September-October 1998 STORMTRACK features the Spencer, SD Tornado


I. COMMENTARY by Tim Marshall

What an active fall season! Lots of hurricanes in both the Atlantic and Pacific oceans and tornadic supercells on the plains. This sure has kept me busy assessing damage. By far, the highlight of the fall season for me was the tornadoes near Watonga and Kingfisher, Oklahoma on October 4th. Never before have I filmed October tornadoes and these were some real beauties. But getting there at the right time and place presents the real challenge to tornado chasers. There are several key decisions to be made. You also have to be lucky.

The first decision was whether or not to chase. This was your typical upper low event coming through the southern Rockies with lee-side cyclogenesis and good dry punch. Negatives were lots of cloud cover in the warm, moist sector, marginal surface temperatures, and climatology to name a few. But, there were more pros than cons, and it was the weekend. So, I decided to chase. I was in Minneapolis on storm duty at the time and flew back to Dallas on the night of October 3rd.

The second decision was where to chase. A plot of the morning maps revealed to me that severe storms were most likely from western Kansas down to northern Texas. Dynamics were best to the north and thermodynamics were best to the south. However, surface temperatures were cooler in Kansas and the cap was still strong over Texas. Thus, Oklahoma looked best to me. Further analysis indicated that storms would likely form along an axis from Woodward to Altus and move to I-35 by sunset. Carson and I headed up I-35 stopping briefly on the Arbuckle Mountains for an update from THE WEATHER CHANNEL. A severe storm already had developed near Woodward and was moving east. I anticipated other storms would develop to the south later on and picked El Reno, Oklahoma as the target town.

The third decision was to pick which storm to chase. When we arrived in El Reno, the Woodward storm was now to our distant north but other storms were developing to our west and southwest. I never considered chasing the Woodward storm since it was quite a distance away, the radar showed it was a big, HP, and I prefer new convection over old. Carson and I decided to be patient. For a time, it looked as if the storms to our west and southwest would develop into a squall line as anvils merged into one big cluster of gunge. However, the storms soon broke apart and now it was time to pick one. A severe thunderstorm warning was issued for a storm to just our west which made our decision easy on which storm to pick. However, this storm had difficulty in getting itís act together and eventually gave up on it. A check of the radar revealed almost a dozen storms to choose from. Meanwhile, a new storm had developed to our west on the outflow boundary of the previous storm. Ah, new convection on a boundary, great! And great it was. The storm developed southwest of Watonga and got itís act together quickly. Soon, a crisp anvil emerged with vertical updraft, mid-level banding, and a lowering at cloud base that was far removed from the visible precipitation area. Ah, what picture perfect structure! I also liked that this storm appeared to be the farthest storm west with itís tail tucked into the clear, dry air.

The fourth decision was where to position ourselves for intercept. It is our goal to film tornadoes with the best contrast with the minimum amount of drive time while the tornado is in progress. The storm was moving northeast at around 30 mph, but such storms can turn more to the right if they become tornadic. At the same time, the paved road network was sparse. So, Carson and I decided to park at a paved road intersection exactly eight miles east of Watonga. The rain free base was to our southwest and closed rapidly. Soon, a tornado developed three miles southwest of Watonga, crossed the road five miles in front of us, then roped out two miles to our north. We filmed the tornado from one location and it was high contrast. Next, we had to position ourselves for the next intercept as a new wall cloud developed to the northeast and was moving quickly away from us. Staying with the storm was difficult due to increased traffic, road construction, and stop signs/lights in the town of Kingfisher. But, we managed to get around these difficulties and watch the second tornado cross the road just to our north. Finally, decisions and luck combined to yield a great chase.


The 3rd Annual 1999 Severe Storms and Doppler Radar Conference will be held March 26-28, 1999 at the University Park Holiday Inn in Des Moines, IA. The conference begins at 1:30 pm on the 26th and runs until noon on the 28th. This conference is geared towards anyone with an interest in weather. Spotters, forecasters, emergency managers, storm chasersÖeveryone is encouraged to attend! Friday evening is chaser night, so bring your videos or pictures to show. The conference will be held at a different place this year: The University Park Holiday Inn. There is a free shuttle from the Des Moines International Airport so it will not be necessary to rent a car. There are plenty of fine eating establishments within walking distance of the hotel. There are also a few fast food places for those people on a budget. The University Park Holiday Inn has a large atrium area so you can easily mingle with the hundreds of other weather-minded people at the conference. Plus it has a waterfall and lush greenery! Rooms at this four star hotel are $68 and can hold up to four people. They come furnished with coffee makers and even data ports. For more information about the logistics of the meeting, mail me (Paul Vincent Craven) at


Dr. Tim Doggett at Texas Tech University announces: " The long awaited decision regarding funding of the West Texas Mesonet Project has finally been made. Yesterday afternoon the Texas Department of Economic Development approved a $2 million grant to fund the Wind Science and Engineering Technology Assistance and Transfer Program. This money will be dedicated to develop a 28-site mesonet in the Texas High Plains region, promote economic benefits stemming from the use of this data resource, and develop a variety of applications to utilize the data. The planned mesonet will include 28 surface stations, 3 atmospheric profilers, and data from a 200-meter tower at Reese Center. Data will be made available across the internet in as near real-time as possible, and we anticipate having the system at least partially in place by this coming spring. We hope that this is a large step toward the implementation of a statewide mesonet called the Texas MesoNet Project. A collaborative effort between Texas Tech and Texas A&M Universities is currently underway to establish this network. This (and possibly other) pilot projects will help to prove the feasibility of the statewide project, as well as demonstrate the economic benefits that can be expected from such a system. I have tried to send this message out to as many people as I could think of who might be interested in this project. Please feel free to pass this information along to anyone who I might have missed. Also, I would be glad to field any questions about the project. Further information about the Texas MesoNet Project can be found at: A west Texas Mesonet Homepage is coming soon! (http://www.cicms.atmo.ttu) e-mail:

Steve Miller writes: "On 5/26/98, an incredible tornadic supercell formed and remained nearly stationary in Tom Green County, Texas where San Angelo is at. This storm obtained an incredible state of outflow/inflow balance while remaining stationary. Tornado warnings were constant for nearly four hours for Tom Green county with about two hours having the TVS right over or within a few miles of San Angelo. Tornadoes and TVS's (Tornado Vortex Signatures) only moved at about 5 mph....10 mph at most. The radar loops were fascinating as other storms in the area moved northeast at 10-15 mph while this storm remained stationary and spinning like a top. The storm slowly propagated southeast into the next county and persisted for another couple of hours as a tornadic supercell until it finally weakened. More interestingly on a synoptic scale, a convectively induced mid level vorticity max with closed circulation developed as a result of this storm (in my opinion). The vorticity maximum transversed northeast across North Texas producing other severe storms and into the Ark-la-tex region where it spinned and remained stationary for about 2 days producing other severe storms and flooding. I must also point out that the setup for this storm had similarities to Jarrell. Upper level winds were weak and not typically favorable for tornadoes. The storm formed along a very well defined and strong outflow boundary from the previous night's convection that had become stationary along a northwest/southeast axis with weak upper winds out of the southwest. The low-level shear was great though with east-northeast winds to the north of the boundary at about 10-15 mph and southeast at 15-20 mph to the south of it. The CAPE was also very high that day exceeding 5000 across SW Texas and above 6000 in the San Angelo area according to ARPS data. There was also a pretty decent cap that day along with abundant sunshine allowing things to really cook prior to storm formation." Steve Miller email: txt@GTE.NET

Robert Gallimore writes: " I have enjoyed reading your articles for many years. Recently, though, I was disturbed by some comments you made in your commentary on Storm Track's web page. In that commentary, you stated that ĎI remain concerned about the number of people becoming involved in this sport and often stress education and safety.í While the subject of education is undoubtedly very important, the idea that chasing as a sport is what angers me. One aspect of chasing that is rarely discussed is the fact that people may be losing their life or having their life forever changed by the vicious wind. I am a tornado survivor. I was quite upset in learning that people were actually wishing the tornado I survived would grow more ferocious, and that they could get a better view of it so they could make money selling their video. I realize you probably get more than your share of letters from upset people regarding storm chasing. But I believe it is time to stop and think about the fact that a tornado or its parent supercell is shear terror for those having to survive it. And the sad fact that people are either having their lives destroyed or taken away from them. It is not a sport to watch with excitement the destruction of peoples lives. And truthfully, it is inhumane to hope that a supercell will cause a tornado so you can get an adrenaline rush and make money at the same time. Next time you go chasing, you should stop and think about the people in Spencer, S.D., Andover, Kans., Catoosa, Okla., and the countless other towns destroyed by a ferocious storm. And you should ask yourself, "Would I appreciate people gaining satisfaction from that which has destroyed my life?" Email:

EDITORíS RESPONSE: I appreciate your comments and I know how you feel. My town was hit by a tornado when I was 11 years old. The damage was incredible. Some of my friends didn't come back to school the next day. I was saddened by this experience and my curiosity of this horrific thing from the sky that took my friends away led me to study meteorology. I have always looked at the sky with respect and my studies are scientific. However, there is another side to this story. If you have chase in Oklahoma you would know what I mean about how chasing has turned into a sport. TV media have led the charge in competing against other stations to see who can get to the storm first and get the best footage/storm accounts. This in turns leads to better ratings and gloating rights thanks to our disaster hungry public. When the media talks to me, they always get around to asking me the NUMBER of tornadoes I have witnessed, much like the number of home runs of baseball greats Mr. Sosa or Mr McQuire. I too am bothered by chasing being perceived as a sport, but unfortunately, that is what it has become.


