DEPARTMENT OF COMMERCE
FREDERICK H. MUELLER, Secretary
WEATHER BUREAU
F. W. REICHELDERFER, Chief
MONTHLYWEATHER REVIEW
JAMES E. CASKEY, JR., Editor
Volume 87 JUNE 1959 Closed August 15, 1959
Number 6 Issued September 15, 1959
A SEQUENCE OF TORNADO DAMAGE PATTERNS
Ferguson Hall and Robert D. Brewer
US. Weather Bureau, Washington, D.C.
[Manuscript received April 6, 1959 revised June 8, 19591
ABSTRACT
1. INTRODUCTION Tallner of the Wisconsin State College at Eau Claire.
The system of tornadoes which struck west-central ]Vis-
consin on the afternoon of June 4, 1958, passed ovcr ex-
tensive forest areas, and left, :L sequence of damage pat-
terns perhaps unique in recent years.1 The storm system
as a whole produced at, least, five major damage paths rull-
ning in an east-nortlxasterly direction. The first and
longest, and the one which accounted for severe damage
to the village of Colfax, was 35 miles long, arld althongll
characteristically narrow over most of its extent, eventa-
ally widened into a. llh-mile path of almost total destruc-
tion. Each successive tornado of the system began to the
south of the previous one, and at, about the time the latter
was reaching the end of its course. The second of the
series produced a somewhat shorter path of about 17 miles,
but was equally violent and caused severe damage to the
northern outskirts of the city of Chippen--a Falls. The
paths are shown in figure 1.
2. DAMAGE PATTERNS
The damage patterns were recorded by a series of aerial
photographs obtained with the assistance of Prof. Gilbert
>Somewhat similar patterns were recorded in Europe b>- J. Letzmann [i l .
Some 130 photographs were taken along the paths of the
first and second tornadoes of the system. Over large seg-
lnents of the paths the forest cover was adequate to per-
Init, quite accurate reconstruction of the damage patterns,
even though, i n the intervening open fields, no sign of
tlwmnge w a s recognizable from the air.
The first tornado of the series was sighted as it touched
down at about 6 p m . south of the village of Woodville.
From visual reports it seems evident, that at this stage the
storm consisted of a distinct, and typical tornado funnel.
Ailmost immediately after reaching the ground it struck
the wooded area sho\.\-n in figure 2. Although the torna-
do% path was almost exactly centered over this wooded
section, it can be seen from the photograph that damage
was much more complete in the southern half (lower part
of picture), where the storm's translat,ional speed of about
50 1-n.p.1~ was added to the rotational motion (and of
course was subtracted in the northern half).* A recon-
strnction of this damage pattern and those immediately
following is sllo\vn in figure 3, where arrows represent
fallen tree directioas. I n many instances the trees lay
crossed over one another, as indicated in the sketch, and
-
2 See a discussion of this effect by G. W. Reynolds [2].
207
208 MONTHLY W E A T H E R R E V I E W JUNE 1959:
c
I
I J J
0 10 20 30 40
SCALE OF STATUTE MILES
FIUVRE 1.-Paths of major tornntlors of storill system.
in all such cases the sequence of directions in which the
trees fell corresponded to wllxt one \~--onlcl expect fro111 the
passage of a compact, cyclonically rotating vortex. As
can also be seen, the "reverse" flow along the northern side
of tlle path became much less prominent as the storm
progressed, suggesting that rotation had slowed somewhat
so that the translational speed was more effective in coun-
tering tlle backward-moving component, or that the trwns-
lation was much faster. This feature w a s an early indi-
cation of the highly variable character of the storm
patterns which was typical of the entire track.
Pronounced damage patterns were infrequent for tlle
next 8 miles of the path, and the storm may have been
on the ground only intermittently or may 1l:rve nlissxl
most of the heavily wooded areas. Howemr, at aljolrt
mile 7 another wooded tract was struck, apparently he:~d-
on, and again the pattern, as can be seen in figlue 4, leaves
unmistakable evidence of a well-formed and classical tor-
nado funnel.
Between miles 11 and 16 the character of the patlc1-11
seemed to change. There was almost none of the nxwks
of a rotating core that had been obvious earlier. Rather,
the predominant feature was what might be called a
JUSE 1959 MONTHLY WEATHER REVIEW 209
FIGURE 2.-Dmnnge pattern in forested area \\-here tornado first touehed down.
in direction. As one explanation, tlle funnel may have
continued toward the northeast,, but then turned sharply
southward and thence back on an east,ward course. hl-
ternatively, t.he original funnel may hare tlissolred as it
moved northeastward, to be replaced by anot,her forming
to the south, as was t.he case on a n~uch larger scale with
the individual storms of the system, as mentioned e,arlier.
Figure 7 is a photograph of the destruction which oc-
curred at point "A" of figure 6, and illustrates the cir-
cularity of the pattern. Also significant here is the sharp
tlernarcation between destroyed and undamaged forest.
