A Concentrated Outbreak of Tornadoes, Downbursts and Microbursts, and Implications Regarding Vortex Classification G REGORY S. F ORBES Department of Meteorology, The Pennsylvania State University, University Park, PA 16802 R OGER M . W AKIMOTO Department ofthe Geophysical Sciences, The University of Chicago, Chicago, IL 60637 Reprinted from MONTHLY WEATH ER REVIEW, Vol. 111, No. I, January 1983 American Metcoro1ogical Society Printed in U. S. A. Reprinted from MONTHLY WEATHER REV IEW, Vol. 111, No. I, January 1983 American Meteorological Socie1 y Prin«d in U. S. A. A Concentrated Outbreak of Tornadoes, Downbursts and Microbursts, and Implications Regarding Vortex Classification GREGORY S. FORBES Department of Meteorology, The Pennsylvania State University, University Park, PA 16802 R OGER M. W AKIMOTO Department of the Geophysical Sciences, The University of Chicago, Chicago, IL 60637 (Manuscript received 6 November 1981 , in final form 13 August 1982) ABSTRACT A remarkable case of severe weather occurred near Springfield, Illinois on 6 August 1977. Aerial and ground surveys revealed that 17 cyclonic vortices, an anticyclonic vortex, I 0 down bursts and 19 microbursts occurred in a limited (20 km X 40 km) area, associated with a bow-shaped radar echo. About half of the vortices appeared to have occurred along a gust front. Some of the others appear to have occurred within the circulation of a mesocyclone accompanying the bow echo, but these vortices seem to have developed specifically in response to localized boundary-layer vorticity generation associated with horizontal and vertical wind shears on the periphery of microbursts. Some of these vortices, and other destructive vortices in the literature, do not qualify as tornadoes as defined in the Glossary ofMeteorology. A more pragmatic defi nition of a tornado is suggested. 1. Introduction Burgess and Davies-Jones, 1979; Burgess and Don­ aldson, 1979; Lemon et al., 1980). In this paper we A significant research effort in recent years has in­ present additional evidence of tornadoes associated volved the development of Doppler radar techniques with a gust front and with downbursts. We also pre­ to identify thunderstorms which produce tornadoes. sent evidence which suggests that some tornadoes T he efforts have been rather successful, identifying may be associated with microbursts. the mesocyclone and tornado vortex signatures as Fujita ( l 976b) originated the term " downburst" to indicators of storms which produce major tornadoes describe the intense downdraft involved in an air­ (Lemon et al., 1977; Burgess and Devore, 1979). plane crash. In association with damage near the Lemon and Doswell ( 1979) have described the de­ ground, Fujita ( 1978) defined the down burst as a velopment of these tornadoes. "strong downdraft inducing an outward burst ofd am­ Not every tornado which develops is associated aging winds on or near the ground." Microbursts are with a thunderstorm possessing a mesocyclone sig­ small downbursts with horizontal dimensions less nature, however. Burgess and Donaldson ( 1979) than 4 km (Fujita, 1981 ). found that several weak and short-lived tornadoes The damage paths of 8 of the I 0 down bursts, 18 occurred in developing echoes without detectable of the 19 microbursts, and 18 tornadoes which oc­ mesocyclone circulations or supercell characteristics. curred on 6 August 1977 are shown in Fig. l. The Later in their lifetimes these echoes developed me­ remaining downbursts and microbursts occurred be­ socyclones and strong tornadoes. Weak tornadoes yond the east and west edges of the figure. The paths also can form outside of the mesocyclone circulation were located usi ng techniques described in Section along the gust front and flanking li ne of a supercell 2. Description of the damage is presented in Sections thunderstorm (Burgess et al., 1977; Brandes, 1978, 3-5. A complete report on this case study (including 198 1) and along gust fronts and downbursts from 40 damage photographs) is given by Forbes and Wak­ non-supercell thunderstorms (Burgess and Donald­ imoto ( 1978). son, 1979; Fujita, 1979; Wilson et al., 1980; Testud The presence or absence of damaging winds was et al., 1980). Additionally, weak tornadoes can form determined essentially unambiguously over the entire under a flanking cloud line behind or to the right of region shown in Fig. 1, as the area was extensively the main cumulonimbus, where radar echoes are covered by 2 m high corn (readily susceptible to dam­ weak or absent (Bates, 1968; Barnum et al., 1970; age). The paths of the 18 tornadoes were unmistak- © 1983 American Meteorological Society N N SPRINGFIELD TORNADOES and DOWNBURSTS August 6, 1977 -::::-- 4::- ~- 9§::::::: m7~ ~ · Na5-..•• ..,._.....-.-.- --... 