I. Introduction

My interest in severe storms and tornadoes spans several decades. The area where I live in Central Texas is roughly 175 miles inland from the Gulf of Mexico, and like most Texans, I grew up believing that hurricanes and tropical storms were a concern only for those living or near the coast. That view was shaken, at least a little, in September of 1961 when Hurricane Carla moved inland at Matagorda Bay (between Houston and Victoria) and then continued northwestward, crossing Central Texas on the morning of September 12th. Carla still had a radar-discernible eye as it moved from Austin to just west of Waco. As it passed, the winds dropped to near calm, the skies cleared enough that I could see the sun's orb through multi-layered clouds, and the squally rain slacked-off to a sprinkle. Within 20 minutes, the rains returned and the wind picked-up from the south-west. It was an intriguing experience for a young person (I was 14 at the time) who was already "in love" with the weather.

Then, in 1980, Hurricane Allen made landfall in far south Texas at Port Mansfield, north of Brownsville. Allen moved northwestward up the Rio Grande and dissipated over the higher terrain and mountains of northern Coahuila and Chihuahua. As Allen crossed the Texas coastline over 250 miles away, its outer rainbands spawned an outbreak of tornadoes from near Victoria northwest to Austin and west to San Antonio. At least 29 tornadoes occurred, most in South Central Texas. Many people who had fled coastal Texas for the relative safety of evacuation centers in Central Texas found themselves staring out windows and doors at tornadoes approaching from the east. Fortunately, no serious injuries occurred, but the airport at San Marcos suffered major damage from one of the storms, later surveyed by Dr. Fujita.

More recently, as Hurricane Jerry moved northward up IH45 from Houston toward Dallas in the Fall of 1989, I witnessed a T/C tornado. We were under the rain shield but on the western side of the storm's center. Cell motion was cyclonic, with cells in Central Texas moving generally north to south. Around 6pm CDT, a tornado developed from a cyclonic swirl I'd been watching in the stratus cloud deck. The tornado was white, and about 6 miles southwest of my home. It remained on the ground for around 5 minutes, then dissipated. No damage or injuries occurred.

I was hooked. I decided then and there that I had to learn all that I could about T/C tornadoes. My particular focus has become whether we can do a better job of predicting where they will occur. And incidental to this inquiry is the issue of whether they can be "chased" in the traditional sense. Some land falling tropical systems are prodigious tornado producers. The table below is adapted from a 1996 N.W.A. Digest article(1) , and reflects the most prolific of such storms through 1996.


Tropical Number of

Cyclone Year Tornadoes

Beulah 1967 113

Andrew 1992 48

Gilbert 1989 47

Danny 1985 39

Beryl 1994 37


II. Past and Current Research

Efforts to understand how and why land falling tropical cyclones produce tornadoes have been on-going for at least the past 35 years. The pace of research and learning has quickened considerably with the completion of the national WSR-88D network, but much of the earlier work continues valid today and has been verified in many respects.

We now know that there are two preferred areas of tornadogenesis in land falling tropical cyclones: the eyewall of the storm and in the outer rainbands (also called "spiral bands"). The study of the eyewall tornadoes has been hampered by the difficulty in distinguishing tornado damage from general hurricane damage, which is usually maximized near the eye of the storm. This was true even when Hurricane Andrew passed virtually over the Miami WSR-88D site in 1994, although one study suggests that cyclonic swirls seen in by the WSR-88D in the eyewall may represent tornadic circulationís. It remains to be seen whether the WSR-88D network will afford opportunities to resolve the damage paths of tornadic cells within the eyewall. It is becoming clear, however, that the WSR-88Ds are very helpful in the case of outer rainband tornadoes and their parent circulationís, which are now thought to be mini-supercells imbedded in the stratiform rain areas of the outer bands.

Several different authors have examined the data sets and concluded that most (but by no means all) outer rainband tornadoes occur in the "right front" quadrant of the storm. There has been some questions of how to define the "right front" quadrant, i.e. in relation to true north or relative to storm motion. The most recent literature suggests that defining the quadrant relative to true north gives a better concentration of the tornado locations. One researcher writing in 1982 actually computed the centroid of T/C tornadogenesis for all cases reported prior to that time, and found that the centroid was approximately 50 degrees east of true north and 150 miles from the eye of the storm (050 at 150). He found that 95% of all T/C tornadoes occur within an area defined as being true north (0 deg) clockwise through 120 degrees east of true north, at a range of 60 to 240 miles. Refer to Figure 1.

As in the case of Great Plains tornadoes, researchers have been studying the antecedent conditions by examining atmospheric soundings taken in "close proximity" to the tornadic cells. This has proven even more challenging than in the case of Great Plains storms, because the environmental conditions associated with land falling tropical systems are harsh, even outside the tornadic cells. An additional problem in many instances is the absence of sounding data from the upwind side of the system because the 'upwind' area is over the ocean itself. The composite "proximity" soundings which have been done reflect relatively low buoyancy, highly sheared environments. This is consistent with what has been found in general tropical cyclone environmental studies. Refer to Figure 2.

Most of the T/C tornadoes are found to be related to mesocyclonic cells within the rainbands. These cells exhibit many properties similar to their Great Plains cousins, but also have important differences. One of the similarities is storm structure, and one of the most important differences is size or scale. In general, the T/C mesocyclones and the cells in which they develop are quite similar in structure but considerably smaller in scale than is found in mesocyclonic storms occurring on the Great Plains. Not only are the cells themselves smaller, but the mesocyclones also tend to be smaller in scale. These reduced dimensions increases the difficulty in identifying the circulationís, even with WSR-88D radar, especially with the current (as of this writing) WSR-88D algorithms. In addition, electrification is extremely variable within the rainband cells, so lightning may not occur at all, even with cells which are producing tornadoes. At least one writer thinks that this is due to the tropical nature of the environment and the relative weakness of vertical motion above the freezing level, which is often ~15,500 ft m.s.l.

There has been considerable speculation about the source of vorticity in T/C mesocyclones. Current theory holds that the rear-flank downdraft (RFD) is a critical element in concentrating vorticity leading to tornadogenesis in Great Plains supercells. Researchers have been unable to detect rear flank downdrafts in most T/C cells which become tornadic, so current theory holds that strong low-level shear (associated with the very energized wind field in tropical cyclones) generates substantial streamwise vorticity, which is then stretched vertically by developing updrafts in the rainband cells. If you are accustomed to looking at wind profiles and hodographs on the Great Plains, looking at similar graphs in tropical situations will take some adjustment, because there is minimal directional shear, with most of the shear coming from the very energized velocity fields associated with the tropical cyclone.

Likewise, skew-T charts are of little help in most T/C situations. By Great Plains standards, buoyancy is meager. It is not uncommon to find CAPE well under 1000 and Lifted Index values around neutral (i.e. 0 to -2 ). The depression of these indicators is a product of saturation through most of the atmospheric column and the tropical nature of the air mass. However, a couple of recent studies have suggested the need to examine upper air prognostic charts closely for the possible entrainment of much drier mid-level air into the circulation, often from the northeast. The entrainment of drier air can rapidly increase the instability (creating a pattern more similar to Great Plains environments) which may lead to increased tornadogenesis, even with relatively weak systems. The evolution of this type of pattern is most likely as the storm re-curves, or after it curves when the storm is moving in a generally northerly track, and may account for the tornado outbreak associated with Tropical Storm Beryl (1993), which made landfall near Panama City, FL. No tornadoes occurred on that day of landfall, but as the remnants of Beryl moved northward through Georgia and the Carolinas the following day, 31 tornadoes occurred.


III. "Chasing" T/C Tornadoes

So on to the question of whether one can "chase" T/C tornadoes. And here's the answer: "no". At least, not in the traditional sense. However, there may be a way to place oneself in a position to 'intercept' the portion of a land falling storm which has the highest potential to produce tornadoes. Through all of the following discussion, please keep in mind that, as with chasing in general, there are risks to property and life involved, and only you can decide whether the risks are worth taking in any particular situation.

Assuming that you've made the decision to try, here's my suggested plan of attack. As a tropical cyclone approaches the U.S. coastline, particularly the Gulf coast from Texas to Florida, monitor the satellite imagery for indications of the storm's intensity and for the presence of one or more outer rainbands. Some tropical cyclones are very compact, and even though intense, may not produce the classic spiral-shaped outer rainbands. There are many examples one could cite, but Hurricane Celia comes to my mind as it was nearing the Texas coast in 1970. Although rated a Category III hurricane at the time, it had a relative absence of outer rainbands.

