The Scottsbluff Tornado July 26 1993

The Scottsbluff Tornado
July 26 1993

Jonathan D. Finch

Related items of interest

Historical Tornado Cases for the Boulder Warning Area

Historical Tornado Cases for the Cheyenne Warning Area

Historical Tornado Cases for the United States

June 03 1958 Wyoming Tornadoes

April 23 1960 Cheyenne Ridge Tornado

April 12 1967 Wyoming Tornado

May 7-8 1965 Front Range Tornado

June 14-17 1965 Front Range Superstorm

The July 16 1979 Cheyenne Tornado

Elevated Mixed Layer

Elevated Heating

High plains and front range topo maps

Severe Weather Occurrences

                                        On July 26, 1993, isolated thunderstorms developed by early afternoon along the foothills from Converse county to
                                        southern Platte and northern Laramie counties. The first tornado of the day occurred 4 miles northwest of Sturgis at
                                        1819 UTC. The rear window of a hatchback was sucked out, the doors of a barn were ripped off and a fuel tank was
                                        tossed 400ft. Golfball sized hail occurred in Niobrara county, WY and Custer county, SD at 2040 UTC and 2210
                                        UTC respectively. A tornado was reported 3 miles southeast of Hot Springs at 2145 UTC.

                                        Severe storms also occurred in very rural areas of northwest Kansas after 19 UTC. There were several reports of
                                        large hail and one funnel cloud report. It is possible that tornadoes occurred in rural areas between Tribune and Colby.

                                        Only one storm hit a populated area on July 26, 1993. This storm probably developed in northwest Laramie county or
                                        southwest Platte county around 1930 UTC and moved east-northeast. At 220 pm MST(2120 UTC), a funnel cloud
                                        3 miles southeast of Hawk Springs was relayed by the Torrington PD to the National Weather Service in Cheyenne.
                                        This was based on information from a volunteer fireman who said that the tornado never touched down. But the report
                                        was contradictory. According to the fireman, the cloud never reached the ground, but drew up dirt in a funnel shape
                                        and they never connected. But the debris cloud does not have to meet the condensation funnel to have a tornado. So
                                        this was in all liklihood a tornado but was never documented as one. This apparent tornado may have occurred before
                                        the report time. This storm continued moving east-northeast and produced baseball to softball sized hail in northwest
                                        Scottsbluff county. A tornado touched down west of Scottsbluff and hit the northwest side of town around 3 pm. This
                                        tornado caused extensive damage as it moved from highway 26 to highway 71. Several homes and businesses were
                                        totally destroyed. A woman sought refuge in a ditch and the tornado moved her 60 feet into a field. This tornado injured
                                        10 people and was officially rated F2. Here is some photos of the tornado taken by Lynn Herdt of Scottsbluff.Some of
                                        these show debris--even trees apparently being lofted.

                                        Pic1    Pic2    Pic3    Pic4    Pic5    Pic6    Pic7    Pic8    Pic9    Pic10    Pic11    Pic12    Pic13

                                        A map of severe weather reports for July 26 can be found here.

Meteorological Discussion

            What a month!!!

                                        July was very active month for severe weather across the plains and front rage. Severe weather with golfball or larger
                                        hail or tornadoes occurred on all but the 25th and 29th. The highest concentration of severe weather along and north of
                                        a quasi-stationary front where upslope flow carried very high dewpoints from the flooded areas of Kansas into Colorado.
                                        Then these storms would often rumble east and south and drench central and eastern Kansas as well as eastern Nebraska.