0ol'l'IOQf DOWNIUlltST MICfll09URST F·KOI• Conto11u TORHAOO Sr.ietiot1 Vorlk n Sfltoks , I f 1 I ! I ~ IO I 12 14 JC1i.,..• ••~ 1 0 I t .S 4 S • l' e t WdH ~ 0801•1 ~ 3:: 0 / / z / ,... -· -l :i:: V Edinburg r -< / :;: m > -l :i:: m IS O ~~7· / ;i:i ~ V Kincoid .,, ;i:i ... m < .,.. o m A :;: L•Y'" ~.O· CHRISTIAN COUNTY M 0 R GA N ....CJ n., .. c!}• .. - ,. ---------- - --- MAC-----OUPI N-- --- - -----0~.~ .,~ >tS<J 9':l00' 8</YJ' I et/!O' ot.'ISf FIG. I. Maoo;,__ ,..1g ot the damage from tornadoes (identified by "No."), downbursts (identified by large numeral), and microbursts (identified by "m") on 6 August 1977. "Streamlines" of damage and F-scale contours are mapped. See figure legend for additional symbol explanation. < 0 r :s:c m JANUARY 1983 GREG 0 RY S. FORBES AND R 0 GER M . WA KIMOTO 222 ably clear. Thus, it is noteworthy that 18 tornadoes these are based upon experience gained during past occurred within a 20 km X 40 km area. This tornado damage surveys performed by the authors, studies by density is roughly a factor of I 0 greater than that of Fujita ( 1978), and discussions with farmers and agri­ the 3 April 1974 "super outbreak" (Fujita, 1975). The culture experts. A rule of thumb invoked during the average spacing between tornadoes 1, 4, 6 and 8 of 6 August 1977 surveys was that some corn is damaged Fig. I, occurring essentially simultaneously, was by winds at the low end of the FO category ( - 20 m 1 about 4 km, whereas the lateral separation between s- ). and damage becomes extensive as the winds ap­ simultaneously-occurring tornadoes on 3 April 1974 proach Fl. was about 45 km. Clearly, the tornadoes of 6 August If the tornado path contains suction vortex swaths, 1977 were not part of a typical outbreak of family then the speed of revolution of suction vortices about tornadoes, such as those presented by Fujita et al. the tornado axis can be estimated from the swath (1970), Fujita (1974) and Galway (1981 ). In fact, the shape, if the speed of tornado translation U can be density of vortex occurrences on 6 August 1977 raises estimated. Using the technique presented by Fujita some questions about the criteria for tornado clas­ et al. (l 970), the shape of the swath (cycloid) deter­ sification. These criteria are disc ussed in Section 6. mines the ratio of revolution and translation veloci­ 1 ties, vu- • The maximum wind speed occurs on the right side of the tornado (facing in the direction to­ 2. Techniques for classifying damage from tornadoes, ward which the tornado is translating) and is ap­ downbursts and microbursts proximately V + U, neglecting radial velocity and neglecting the speed of winds rotating about the axis Damaging wind speeds on 6 August 1977 were of the suction vortex. Forbes ( 1978) calculated speeds estimated using the Fujita ( 1973) scale, referred to as of rotation averaging 39 m s- 1 in well-developed suc­ F-scale. Fujita ( 1981) presents details of the estima­ tion vortices, however. Thus, total wind speed may 1 tion off-scale on the basis of damage to trees, vehicles exceed V + U by more than 39 m s- • 1 and structures. For example, FO (18-32 m s- ) winds Tornado 9 was classified as F3, on the basis of both produce some damage to chimneys, antennas, bill­ damage to structures and the shape of looping marks boards and tree branches. A few shallow-rooted trees in the suction swath patterns. These loops were more 1 may be uprooted. Fl (33-49 m s- ) winds normally broad and more pronounced than in any of the other damage roofs, overturn mobile homes and uproot tornadoes, and suggested that vu- 1 = 3. With a trees. Some trees may be snapped. Once winds reach translation speed of about 15 m s- 1 and the rotation 1 1 F3 category (70-92 m s- ) , roofs are removed and about the suction vortex axis estimated at 20 m s- , some walls are torn off well-constructed frame homes, the maximum wind speed in tornado 9 was estimated and most trees are uprooted or snapped. as 80 m .s- 1 (sum of translation, revolution and suc­ Tornado path lengths and path widths were based tion vortex rotation). Looping marks in suction upon the extent of the FO damage. Scales for path swaths observed by the authors on other occasions length and average path width were assigned using suggest that vu- 1 can be as large as 3 or 4 in tor­ the Pearson scales (Fujita, 1973). All reported tor­ nadoes translating 20 m s- 1 or slightly faster. Thus, nadoes are now rated by the National Severe Storms extreme structural damage is not necessary for cate­ Forecast Center using these Fujita and Pearson scales, gorizing a tornado as F4 or even F5, though present the FPP scheme (Kelly et al., 1978). tornado statistics certainly contain that classification Wind speeds deduced on the basis of damage ap­ bias. pearance must be considered only estimates of the In the absence of suction vortex swaths, easterl y true wind speed.