A storm which has the appropriate rainband structure, should be monitored as to the projected path as forecast by the Tropical Prediction Center (TPC), which issues projected position information at 12-hour intervals every six hours out to 36 hours. Plot these positions on an accurate chart, then plot the offset "centroid" position mentioned previously in Sec. II (050 at 150). An easy way to do this is to make a clear plastic overlay showing the relationship of the eye of the storm and the "centroid" with true north relative of the eye also marked. Be sure to match the scale of the overlay to the scale of the map you're using. With this, you will be able to plot the 'path' of the "centroid" at the storm makes landfall.

Then, to allow yourself some leeway, identify a swath roughly 30 miles either side of the centroid path, which will give you a sixty mile wide zone which should, assuming the TPC projected path verifies, "capture" most of the outer rainband tornadogenesis in the period covered by the TPC projection. If you wish, you can subdivide the 12-hour projections into 6-hour or 3-hour projections, but be careful of changes in the speed of the storm. Storms making landfall and/or re-curving (i.e. turning northward) often accelerate in forward motion. Obviously, storms deviating substantially from the projected path will result in "centroid" deviations, too. Refer to Figures 3 and 4.

Now it's time to hit the road for a position near the centroid path. Where alternatives exist, choose locations which afford (1) daylight conditions, (2) roadways not likely to flood, and (3) areas with better road networks, allowing rapid travel in as many directions as possible. If at all possible, you will want a way to obtain radar updates while on the road. If you don't have access to real time data (a la Carson Eads and Tim Marshall), think about having a friend in the non-affected area who you can call by cellular phone for ~20 minute updates. And here's what you or the friend furnishing the updates should be looking for in the radar data: discrete cells in the outer rainband(s) within the 60-mile wide centroid swath which: (1) have persistently (~20 mins or more) higher reflectivity than surrounding cells; and (2) show signs of developing rotation. The latter can be based on either velocity data, if available, or inferred from shape (e.g. kidney bean) of the cells.

Outer rainband cells in T/C situations which are likely to produce tornadoes contain small-scale mesocyclones, which although smaller in dimension (both width and depth) than on the Great Plains, exhibit shape and velocity signatures which correlate well with plains storms. Because of the smaller dimensions, however, range from the radar begins to impact velocity data closer to the radar than in larger storms, and if you have it (most don't), spectrum width data (a WSR-88D product not generally available to non-NWS users) can be helpful in identifying mesocyclonic cells beyond the range of the usual velocity (base and storm-relative motion) products. Tornadoes, if they occur, are most likely on the upwind side of these cells. Remember that these cells will likely be moving rapidly because of the energized wind fields associated with the storm.


Here are some other considerations: (1) Once a persistent outer rainband cell becomes mesocyclonic, watch for other cells downwind (!) of the initial cell to also develop mesocyclones. It isn't completely clear why this happens, but this helps explain the occurrence of "swarms" of tornadoes sometime associated with outer rainbands. (2) The absence of well-developed rainbands is a negative for focused tornado outbreaks during landfall. but this does not necessarily preclude the development of a remnant tornado outbreak on the day following landfall. (3) Remnant outbreaks are likely if the wind field stays energized and significant dry air intrusion is forecast or occurring in mid-levels, especially in the eastern quadrant of the storm. An additional positive would be the development of areas of enhanced insolation where the skies become partly cloudy during daylight hours. (4) There is some indication that even eastern Pacific storm remnants moving northeastward across Texas and Oklahoma can enhance the wind fields enough to produce at least a threat of tornadoes, but the threat does not appear to be as focused as in other cases.


(1) Vescio, Weiss & Ostby, 1996, Tornadoes Associated with Tropical Storm Beryl, NWA Digest, Vol. 1, No. 21, 2-10.


I waited for a short distance chase opportunity and Friday, May 8th could not have been much better. Storms were forecast to break out west of Dallas-Fort Worth ahead of a well developed trough with an associated surface low and dryline. By 1:30pm, I was heading west toward Decatur on Highway 380 to rendezvous with the dryline.


By 4:00pm, I started to worry that the storms would stay too far west to see before dark. The veil of cirrostratus to the west looked like blow-off from very distant storms. I stopped in Graham, Texas to download a radar picture. There were a few storms well to the northwest. However, I decided to move southwest hoping that something would develop and more toward my position. Finally, as I drove south to Brekenridge on Highway 67, I began to see some small turrets peeking through gaps in the cirrus. At about 5:30pm, I pulled into a McDonalds in Brekenridge to download another radar picture. Just when I started losing hope of seeing anything more than a layer of cirrostratus, Craig Green pulled up beside me. I had never met Craig, but I welcomed the opportunity to partner with a knowledgeable spotter and chaser. He pulled out his map, and we studied a radar picture. There was a strong cell in Haskell County. We decided to head north on Rt. 183 to intercept the cell near Throckmorton. By about 6:10pm, the rain free base was visible to our north. At about 6:20pm, we pulled over about six miles south of Throckmorton. There was a group of chasers also watching the tail-end charlie storm. We watched a small, poorly organized area of rotation at the cloud base. A downburst produced a large dust foot to our southwest around 6:30pm, and gust front. As the storm outflow started to bear down on us, the whole chase group fled back south on Highway 183. Driving south, we caught the leading edge of the rear flank downdraft with strong crosswinds, rain, and sporadic golf ball-size hail. Although we did not see a tornado, this storm had an attitude, and we didnít escape the fury without some battle scars. We later learned that a trailing chaser had been hit with baseball-size hail along our outbound route. We retreated south and east, eventually making our way back to Graham. There was some rotation evident in the cloud base, but nothing ever became well organized. We eventually moved south to Interstate 20 and returned to Dallas skirting the storms to our north and west.

Many factors favoring severe storms were present in this situation including strong dynamics. Using Jon Daviesí article from the March-April 1995 issue of STORMTRACK and some approximations from soundings, I estimated the storm relative helicities (SRH) at 660 m2/s2. Estimating the convective available potential energy (CAPE) graphically from a Skew-T log-P chart, I got 1515 j/kg. This computed to an energy-helicity index (EHI) value of 6.2, a respectable number, even if itís off by a factor of two. At 7:00pm (00z), 130 knot jet stream winds were recorded over Fort Worth and 60 knots at 500mb with strong diffluence between Fort Worth and Oklahoma City.

In retrospect, I remembered Craigís comments about the lack of low-level moisture. This may have been one factor that prevented a tornado outbreak. Dewpoints were only in the mid to upper 50ís (F) at 4:00pm. Jon Davies may have explained this best in his discussion of tornadoes and the lifting condensation level (LCL) in the May-June 1998 STORMTRACK. The 4:00pm, (21z) surface temperature-dewpoint spread (T-Td) at Fort Worth was 31 degrees F. This gives an LCL of 1740 meters above ground level, very similar in this respect to the April 24, 1997 Midland case cited by Jon in his article. That storm produced only downbursts, much as this one did.



The first thing I was aware of was the sensation of being in a moving vehicle. For a few brief moments I could not figure out where I was. Am I still on that bus to Kearney? What day is this? Is this a dream? Am I dreaming I'm having a dream? As I continued to recover from my altered state of consciousness, I began to hear the sounds of the wind whistling through a car window, an AM radio, and what sounded like a weather report coming from a scanner. I could also hear a voice, saying, "Cheryl - are you awake back there?"

Then I remembered where I was. It was May 22, 1998. I was riding in the back of veteran storm chaser Bill Reid's Nissan Pathfinder, and we were somewhere in Kansas. Today was the first day of my very first chase trip. It had taken years of dreaming, a few months of planning, a two hour plane ride from Calgary to Denver, and an all-night bus ride to Kearney, Nebraska, where I was met by Bill about an hour after sunrise. Up to that point I'd managed to get about two hours sleep in the past 30, give or take. Shortly after meeting Bill I attempted to doze in a sleeping bag on the floor of Bill and veteran chaser Keith Brown's motel room, all to the comforting drone of the Weather Channel. I had made it. My first storm chase adventure was underway.

My first chase day officially began a couple of hours later. Right before checking out of the room, Bill had been busy at his laptop downloading weather data and checking the outlooks for the day. There were two areas to choose from - northeast Colorado and central Kansas, where it was thought that the potential for severe storms was greater. After Bill and Keith analyzed the weather data and came up with a preliminary chase plan, we proceeded to check out of the motel room, grabbed a bite at the local McDonald's, and headed south.

I really don't remember much at all about those first few hours on the road. I was in the back seat of Bill's Pathfinder, surrounded by all kinds of suitcases, camera cases, tripods, etc. There was nothing really to see for a while, and besides, that pillow next to me looked pretty darn inviting, so I proceeded to doze. Every now and then I'd rise from my stupor to look out the window, ask questions, and listen to Bill crack a joke and Keith give a monosyllabic reply.

Eventually we came to Stockton, Kansas where we had to make a decision. Was it east to Central Kansas or west to Colorado? After discussing it between themselves, Bill and Keith decided to go to Colorado because there would be better potential for photography - since there was all that smoke and haze in Kansas. Sounded good to me. So, we headed west. I noticed that as the day went on and we got closer to our target area, Bill and Keith got more businesslike. There were fewer jokes, the radio and scanner were used more often, and at one point they hooked up Bill's portable TV underneath the dashboard. We stopped by the Goodland NWS office, and were advised that if we wanted to see supercells we should probably head back east, into the moderate risk zone. There was however, a storm to the northwest which had produced a lot of hail around Fort Morgan. We thanked them and headed west on Interstate 70, and Bill and Keith got their camcorders ready.