            July 24

                                        July24 was quite an active day in the central plains. Several damaging tornadoes occurred, along with large hail.
                                        Severe storms developed in northern Kansas in the afternoon and then progressed into southeast Nebraska in the evening.
                                        A strong shortwave trough progressed across the northern plains on July 24. Despite the very warm 700mb temperatures
                                        storms developed across northern Kansas and became tornadic in southern Nebraska. In fact the entire troposphere
                                        was warm on the evening soundings from Topeka, Omaha and Dodge City with 500mb/400mb/300mb temperatures
                                        from -4 to -6C, -15 to -17C and -31 to -33C respectively. The Topeka sounding had extremely high dewpoints of
                                        25.6C, 24.3. 23.6, 21.8. 21.4,18.4, 16.3 and 15.3C at 974, 925, 900, 859, 850, 839, 833 and 824mb respectively.
                                        Until I worked on this case, the Topeka sounding from August 7 1962 00 UTC was the most moist sounding I had ever
                                        seen. In the 1962 sounding, the surface dewpoint was higher (27C) but the dewpoints between the surface and 850mb
                                        were a little lower. However, the 1962 Topeka sounding was more unstable (for surface bases convection) owing to
                                        cooler 500-300mb temperatures and a higher surface dewpoint. The 21 UTC surface chart showed a thermal low over
                                        central KS and possibly an outflow boundary or warm front near the Nebraska/Kansas line. The main synoptically
                                        forced surface low was in the northern plains. Storms developed just northeast of the thermal low.



           July 25


                                        A cold front pushed south into west Texas and northwest Oklahoma behind the shortwave trough from the previous
                                        day. By 12 UTC July 25, this front was located from northeast Kansas to near Clovis. At 21 UTC July 25, this front
                                        was returning north on the western end, stationary over northwest Oklahoma, and still moving southeast as a cold front
                                        through Missouri. Rich surface moisture was noted from northeast Kansas into the Texas panhandle. Low level drying
                                        was beginning across central Texas. The 500mb chart at 00 UTC July 26 (5 pm MST) showed the northern plains
                                        shortwave trough still progressing into Minnesota/Wisconsin. The 700mb chart featured a warm plume across the
                                        southern Rockies. Very cool 700mb temperatures were noted over the northern Rockies, with a strong baroclinic zone
                                        from northern California into Wyoming. The next shortwave trough was evident over the pacific northwest. By 06 UTC
                                        July 26 (11 pm MST), the front was still changing orientation with the eastern end moving south and the western end
                                        moving north as a warm front. Low level upslope flow and moisture advection was beginning across southwest Kansas.



            July 26

                        Morning Hours

                                        At 12 UTC July 26 (5 am MST), the front was becoming harder to find on the western end. Also, radiational cooing on
                                        the front range and Rockies made frontal analysis even more difficult. Its is clear that moisture advection had occurred
                                        overnight in western Kansas. A mesoscale convective system was moving east across southwest Nebraska. The leading
                                        edge of the outflow associated with this system had long passed Akron, was just passing through McCook and was
                                        approaching Goodland. The 12 UTC 500mb chart featured a strong shortwave trough centered over central Idaho.
                                        Strong southwest flow(by mid-summer standards) was located from northern Utah into central Wyoming. As one would
                                        expect, a 700mb warm plume (+15C) was present across the central and southern Rockies. A strong 700mb baroclinic
                                        zone had progressed east into Salt Lake City and Lander. Rich moisture was present at Dodge City and Topeka with
                                        850mb dewpoints from 16 to 18C. One jet streak was exiting the plains at 12 UTC, but another was obvious over
                                        northern Nevada and northern Utah.

                                        By 15 UTC, the outflow boundary generated by the morning MCS stretched from east of Cheyenne to between Limon
                                        and Akron to south of Goodland to south of Hill City. On the western end of this boundary, but on the cool side, low
                                        level upslope flow was transporting higher dewpoints into northeast Colorado. The dewpoint at Akron was up to 62F.
                                        Rich moisture was also located south of the boundary in western Kansas but not in eastern Colorado. The dewpoint at
                                        Sidney was probably too low. This may have been a chronic problem since the mid 1970s as my other case studies
                                        would seem to indicate.

