the convergence zone (Miller 1972). They then developed five scenarios in which most severe storms could be categorized using these criteria, which allowed them to easily identify regions in which tornadic activity was most likely to occur. Figure 1 shows the Miller Diagram associated with this particular event and highlights the region where severe weather would be likely to occur according to these variables. It seems reasonable to begin an investigation of this sort with a Miller Diagram, as it shows a big picture of all the most important variables that interacted in the development of the storm. Throughout