I was feeling pretty awake now and could actually pay attention to the world around me. This part of Colorado was pretty flat and kind of desert-like, and the sky seemed to stretch on forever. I could see lots of cumulus clouds and some high clouds off to the NW. We were beginning to see the anvil from our storm. As we got closer I could see that the anvil had hard edges and was scalloped in places, and I could also see that it was very, very big.

Eventually we came to Joes and headed north on Highway 59. About 12 miles north of Joes we stopped to take a look at the updraft base. The wind going into that storm was stronger than any inflow I'd ever seen, and it howled and whistled and moaned. I had never heard the wind sound like that before, even during an Alberta blizzard. The updraft base looked huge, dark and ominous. Even I could tell that it was rotating. Almost immediately after we stopped a loud tone came from Keith's scanner --- a tornado warning had been issued for our storm.

Bill and Keith wanted to get a closer look. We got into the Pathfinder and crept cautiously north along Highway 59. A dust whirl crossed the road 1 Ĺ miles or so north of us, and shortly after this we were hit by strong west winds, as well as a little bit of rain and hail. (A few days later we returned to this spot and found power poles knocked over right where the dust whirl had been. The winds which had hit us were probably associated with the hook echo from this developing tornado.)


After being hit by these winds we quickly turned around and headed a few miles south to Abarr, where we were in a safe enough spot to stop and get another look. This time inflow winds slammed the car door right into me as I was trying to get out. I took some slides, and Bill and Keith busily aimed their camcorders at this thing and discussed the nature of the rear flank downdraft (RFD). I watched and pointed my camera at a white curtain of hail which moved from left to right towards the updraft base. A minute or two later we were almost knocked over by strong RFD winds. I decided not to bother taking any more pictures at this point --- it was all I could do to stand up. On my way back to the Pathfinder I took one last look in the direction of the storm, and promptly got a small hailstone right in the eye. The storm was beginning to intensify and it was time to go.

We zig-zagged on dirt roads east and south, east and south. We got to Highway 36 and continued to head east into Kansas, passing towns like St. Francis and Bird City. The air to the east was murky and dusty. The car radio, the TV and Keith's scanner were all on at once, and Keith's scanner, which was tuned to the NOAA Weather Radio, was going crazy ---- severe thunderstorm warnings were being issued as this storm headed into Kansas. All the while this storm was getting stronger and stronger, and closer and closer.

By now it wasn't the only storm around, either. It looked like some of its friends had come to join it. A tornado warning had been issued by the Goodland office for the storm directly to our southwest. I could see this storm's wall cloud from my window but visibility wasn't good enough to see whether or not it was producing a tornado. We didn't have the luxury of taking the time to really get a good look either, since our storm was still making its way towards us.

West of Atwood, Bill looked over his shoulder and exclaimed at the storm structure. Since it was safe enough at this point, we were able to get out for a quick look. We gazed in awe at the storm. Above the laminar base a cylindrical updraft towered thousands of feet upward. It was dark, gray brown tinged with pearl gray. The anvil was studded with mammatus which were also highlighted with pearl gray. Lightning crawled across the mammatus-laden anvil and was tinged with purple. White and yellowish CGs were coming out of the base of this storm. It was ominous, menacing, and one of the most beautiful things I had ever seen in my life....and it was heading right towards us. We fled.

Bill had to work hard to battle strong inflow winds and keep ahead of the storm. Tumbleweeds swept across the road and were illuminated by constant flashes of lightning. The mammatus-filled anvil spat marble-sized hail at us from time to time, and we heard reports of baseball-sized hail over Keith's scanner. I saw CGs to my right almost constantly for a while. It was pretty obvious at this point who was in control here, and it sure wasn't us puny humans.

We raced east through towns that were in the path of the storm and watched as people scrambled to get themselves and their cars protected from the large hail that was coming their way. Eventually we made it to Oberlin, Kansas, with about 10-15 minutes to spare before the storm was to hit there. Bill found a car wash and planned to pull into one of the stalls before the storm hit; however, we had to wait in line, because people were washing their cars!! I thought these people were absolutely NUTS --- don't they ever look at the sky?

Eventually one guy with his immaculately washed (not for long) white van pulled out and we were able to pull into the stall and wait for Armageddon. Luckily for us, all we experienced was heavy rain. We breathed a sigh of relief and turned our thoughts to finding a place to stay for the night. By that time it was close to 11 p.m. - we had been chased by this thing for hours.

We were lucky enough to find the last room in town, and finally found some greasy pizza to eat at a local gas station. At around 1 a.m. I was able to crawl into my roll-away bed and relax. However, I couldn't sleep for a long time. Images of that storm kept going through my mind. Although I knew that we'd been in the way of something very powerful and very dangerous, I realized that we had also been in the presence of something very great, and I was lost in the wonder of it all, and grateful to Bill and Keith for their expertise and to Bill for his good driving. I was very content when I finally drifted off to sleep that night..........

And that was just the first day of my first chase trip



Without a doubt, this day was the most unstable day of the year so far in Oklahoma. Convective available potential energies (CAPES) were greater than 5000 j/kg with surface dewpoints well into the 70ís (F). A warm front was located along the Oklahoma-Kansas border with a trailing dryline extending into the Texas panhandle. The main problem today would be where to go. Should I play the triple point in Kansas?, head to the Texas panhandle for a dryline storm?, or try to see a storm through the Mexican haze along the warm front?

Val Castor and I decided Kansas was the best place to be, but it was too far for us to go. So, we decided to head west on I-40 and keep our options open, should we have to go north. We were eyeing a significant area of convergence near and south of Childress, Texas. As we headed west, Mark Hill, another chaser further south, reported towers to his south and southwest. We decided to stop at Clinton, Oklahoma and further evaluate the situation. We heard that a supercell had formed well to our north in Kansas. The storm was heading southeast but it would be a long haul for us to try and intercept it. Still, we decided to head north as the storm was putting out an outflow boundary and we hoped new storms would develop along it. We stopped at waited east of Waynoka, Oklahoma as a cumulus field formed overhead. The cumulus was well defined with vertical development. It appeared we had finally found a pocket of good convergence. After calling the National Weather Service in Norman for information, we were told the boundary extended from near Stillwater to where we were and further west-northwest. Development was slow and I suggested moving back west to try and be close enough to the dry line. As we left, I could see a narrow tower going up rapidly through a break in the clouds. However, it fell apart after about ten minutes. Then, two towers developed, one to our northeast near Alva and the other southeast of us near Fairview. The northern tower was going up rapidly compared to the south. Our minds were made up and we headed towards Alva as the storm exploded right before our eyes. This was definitely a supercell to be. The convection was rock hard. However, the closer to the storm we got, the more obscured it became in the "Mexican haze" -smoke and haze associated with the forest fires advected northward from central America.

We intercepted the storm near Burlington. The storm had strong inflow and a large rain-free base, but little if any precipitation. The storm moved slowly east and we god hit with a strong rear flank downdraft which raised dust and spun up a couple of small, but vigorous gustnadoes. We turned east on Highway 11 near Bryan. A few miles up the road, we noticed a concentrated area of dust to our north beneath the short flanking line. We stopped to check it our and could see it was weakly rotating. The main updraft was further northeast. After about one minute, the circulation tightened into a small dust tube. This was likely a landspout given its location and it dissipated within another minute or so. We continued east about nine miles east of Byron. The updraft to our north was now rotating quite nicely. My brother Kevin had met up with us when I looked over and there it was -- a tornado spinning up in a field about one mile to our northeast. There was a small, stubby funnel aloft with a cloud of rotating red dirt at the surface. We had perfect viewing as the tornadoes moved slowly southeast and were able to stay right behind it as it moved towards Highway 11. The funnel never came more than half way down, but the circulation at the surface was present the entire time. As we got to within 1/4 mile, we could hear the tornado which sounded like a waterfall. Cows were fleeing in panic as the tornado approached. The tornado crossed the road right in front of us. Power lines came down with a flash. The tornado continued southeast moving through a grove of trees. We drove up to the damage path and were able to get around the power lines. Large trees were uprooted and there was a strong smell of vegetation and dirt in the air. As the tornado moved through the trees, large limbs were thrown high into the air and our video shows one large tree coming down. We didnít even notice it at the time, but upon watching our video later, the tornado appears anticyclonic!

The tornado continued southeast through open fields before dissipating after about three or four minutes on the ground. When it ended, it had a landspout look to it with dust tube. The storm base was much larger now and circular. We continued east towards Wakita (where some of the TWISTER movie was filmed) and saw several weak, dusty spin ups only a few hundred yards away. As we stopped to watch one of the spin ups, an inflow jet developed right over us lifting dirt and debris into the air. Winds gusted to near 60 mph. A funnel was seen over Wakita, but no damage occurred in town. We did get hit with one inch diameter hail. We continued east on Highway 11 towards Medford and saw another tornado about one mile to our northeast just as we crested a small hill. There was no condensation funnel, but a large, dusty circulation. We stopped due south of it and suddenly turned toward us as the storm outflow pushed our way. We let it close in to within 200 yards before we hightailed it back west. It quickly broke up into several smaller sub-vortices before dissipating.