                                        By 18 UTC (11am MST), there was a strong surge of moisture upslope onto the central high plains. A dryline was taking
                                        shape from eastern New Mexico into far eastern Colorado. Now that heating had broken the morning inversion over the
                                        Rockies, a surface front cold front could be analyzed from east of Rawlins into central Utah. The dewpoint at Akron was
                                        up to 64F. The mesonet observation at Wheatland, WY indicates a 62F dewpoint. So rich moisture was already banked
                                        up against the mountains to the north of Cheyenne by 18 UTC. The the rich moisture seems to be just east of Cheyenne.
                                        Also notice the strong moisture gradient between Limon, CO and Stratton, CO.

                        Convective Initiation in Wyoming

                                        At 19 UTC, convective initiation was immiment to the northeast of Laramie and also in northwest Kansas. Let us first
                                        examine the severe storm environment in southeast Wyoming.

                                        The Scottsbluff tornadic storm developed out ahead of a strong pacific cold front. By 19 UTC, the Laramie range
                                        marked the western edge of the moist air that was surging upslope into southeast Wyoming. To the west and probably
                                        on top of the Laramie range, dry and potentially hot air was noted. More moist upslope flow was noted to the east. A
                                        surface warm front separated these two airmasses. This warm front moved through Laramie by 17 UTC and was
                                        probably located near the Laramie Range by 19 UTC. An outflow boundary from overnight and morning thunderstorms
                                        over Nebraska was located across southeast Wyoming. The surface boundaries, moisture distribution and terrain features
                                       are shown here.

                                        The surface cold front was through Rawlins just after 17 UTC, Laramie just after 20 UTC and Cheyenne around 22 UTC.
                                        This front was generally progressing to the southeast at 25 to 30 mph. The progression of the front is known here with
                                        fairly high confidence since I was able to use time series of surface observations from Rawlins, Laramie and Cheyenne.

                                        Given that the tornado hit Scottsbluff around 22 UTC and since the storm moved to the east-northeast after initiating
                                        along the eastern edge of the Laramie Range, the initiation point for this storm was likely around 1930 UTC (assuming
                                        a storm motion of about 30 to 35 mph). This seems reasonable given the 6 km winds of 55-65kts (we will discuss the
                                        shear profile in a later section). Even if the storm initiation time is off by 30 minutes ( as early as 19 UTC and as late as
                                        20 UTC), the cold front would still have been to the west during initiation. Therefore, it appears that the storm developed
                                        ahead of the cold front. More specifically, the storm developed along the warm front near the Laramie Range.

                                       The storm developed at the interface of the warm dry air over and west of the Laramie Range and and the warm and
                                        moist air to the east. The warm and moist air to the east was enhanced by the overnight and morning MCS over Nebraska.
                                        On the same chart as the frontal progression, the 700mb temperature field at 19 UTC is presented. How did I develop
                                        this chart? Well, from analysis of many high plains severe weather events, I have found that 700mb and even 500mb
                                        temperatures can often be estimated using surface observations at 19 UTC (local noon) or after mixout occurs. The
                                        time of mixout depends on windspeed and cloud cover. But under dry and breezy conditions, this often occurs by 19
                                        UTC at elevations above 6000ft. I took the 19 UTC surface temperatures at Eagle, Laramie, Colorado Springs, Limon
                                        and a few other locations up the dry adiabat to 700mb. To supplement these data, I used the 12 UTC and 00 UTC
                                        700mb temperatures at some surrounding stations. For the immediate purpose of estimating 700mb temperatures near
                                        the initiation point, the later was not essential. This technique allows for a more detailed 700mb temperature analysis.
                                        This is especially true if a cold front does not progress through the stations between 19 UTC and 00 UTC.

                                        The 700mb temperature was around +15C near the storm initiation point. I estimated the 700mb temperature to be
                                        +15.5 C at Laramie, +16C at Eagle, +17C at Limon and +19C at Colorado Springs. The 00 UTC 700mb temperature
                                        at North Platte was +15C. Along a line from Laramie to North Platte, 700mb temperatures were similar.

                                        So convective initiation occurred early in the afternoon despite the warm 700mb temperatures. However, Keep in mind
                                        that elevated heating often helps with cap removal on the high plains even when 700mb temperatures are as high as
                                        16-18C. Magic numbers or rules of thumb are often of little use.