SOUTHERN NEW YORK CHASE: May 31, 1998 by Mike Waselnak

On Sunday May 31st, 1998, the Storm Prediction Center (SPC) placed most of New York in a rare "HIGH RISK" area. Over New Yorkís central and southern tier region, severe weather indices (based on modified sounding data) were impressive. Lifted indices were -9, CAPES were over 3000 j/kg, helicities were above 245 m2/s2 and the EHI was 4.86. Strong southerly winds pushed surface dewpoints into the upper 60ís (F) and afternoon surface temperatures reached into the mid 80ís (F). A cold front was forecast to sweep across the area during the evening, but I figured the outflow from a dying squall-line/derecho over western New York would provide a trigger for pre-frontal deep convection during the late afternoon.

At 2:33pm, tornado watch #478 was issued for central and eastern New York. Supercells rapidly formed over western and central New York; upper level winds moved them east at 50 mph. I knew that chasing these storms would be difficult due to the rolling terrain and extensive tree cover, so careful positioning was a better option. My target storm was a large supercell that was moving east along the New York/Pennsylvania border. The National Weather Service issued tornado warnings on this storm as it moved toward Binghamton. I intercepted the storm in Johnson City, just west of Binghamton. The storm had an HP supercell appearance and I encountered torrential rain and pea to half-inch diameter hail as the storm approached. The rain and hail decreased rapidly in intensity only to have the hail become larger, about one inch in diameter. A SKYWARN spotter four miles to my west reported two inch diameter hail. The hail finally stopped as the rain free base and southwest flank of the storm approached. Cloud-to-ground lightning strikes became frequent. A clear slot began to open up the clouds west of my position. I noticed an area of cloud base rotation, but no defined wall cloud. Reporting this to SKYWARN, I waited for the storm to pass then drove east up Airport Road north of Binghamton. I pulled to the side of the road and noticed a new storm was approaching from the west, but it was a little further north than the previous supercell.

I turned west onto Commerce Road then south onto East Main Road which gave me a good hilltop view of the new storm. The gust front was ominous looking. The rapidly expanding shelf cloud had a smooth, striated look. I decided to move further south to avoid the gust front and wanted to get a look at the southwest portion of the storm. Just as with the previous supercell, this storm showed signs of becoming rain wrapped. As I crested a rise a little further south, I saw a large lowering from the rain free base to my west. I stopped and watched scud move into it from the south, north, and directly underneath. I didnít notice any rotation, but the vertical motion was intense. A tornado warning was issued by the Binghamton National Weather Service at that time. Since the lowering was rapidly moving toward me, I continued south to avoid it and soon lost sight of it in the hilly terrain.

CHASE STRATEGY: MAY 30, 1998 by Tim Marshall

My chase vacation was half over when this day presented itself. Carson and I had driven up to Salina, Kansas the previous evening along with a film crew from Japan. We awoke to the sound of car doors slamming outside our motel rooms and chasers talking. I peeked outside the window to see the sub-VORTEX crew preparing their vehicles. They evidently arrived late last evening. Bright minds think alike! ---so I thought. Looking at the morning data, I could see a surface low north of Grand Island with a frontal boundary extending northeastward through southern Minnesota. Northeast surface winds were at Norfolk, NE and Sioux Falls, SD; the result of nocturnal thunderstorms. There was a nice nine degree difference in temperature between Omaha, NE and Sioux Falls. The dewpoint axis intersected the front in south-central Minnesota. Warm, humid air was moving north through Kansas; the atmosphere was highly capped there into southern Nebraska and southern Iowa. I anticipated the surface low would move into northwest Iowa by late afternoon and storms would fire along the warm front in southern Minnesota and along the front in western Iowa. I oriented my forecast ellipse east-west along the Minnesota-Iowa border.

The upper air maps showed the huge, persistent ridge over the central U.S. At 250 mb, the 12z jet stream was positioned along an arc from near Rapid City, SD to Minneapolis, MN. There was ample low-level turning with height for supercell spin. We left early and planned to stop near Des Moines, IA for a weather update. As we approached Des Moines, we heard radio reports of a tornadic supercell near Ames, IA heading east-southeast (about 30 miles to the north). I thought this was going to be an easy catch. We then started hearing transmissions from the sub-VORTEX crews who also were closing in on this storm. Just as the storm crossed I-35, the storm began going down hill and quickly died for some yet unknown reason. A plot of the 20z surface data showed we were in the warmest easterly winds with the surface low between Omaha and Sioux Falls. SPC had issued a tornado watch for southern Minnesota and northern Iowa which included my entire forecast area. We waited the rest of the afternoon in Ames, IA watching THE WEATHER CHANNEL for any new convection. Around 4pm, we noticed a wave of high based convection moving rapidly eastward across South Dakota on the radar loops but didnít think it would get to the deeper moisture until after dark. By 6pm, the deeper moisture had moved into southeast South Dakota --to late for us to get there. We watched in agony as supercells developed on the radar -a good 300 miles to our northwest.

MKC AC 301430


VALID 301500Z - 311200Z







Hey, why didn't someone tell me about Dakota magic before? I've been hanging out around the Oklahoma and Texas panhandles all these years, waiting for the Holy Grail of Tornadoes to unfold before me. Maybe I should trade in Amarillo for Aberdeen.

The 1998 chase season was weird. I arrived in Amarillo on May 14th, and soon learned of a tornado warning. YES!! Oops, the warning was for Agoura Hills, California. That is where I live. That is where I was less than two days prior. It looked like one of those years. It looked like El Nino was turning the chase season upside-down. I abandoned the Lone Star State and placed myself in Nebraska. A couple of days passed with nary a cumulus cloud, but then, suddenly, a major turning point! It was the ultimate right-mover, right there on my motel room television in Hastings. Newt Gingrich was doing the "Look Ahead" segment on The Weather Channel! I chased supercells on six consecutive days after El Newto's guest appearance.

I was in Ogallala, Nebraska, on Saturday morning, May 30th, 1998. With me were Cheryl Chang and Keith Brown. We three had been teamed up for about ten days, and we were treated to really cool storms on more than half of those days. Cheryl lives in Calgary, Alberta, and her first chase trip was nearing its end. Keith was looking at his last chase day, as he had to be in class at Norman on June 1st. Martin Lisius had met up with us the previous evening, and he was rather desperate to see a decent storm---a storm free of the smoke which was plaguing much of Tornado-less Alley this spring. On Friday, Martin drove from Arlington, Texas, to Big Springs, Nebraska, only to see some distant lightning after dark. Martin was a man on a mission on Saturday. Knock-knock-knock. "It's 7 a.m., wake up people! We have to drive to Iowa!"

Iowa, Illinois, Indiana...geez, what's the difference. They might as well be in Iberia. They're too far away. A chaser's got to know his limitations. I'm a High Plains chaser. There are haze and trees in Iowa. Too much of Nebraska is between me and Iowa. Wake me up at checkout time. I had to chase, though. The severe-weather setup looked pretty good. It was the last chance for Cheryl and Keith. There would be no storms for several days after today. Besides, what would Martin think of me if I wimped out? Keith joined Martin in his Explorer, and Cheryl and I planted ourselves in the Pathfinder. On to Iowa!

By noon we were in Grand Island, and Martin stopped at a Bosselman's truck stop to fill his laptop with unleaded data. The forecast charts continued to suggest that the center of the severe-weather-parameter universe would be near Storm Lake around 00Z. We could get close to Storm Lake by 4 p.m. if we continued our brisk pace towards Omaha. It would be nice to get out of Nebraska. Thick high clouds were covering the entire state, it seemed. There was a cap to break, and it was not going to break where there was no sunshine. Ain't no sunshine in the Cornhusker State.

At York, 40 miles east of Grand Island, Martin exited Interstate 80. Keith had been digesting the newly downloaded weather data, and a couple of NWS discussions indicated that a developing surface low-pressure system was moving east-southeastward through southern South Dakota. The low was forecast to be in extreme northeast Nebraska by mid-evening. We agreed that it would be propitious to place ourselves in front of this low, and to try to lose this depressing high-cloud shield. The moderate-risk area was in Iowa and Minnesota, but we were now playing the low. We headed north on U.S. 81 through Columbus and Norfolk, helped along by humid southerly winds.

Around 5 p.m. at Willis, Nebraska, the high clouds thinned considerably and some flat cumulus were overhead. We were along an area of weak convergence, with southwesterly winds at Norfolk and southeasterly winds at Sioux City. The temperature was 82F, and the dew point was 72F. Had we found the spot where we wanted to be? On our hilltop near Willis we scanned the skies for convection, and waited. Nothing. Our chase prospects looked bleak. Keith muttered that it looked like a major "bustola capola". NOAA wx-radio from the Siouxland said that storms could be expected after 10 p.m.---not good. Maybe our forecast had been bad. Maybe the mid-level cap was getting stronger. Maybe I should have stayed in Ogallala. It was 6 p.m. and time for drastic action, so Martin called Jason Jordan at the NWS office in Fort Worth. Jason said that there were some storms near Huron and Chamberlain in South Dakota. Wow--- those towns are not very close, but at least the cells are moving east and southeast. We figured that if we were going to see any storms on this day, we would have to try to catch something associated with this activity. I had never chased in South Dakota before.

We were close to Sioux City, so we motored east a little, hopped onto Interstate 29 north to Sioux Falls, and then went west on Interstate 90. Along the way I learned of severe thunderstorm warnings for Hand, Sanborn, and Miner counties, all to our northwest. We made good progress at 75 mph on the Interstates, and a new, hard thunderstorm anvil soared overhead just west of Sioux Falls. It was almost 8 p.m., and sunset was still a ways off (thank God we were so close to the North Pole!). I had a good feeling about how things had changed for us in two hours. Radio station 1390 AM in Madison reported that a new storm, near Mitchell, deserved attention. That must be our anvil maker! It's dead ahead. It's moving towards us. It's a tail-end Charlie. It's in a Tornado Watch. Life is good!

Martin and Keith exited I-90 onto U.S. 81 (again!), and I followed. Martin missed the road west out of Salem, and we convened just north of the small town. Lightning flickered in the anvil above, and we knew that there was a strong cell to our west---we just couldn't see it. A scrawny updraft just to our west was obscuring the area of interest. Soon, a tornado warning was issued for northern Hanson County, the next county west. Moments later, AM 1390 said that there was a tornado approaching Farmer, just 15 miles west of Salem! The storm was moving east at 35 mph. It was about 8:25 p.m., and time to head west again.

We quickly cleared the annoying condensation of the intervening mini-cell as we drove west from Salem on Highway 38. The sky opened up dramatically, and, along the horizon to the west-northwest, was a strange yellow tinge. Trees occasionally blocked our view, but there was something bisecting this shallow band of light. Could this be the tornado that was reported? One more line of trees to clear...OH MY GOSH---YES!! It's HUGE! Look up there---an awesome striated rotating updraft! In front of us, and headed towards us, was a classic Midwestern supercell, with a fully developed and beautifully backlit tornado.

This is the moment that chasers yearn for! This is why we drive 10,000 miles every spring! This is why we save all of our money for gas, motels, maps, cameras, film, laptops, scanners, and cell phone bills. This is the moment that we fantasize about while driving down that deserted two-lane road in West Texas after another busted chase. This is the moment of fulfillment and exhilaration, and all of those other feelings that are impossible to explain to someone who doesn't give a thunderstorm a second thought!

Elation notwithstanding, I had to concentrate. Where should I stop? I could practically drive right next to the tornado by continuing west. I was leading the two-vehicle chase team down Highway 38, and I elected to stop relatively soon. The tornado was still about 5 to 6 miles away, but it was approaching rather rapidly, it was not going to dissipate any time soon, and the lighting and contrast were too good to pass up. I turned right on a paved north-south road and almost immediately pulled into a turnout on its west side. (This road was six miles west of Salem and four miles east of Spencer. Martin and Keith stopped along U.S. 38 about an eighth of a mile behind Cheryl and me.) I did not know that the town of Spencer was between myself and the tornado. A large and empty farm field afforded an excellent view of the storm, and I nervously set up the tripod and the camcorder. Make sure it's level, Bill! Make sure the lens is clean, the unit in focus, and NOT on pause!

The time was 8:35 p.m. A few small hailstones struck the Pathfinder, and winds were from the east-southeast at about 20 mph (compared to light winds in Salem ten minutes earlier). Just north of due west spun a scene of congruity and chaos: above, the mother cyclone, in perfect tropospheric harmony; below, total tornadic tumult. The seething vortex distributed dust, dirt, debris, and accessory clouds in fast-forward speed. Dust plumes and curtains, condensation skirts and collars, spokes and fingers---all whirled around in an unbridled frenzy. The funnel motion on this thing was unreal! Upward translation along its north fringe was at warp speed. The small town of Spencer found itself in the direct path of the tornado at 8:37 p.m. Six residents perished, and most of the town was demolished.

The tornado maintained its size and strength as it drew nearer. Early on I thought that it might pass just to our north, but now it was only a couple of miles away, and due west. The orange-yellow background turned a faint peach color, not much brighter than the gray tornado. The tornado hid itself momentarily in dust and condensation, and then emerged completely for new photo-ops. The front edge of the rotating updraft moved overhead, and lightning activity suddenly increased. I told Cheryl to get inside the vehicle, and I tried to stay low. The landscape lit up a couple of times and thunder crashed within two seconds. It was too dangerous to be outside. The camcorder came off of the tripod and into Cheryl's hands. I tossed the metal Bogen tripod into the vehicle, closed the rear hatch, and jumped into the driver's seat. Whew! A violent-class tornado was just down the road, but I felt a LOT safer now!

For about 20 seconds we watched the increasingly black funnel accost the farmland to our west-southwest, less than two miles away. We could not stay here. The tornado was going to be just to our south in a couple of minutes. I decided to blast south towards Interstate 90. Cheryl pointed the fully zoomed-out camcorder towards the west as I accelerated south of Highway 38, and the surface circulation soon filled the entire video frame! A developing squall line was advancing upon the tornadic mesocyclone, and conditions were changing fast. The tornado, though wider, was weakening, and it was incorporating a lot of dust and dirt. After about two miles, or halfway to the Interstate, I turned the Pathfinder around for a quick look. Yikes! A wall of blackness was rapidly descending upon us. There was almost zero contrast. We knew that there was a tornado in there, but this storm looked like something from the Kansas Dust Bowl Days. Plumes of dust were zeroing in on us. The storm was turning outflow dominant, and these chasers needed to get away fast!


We bailed eastward on I-90, south on U.S. 81, and east to Canistota. A new supercell was in the works south of the Spencer storm, which spawned tornadoes (which we never saw) near Alexandria. The sirens were wailing in Canistota, where storm core and severe outflow appeared ready to converge with the vertical storm tower to the west-southwest. Our visit to Canistota was very brief. We surged south and east again (on Highway 42) to avoid the ugly jaws. It was getting dark, and I wanted to be as far away from this double-headed monster as I could! New tornado warnings were issued for towns just behind us, lightning streaked through the mammatus in front of us, and I just held on tight and flew eastward against the strong inflow.

I called Martin after we rode out a core in a motel parking lot in Sioux Falls. He and Keith were out of harm's way, south of Sioux Falls. I stepped into the motel, and found the lobby and hallway filled with folks, many of whom were quite distressed. The lobby television showed an endless stream of radar loops and warning scrolls, and we learned that Spencer had been struck by the tornado we had seen.

Martin and Keith found Cheryl and me at the motel, and we searched for an open restaurant in Sioux Falls. It was celebration time! Our chase was a tremendous success! On the restaurant television, however, the news was becoming increasingly disheartening. There were fatalities in Spencer. Our meal didn't taste so good all of a sudden.

It was a scary evening for many South Dakotans on this Saturday, and it was a roller- coaster ride of emotions for my friends and myself. Four storm chasers, who started the day in southwest Nebraska and wound up between Salem and Spencer in southeast South Dakota, caught a storm that they will never forget. Sadly, six souls from Spencer were caught in a storm they will never remember.



I left Ogallala, NE that morning and headed east on I-80 with an initial target of northeast Nebraska and northwest Iowa. With dewpoints in the low to mid 70ís, moderate south to southwest winds, CAPES between 2000 and 3000 j/kg, lifted indices around -6C, a dryline approaching from the southwest, and a cold front extending from a low in South Dakota, my forecast area looked prime for severe storms and possible tornadoes. This area also was located on the western edge of Storm Prediction Centerís (SPC) moderate risk area. The only real negative appeared to be the heavy cirrus overcast that persisted across the entire state of Nebraska. A stop at the Bosselmanís truck stop in Grand Island, NE to retrieve the latest data at 1230 CDT, still confirmed my target. I met William Reid and his group and they were thinking the same general area.

Cirrus overcast continued passed Omaha. The NOAA weather radio was reporting sunny skies to the north. I set up the DSS dish to have a look at the TWC radarís at 6pm CDT. There was a line of strong storms located south-southwest across central South Dakota and moving east-southeast. I got back on I-29 and headed north from Onawa, Iowa. Approaching Vermillion around 7:30 pm, I could begin to see the faint outer edges of thunderstorm anvils to me north and northwest, which I judged to be around 50 miles away. Nothing looked real exciting at this distance. Approaching Yankton, spotter reports were coming in along I-90 from Chamberlain east to Farmer, South Dakota of 80 mph winds and large hail. I lose some time getting a room at Yankton, then head west towards this storm. I even stopped for a bite at Dairy Queen 45 minutes earlier! Why or why did I do this? I WASNíT EVEN HUNGRY!

I proceeded north on Highway 81 and was distracted by rapidly building towers to my southwest. After about three minutes, I decided to continue heading north as this new convection did not appear to be well organized. The time was 8:35 pm. Little did I know that people were losing their lives in Spencer, about 57 miles to my north. I drove another 16 miles to the north and stopped on a hilltop with good view to the north. I could see a lot of cloud-to-cloud and cloud-to-ground lightning but no sign of mid-level rotation from this distance. It was getting dark now and I was too far away to see the base of this storm. After driving 524 miles to get to this point, the storm seemed so close (only 43 miles away), yet, I knew I had run out of time. All I could do now was enjoy the beauty of the retreating cloud towers reflecting the colors of the setting sun -the first and only real storm structure I had seen on this chase vacation. Lesson #1: Never become complacent and assume your chase efforts will end in another "bust" before dark.