                        Convective Initiation in Kansas

                                        Storms initiated south of Goodland around 19 UTC near the intersection of the outflow boundary and the dryline.
                                        Very rich moisture continued to flow upslope into northwest Kansas. Even though 700mb temperatures were very
                                       warm(15C) in northwest Kansas, elevated heating north of the boundary, along with 70-72F dewpoints by 20 UTC
                                       resulted in rapid convective development. .


                        "Cool" and Moist Upslope


                                       Even though surfae temperatures were fairly "cool" in the upslope region from northwest Kansas into southeast WY,
                                       potential temperatures were fairly high. So this airmass was actually potentially warm and moist. For example, the
                                        surface temperatures/potential temperatures at Wheatland, Akron and Goodland at 20 UTC were 79F/103F, 81F/106F,
                                        and 86F/107F. So potential temperatures ranged from 103F to 107F on the "cool" side of the outflow boundary.

                        Shear, Hodograph and Surface Based CAPE


                                        here. Unfortunately, the flow from just above the surface to about 600mb cannot be known with any accuracy. This
                                        is often the case on the western high plains and front range. The 500-200mb winds are fairly uniform and are known
                                        with much higher accuracy. The 6 km winds (400mb) were 55 to 65 kts from the west-southwest while the surface winds
                                        were from the east-southeast at 20 kts. Needless to say, the 0-6 km vertical wind shear was excellent. But the lack of
                                        wind information at low levels precludes the devleopment of an accurate hodograph. Given the storm motion, the
                                       storm relative flow was quite good at the surface and from 400mb to 200mb.

                                        To assess the shear profile, let's look at the 00 UTC upper air charts. Since the Scottsbluff tornado occurred at 22
                                        UTC, the wind profile was probably very similar to the 00 UTC raob winds. Since the 400-250mb winds were
                                        gradually increasing during the day, the wind speeds in this layer may have been slighly lower than the 00 UTC values.
                                       The following table contains wind speeds at various levels based on the 12 UTC and 00 UTC UA charts.


Level	850	700	600	500	400	300	250	200
12 UTC	XXX	XXX 11	XXX 4	23045 -8	24050 -19	22560 -35	22570 -46	23590 -54
00 UTC	XXX	XXX 15	XXX 5	23045 -8	24065 -20	23580 -34	22585 -43	23585 -53

                                        The 00 UTC 700mb chart shows light winds at Denver. However, this may be a local feature associated with the
                                        Denver cyclone. The surface observations from Laramie from 18 to 20 UTC (just upstream from the tornadic storm)
                                        show 20kt southwesterly surface winds gusting as high as 29kts. In the absence of any other clues, we will use
                                        22525kts as our 700 wind vevtor and 22530kt and 600mb wind wind vector.

                                        Therefore our wind profile probably looked something like this:

                                                870    11020    SFC
                                                850   12030
                                               700   22525    1.8 km
                                                600    22530    3.0 km
                                                500    23545    4.6 km
                                                400    24060    6.3 km
                                                300    23075    8.3 km
                                                250    22580   9.5 km
                                                200    23585   11.0 km

                                        Assuming the storm motion vector was 25030kt, the storm relative flow was impressive at all levels. The approximate
                                        hodograph shows a 0-3km srh of 363. Keep in mind that this is a very rough approximation. The actual values are
                                        very sensitive to wind direction and wind speed from just above the surface up to 600mb. This happens to be the layer
                                        that we have the least confidence in. I used a bunkers method storm motion of 25025kts. This is about 5 kts weaker
                                        than my expected storm speed. If 25kts is accurate, then the storm must have developed at 1900 UTC instead of
                                        1930 UTC. The Bunkers method did give the same storm direction as the actual estimated storm motion from the storm
                                        reports (250 deg).

                                        The surface based CAPE was fairly easy to approximate since this calculation is sensitive to qualtities that are known.
                                        The approximate sounding for Scottsbluff for 21 and 22 UTC shows 3500 j/kg.