It was the most intriguing storm chase of my life. Iíd been roaming all over western Nebraska the last two days having good luck picking up signatures of wind, humidity, and cloud that led me to interesting storms both days, but no tornadoes. A fast moving upper trough out of the Rockies promised to spoil day three by moving through the plains and bringing down cold air to meet the cap which was moving up to the Nebraska-South Dakota border. At 10 am (MDT), the surface wind already was shifting to the west at Chadron, Nebraska, but at least no cold front was in sight. "Go east, old man", I said. Maybe I could find something in Iowa if I could outrun the upper trough.

By noon, I had reached a point ten miles west of Valentine. Sure enough, the wind was west and rather brisk. Depression set in. Would the wind keep shifting to the west as fast as I drove east? The sky was clear except for an equally discouraging dense layer of cirrostratus over the southern 3/4 of Nebraska. But to play it safe, I drove north to I-90, then east to Reliance, South Dakota. Sure enough the wind was west, but the sky was crystal clear. But some excitement lay ahead. On the other side of the Missouri River, there was hazy cumulus and the dewpoints jumped to 78 degrees (F). Winds were west-southwest. At 1600, I reached mile marker 296 on I-90. A quick check of my laptop verified that the winds were southeast at Sioux Falls and south at Huron. I will outrun the upper trough! I was in a sea of hazy small cumulus moving slowly from the west. It sure looked like the atmosphere was capped.

The next three hours became a mixture of delight, terror, and agony. At 1630, a little patch of altocumulus castellanus (ACCAS) built downwards towards the lazy, hazy cumulus which took itís cue and exploded into towering cumulus in a matter of minutes. The first supercell appeared in the direction of Huron, SD so I raced northeast to Highways 34 and 281. Then, another supercell developed to the south. I remained between the two storms on Highway 34. As expected, the anvils that poured out from these two supercells enveloped the eastern quadrant of the sky with rain and hail. Middle and high clouds were present in the storm free area above Highway 34. Since the north storm was the closest, I kept up with it. The storm produced a brief funnel but soon ran out of "pure" moist, inflow air. Soon, I found myself enveloped in heavy rain and hail -the likes of which Iíve never seen before. The storm activity started moving southward and I played tag with the outflow for a good hour.

Enjoying this game of tag, I lost sight of the supercell to my south. But then it happened. Through the rain, I could see a tornado to my southeast with plenty of debris around itís base which I first thought were trees. I saw the tornado for a few seconds before it became rain and debris wrapped. Nothing can symbolize in words what the sky looked like. No artist could duplicate on canvas or with music the color, texture, and movement of mother natureís most spectacular display of violence and beauty. The large tornado seemed insignificant, mostly shrouded in foreground rain, to the magnificence of the immense supercell above it -later illuminated by the setting sun. A minute after I saw the Spencer tornado, two other tornadoes quickly formed and dissipated near I-90.

Then the agony set in. As I began the long drive back home to Salt Lake City, UT, an endless stream of flashing red and blue lights was coming the other way to a town I had just passed at mile marker 354. Listening to the radio, I heard of the tragedy at Spencer and cried a little. It was hard at that moment to compromise oneís elation and excitement with anotherís brutal encounter of death and suffering.



Spencer, South Dakota by Martin Lisius

It was Saturday evening, June 13, 1998. I was having dinner with some of the top storm chasers in the country at the Wagon Wheel Tavern in historic Marysville, Kansas. Fellow chasers Carson Eads, Tim Marshall, Alan Moller, Gene Rhoden and I sat down to a late meal after chasing a fast-moving, high-precipitation supercell along the Kansas-Nebraska border for several hours. The storm started out well to the west of Highway 81 and moved rapidly to Southeast Nebraska. At times, it looked like the storm might produce a view-able tornado, but the high-precipitation structure and rapid movement made things difficult for us. We abandoned the storm northeast of Marysville as our visibility deteriorated.

Before arriving to dinner, we heard about a tornado that had just tracked across the north side of Oklahoma City. All of us had driven to Northern Kansas, separately, from the Dallas-Ft. Worth area. We had passed through Oklahoma City just hours earlier. We were surprised to hear about the tornado that had had occurred much closer to home. During a pause in the dinner conversation, I sighed and then said quite profoundly, "tornado forecasting is very difficult."

Veteran storm chasers often look at a multi-state area when forecasting where they think tornadoes will occur before departing on a chase. They may consider a 300,000 square mile region. If they are lucky enough to see a tornado that produces a path one-half mile wide and ten miles long, for example, then simply put, they just pinpointed a five square mile area out of a possible 300, 000. Forecasting and intercepting tornadoes is literally like looking for a needle in a haystack!

When I departed Dallas-Ft. Worth at 4 a.m. on the morning of Friday, May 29, I had no idea that the needle I would eventually find would be in Spencer, South Dakota. My target for that day was South Central Nebraska. Specifically, I was targeting Hastings, where much of my family lives. Conditions appeared to be coming together there for a possible tornado event later in the day. Thermodynamics and wind shear both looked good. That was until I reached Central Kansas and got a call via cell phone from storm chaser Bill Reid. Bill relayed data that indicated that outflow from nocturnal storms had stabilized the atmosphere over much of Nebraska and Northern Kansas effectively eliminating any chances for tornadoes later in the day. His updated analysis showed that there was a slim chance of tornadoes perhaps in Northeast Colorado. So, at Salina I turned west on Interstate 70 for the Rocky Mountain State. By 8 p.m., I was in Colorado and could see distant, weak thunderstorms near the intersection of the Colorado, Wyoming and Nebraska border from my location north of Yuma. Things were not looking good. Soon after a mild attack of despair began, I received a much-needed pep call from Bill who was just over the border in Big Springs, Nebraska passing time with chasers Keith Brown and Cheryl Chang. Bill announced that it was basically a bust for everyone and asked me to meet up with him. I had not chased with Bill since 1996, and was eager to visit with him.

Bill, Keith, Cheryl and I sat down to a hearty meal in Ogallala. After dinner, we drove to the edge of town to watch a distant and meager lightning display originating from somewhere over the Nebraska Panhandle. Back at the motel, the radar indicated that this activity was isolated and that large-scale convection may not form, leaving an unstable atmosphere for chasing on the Plains the next day.

I rose with the sun on the morning of Saturday, May 30. I was very eager to download weather data on my laptop to see how and where conditions were coming together for a possible chase. One of the first discussions I read was a public outlook issued by the Storm Prediction Center (SPC) that called the event shaping up "a particularly dangerous situation" with the possibility of damaging tornadoes. Their area of concern was primarily over Iowa, Eastern Nebraska, Southeast South Dakota and Southern Minnesota. A quick look at surface and upper air data supported their outlook. I was soon on the phone attempting to wake the Reid gang with the news. After several failed attempts to contact them by phone, I decided to drive over the Platte River to their motel in Downtown Ogallala. Once stirred from bed, the three weary but expedient chasers were ready for the road and our long trek eastward across the Cornhusker State. Our initial target was Northwest Iowa via Omaha. We would split the four chasers up into teams of two. Keith rode with me in my Explorer while Cheryl rode with Bill in his ubiquitous and storm-battered Pathfinder.

We broke for lunch at Bosselmanís Truck Stop in Grand Island, hungry for both food and data. A quick download of data was facilitated by a special digital phone booth there which featured a data jack and small table. Over a tuna fish salad sandwich, compliments of Mr. Brown, we analyzed data. After a hasty glimpse at conditions, our initial target remained intact. However, as we drove east on Interstate 70, Keith read deeper into our noon (1700Z) data download. A surface analysis indicated that winds were veering to the southwest over Central Nebraska. In addition, the huge cirrus cloud shield that I had seen on the satellite loop early that morning was now covering much of Nebraska including our location. This was not good. We stopped at York and decided to go north to Northeast Nebraska as quickly as possible where the noon analysis indicated backed winds and sunny skies would await us.

By the time we reached Norfolk, winds had veered to the south-southwest and the cirrus shield above was thick. North of Norfolk, we turned east toward Sioux City, Iowa. As NOAA Weather Radio in Sioux City drifted in on my HAM radio, we could hear reports of sunny conditions from Yankton into Northwest Iowa. Just west of Sioux City, we crossed over a wind shift line and stopped. It was apparently a local outflow boundary from a morning storm that tracked through the county based on old warning statements we read in our data. Winds were easterly at about 5 kts. there. A small towering cumulus grew overhead. For a few minutes, we were excited. Things looked promising. But then the tower collapsed along with our spirits. We continued toward Sioux City and our original target.

It was late afternoon when we stopped on the west side of Sioux City, to view a lifeless sky. What was going on? There should be storms by now, we thought. I called fellow chaser and friend Jason Jordan at the National Weather Service Forecast office in Ft. Worth. Jason relayed information about building storms well to our north in South Dakota. Meanwhile, Bill was monitoring an AM radio station in Yankton. A tornado watch had been issued north of Sioux Falls, South Dakota. It would be a stretch to reach it by dark but we had no other choice. We had traveled far on our quest. We couldnít give up now. So, it was north to Dakota!

Based on information gathered from the AM radio station in Yankton and from Jason, we decided on a target area near Mitchell, South Dakota. Two, then-existing thunderstorm cells would merge there about the time we expected to arrive. The southern storm was moving east, while the northern of the two was tracking southeast. It was after 6 p.m. now, so it would be tough to make it there much before sunset. That was until we realized that the sun would not set until around 9 PM! Finding this out was a jump-start for us. We continued to monitor the radios as we passed by the Gateway 2000 computer factory painted up like a giant Holstein, and located in a pastoral meadow alongside the blue waters of the Big Sioux in North Sioux City, South Dakota.

While on Interstate 29, south of Sioux Falls, Jason Jordan relayed updated information which indicated that the two previous storm cells had fallen apart. Still, the Mitchell target seemed as good as any. We stuck to it. As we entered the Sioux Falls area, we could make out a sharp anvil edge overhead. We turned west on Interstate 90 toward our target area. The southern edge of the anvil was directly overhead, with pulse waves evident, and paralleled the east-west Interstate. Our mission now was rather simple. We would follow the anvil westward to the storm it was anchored to.

We exited the Interstate at US Highway 81 near Salem. The anvil edge had led us to the southern-most cell in a line that stretched northward. Our plan was to turn west down Highway 38 toward Spencer. However, we missed the turn in Salem and continued north about a mile and pulled over on Highway 81 to view the storm. Winds at Salem were south-southwest at about 5 kts. Our storm displayed decent, but slightly mushy towers. I was not impressed. Soon, I spotted a small and newly-developed anvil just to the west of the convection in front of us. A brief moment passed and a tornado warning was issued. Radio reports indicated that the tornado was moving in our direction. Should we wait for it to come to us? Bill didnít like that idea. So, the decision was made to go west on 38, our original plan.

I had never chased a storm in the Southeast South Dakota area before. I was impressed with the terrain. It was very much like South Central Nebraska, in the Hastings area, for example. Gently rolling, mostly flat topography, with only scattered trees and an excellent, section road network. The countryside was occupied mainly by corn fields and pastureland.

As we headed west on Highway 38 from Salem, Keith exclaimed that he could just make out striations in our newly-discovered western storm. A few miles more and we could see a lowering nearly touching the ground just to our west-northwest. At least thatís what it looked like. An area of trees just to our right was obscuring our view to the ground. We cleared the trees and the "lowering" turned out to be a fully-developed tornado! We cleared a small hill and pulled over at a flat, treeless area 4.4 miles due east of Spencer on Highway 38. Bill and Cheryl continued west a few hundred yards further.

Adrenaline was rushing for Keith and me as we unloaded camera gear. My priority was to begin shooting the tornado as soon as possible with my 35 mm motion picture camera. Pointing my spot meter at the tornado, I received a reading of 2.8 at an ISO rating of 250. Luckily, the zoom lens I was using was fast, with a maximum aperture of 2.8. Perfect. The tornado had quickly developed into a small wedge. It was to our west-northwest and spinning hard. I shot about 200 feet of film and then grabbed my Betacam SP camcorder while Keith shot Hi8 and took stills. For us, the moment was about concentration, getting the image and doing it right. There was not much time for anything else. The tornado was moving rapidly east-southeast, at about 30 kts. in what appeared to be our general direction! Frequent cloud-to-ground lightning was hitting so close to our position that the thunder sounded like shotgun blasts with little reverb. At times, I could feel the heat of the lightning on my face. Keith and I moved about in crouched positions to reduce our risk of death by electrocution.

A citizen pulled over and asked us, "is that what I think it is?!" Keith and I replied that it was a violent tornado. He asked if it was safe for him to continue driving west. We told him no. He said that he was on his cellular phone with a local radio station and asked if he should report it. We told him yes and asked him to tell them that it was a large and dangerous tornado.

Within five minutes of arriving at our site (appx. 8:38 p.m. CDT), the tornado had developed into a mighty and majestic wedge back-lit by the orange light of a setting Dakota sun. The vortex was dark gray with a lighter "skirt" rotating wildly above, highlighted with a hint of blue from the clear sky to the south. The storm base of this relatively small supercell resembled an inverted wedding cake and was classic in form. Surface winds were dramatically different than what we had observed just 5 miles to the east and 10 minutes earlier. They were now strong, at about 25 kts. and gusty, from the east-southeast. It was amazing how this supercell simply wrangled the local atmosphere. Was the storm the cause or was it the effect of this change? Looking back, it appears that the earlier convection to the north may have put down a southeast to northwest aligned outflow boundary, increasing convergence and causing winds to back.

The tornado was now nearly a mile wide and entering Spencer to our west. Muted power flashes were observed at the base of the funnel as electrical lines, and possibly transformers, arced and exploded. A violent inflow jet was evident over Highway 38 as the powerful tornado ingested black soil from a plowed field. It was apparent that the tornado was going to pass over Highway 38 just to our west.

Since I had been driving earlier, I did not have the acute knowledge as to our position relative to Spencer that Keith had. As team navigator, Keith knew that Spencer was possibly being devastated. I saw an area of trees between our position and the tornado. I suspected that the trees represented the location of the town and that the tornado was skirting along its west side. Neither of us were certain, so our concern remained with the storm itself which was now south of Highway 38 growing larger and becoming wrapped in dust. At one point, the northern edge of the apparent vortex seemed to hover over Highway 38 for a moment while the tornado grew larger. At this point, it appeared that the dark giant was tracking east, toward us! Path analysis would later indicate that the tornado was actually moving east-southeast as it crossed 38, 1.5 miles to our west, eventually turning southeast, south of the highway. Just to be safe, we made a hasty departure, tossing cameras and tripods into the back of my Explorer. We headed east down 38 glancing back at Bill and Cheryl who were heading south, crossing in front of the tornado!

Soon, hail and driving rain raked us from the north. We motioned motorists along the highway to head east. I dialed 911 on my cell phone and reported that the tornado was tracking rapidly toward the Interstate. Once again, I emphasized that it was a large and violent tornado. Keith quickly reviewed the map and prepared a route that would keep us east of the tornado as we zig-zagged southeast.

By the time we reached Interstate 90 at US Highway 81, the tornado was heavily shrouded and obscured by dust to our west-northwest. Along the way, Keith mentioned that a flash of lightning revealed what appeared to be a small cone-shaped tornado within the dust.

We tracked the storm southeastward to Parker driving in and out of violent straight-line winds. We avoided making the dangerous mistake of stopping downwind from trees. The storm was racing now. Our fear was that tornadic activity might still exist deep within the dust. We werenít taking any chances after seeing what the storm had produced earlier. I recall hearing the squeal of my tires numerous times during the retreat.

Darkness had arrived and the storm had all but collapsed by the time we reached Interstate 29 south of Sioux Falls. Radio reports were indicating a new tornado warning to the north over the city. We headed to Sioux Falls to take a look and received a call from Bill who was waiting for us there with Cheryl.

Keith and I spotted Bill and Cheryl parked in town. We were greeted by sweet, rain-cooled air and two weary but happy chasers. We found a restaurant still open and serving steak, the symbolic prize for a successful chase.

A steak dinner was well-deserved by this chase team. Our intercept had begun hundreds of miles earlier. We had thwarted all the obstacles and met all the challenges Mother Nature could muster. Analysis, more analysis, hope, trust, perseverance and teamwork led us to that little fragment of Tornado Alley at Spencer, South Dakota. We had found the needle in the haystack. And, I was able to capture possibly the first violent tornado on 35 mm film. All was well.

Still, our journey was not complete. A television at the restaurant was tuned into a Sioux Falls station broadcasting live from Spencer. We grew silent as news indicated that the tiny community had been destroyed. Reports said that several had died and dozens were left hurting. Our logistical victory was now a human tragedy as well.

Keith and I were greeted by sunny skies, and cool, blustery north winds as we arrived in Spencer early the next morning. Highway 38 was busy on the south side of town as a media camp had been established. All roads to Spencer were blocked by order of the McCook County Sheriff. Large trees were stripped clean of branches and leaves, a farmer placed dead cattle into a truck using a front end loader. The sounds of earth-moving equipment drifted in the wind. We saw a car in a tree, farmsteads destroyed, a toppled grain elevator, and what looked like a bombed-out town. We measured the tornadoes path at .8 mile wide at Highway 38. Northwest of town, we estimated damage along an east-west section road at 1.5 miles long. The damage was incredible.

As storm chasers, we live two lives, one as scientists, the other as humans. As scientists, we concentrate on the physics and the logistics. We are excited and awed by the power of nature. Violent storms and tornadoes are beautiful to us. Sometimes we forget, if only briefly, that to many others violent weather is a curse. Spencer was destroyed by nature. Simply put, an amazing amount of atmospheric energy came together there for a moment. It was haphazard. Unlike man, weather does not kill by intent.

Tornadoes are difficult to forecast. Each time we find the needle in the haystack, we take a step. Sometimes, itís a step back, as nature shows us how little we really know. But, over time, perseverance will move us forward and closer to fully understanding tornadoes. The more we know the more we can help humankind.

Martin Lisius is an award-winning producer and director for film and television. He can be contacted through his companyís Internet site at