After thus tleviating in its course, the storm crossed the
Red Cedar River and struck the village of Cedar Balls,
then continued its march to the east-northeast toward
Colfax, some 8 miles distant. The area here is intensively
farmed ancl the infrequent wooded sites in the storm'!
210 MONTHLY WEATHER IZEVIE\V
~"I G U K E 3. (At 1eft)"Reconstrncted damage
Imttern mer first segment of tornado path.
Arron-s represent fallen trees.
FIGURE 1. (J3elow"Damage pattern showing
distinct circularity.
JUNE 1959 MONTHT,Y WEATHER REVIEW 21 1
FIGURE 6.-Reconstructed pattern near midpoint of path.
212 MONTHLY WEATHER REVIEW JUNE 1959
FIGURE ‘?.-Circular pattern at point “A” of figure 6. S’ote sharp dernarcation between clamaged and undamaged trees.
path were, with two exceptions to be mentioned later? in-
adequate to define a clear-cut damage pattern. A t t h i s
time, however, the storm was described by witnesses, not
as a classical funnel, but rather as a huge bowl-shaped
protuberance about half a mile wide, that extended to the
ground from the low cloud deck of the squall line wit,h
which it was associated. Heavy rain preceded and ac-
companied the storm passage during this period.
The last 6 miles of the storm’s path (miles 29 to 3 5 ) are
pictured in figure 8. The shaded area is the built-up sec-
tion of the village of Colfax, the southeastern one-third
of which suffere.d almost complete destruction, with wide-
spread but more irregular damage elsewhere. Perhaps
the most impressive feature of the patterns is the sharp
lins of dernarcat,ion bet.ween the opposing flows from t,he
north and from the southwest. From place to place along
this line can also be noticed the confused and at times
circular patterns characteristic of high-speed vortex mo-
tion. Substantial alterations in pattern as the storm
progressed are quite noticeable and might indicate highly
JUNE 1959 MONTHLY WEATHER REVIEW 213
FIGURE 8.-Reconstructed pattern of final portion of path. Built-np portion of village of Colfax shaded.
unsteady storm structure, except that here the terrain is
.,quite rolling or even hilly, and nmy llave been sufficient
,.to distort the flow. Taken as a whole, however, t,he pat-
terns suggest a central high-speed funnel or vortex sw-
rounded by an intense cyclonic circulation giving clestruc-
tive winds out to a radius of % mile. The higllly con-
vergent appearance of the damage also suggests, however,
the presence of intense inflow toward the center of the
storm sufficient to distort the pattern normally produced
by a moving rotary system of winds. Figure 9 shows a
view of Colfax, looking northm-ard from the southern
village limits.
3. FLOW MODELS
Three idealized flow models were constructed to see
how well they might account for t,he actual damage pat-
terns. These consisted of-
(a ) a translation speed of 25 m.p.s. (56 n1.p.h.) com-
bined with a VR vortex in which the winds were set at
55 m.p.s. (123 m.p.11.) at a radius of 100 m. from the cen-
ter and decreased in proportion to increasing radius ;
(b ) translation combined with a sink with converging
windof 60 m.p.s. (134 m.p.h.) at the 100 m. radius and
decreasing with increasing radius in the same fashion ;
(e ) translation combined with both vortex and sink
(this of double the above st,rength).
The sequences of winds which would be experienced at
the surface during the passage of these systems are shown
in figure 10. Model (a) would undoubtedly account for
t.he patterns of both figure 3 and figure 6, where rotation
and translation seem predominant and the degree of “cir-
cularity” of the patt.ern would depend upon the forward
speed of the system relative to the intensity of the vortex
circulation. Thus t.he strongly circular patterns at the
beginning of figure 3, where the tornado first touched
ground, and at point “A” of figure 6 imply high vortex
speeds or low forward motion. I n the first case it seems
possible that the funnel may have built up to very high
speed just before touching the ground. Then due to sur-
face friction this speed may have been reduced to give
the translational type of pattern immediately follo\Ting
and continuing for the next mile or two. I n the case of
the circular patterns of figure 6, a marked deceleration of
the system would seem likely, accompanying the change
in direction or alternatively in connection with the sub-
stitution of a second funnel for the first. This furnishes
a possible explanation for the recurrence of circularity,
but does not necessarily account for the sharp demarca-
tion between damaged and undamaged forest noted in fig-
ure 7 . The latter suggest,s the funnel was more in the
nature of a spinning column of air, perhaps in solid rota-
tion, moving through a relatively undisturbed environ-
ment not. partaking of t.he tornado circulation. It is
possible that this particular vortex was indeed dying out,
as suggested prexiously, and mas merely “coasting,” sep-
arated from the VR c,irculation which was then attached
214 MONTHLY WEATHER REVIEW JUNE 1959
JUNE 1959 MONTHLY WEATHER R,EVIEW 21 5
I
",
C
t t t t l l t
Frc;r:n~ lO.--Hypothetical models of wind structure :
(a ) vortex + translation, (b ) sink + translation,
and (c) rortes + sink + translation.
model (a ), bnt nligllt rery we11 occur with 111oclels (0)
and (c ) near the axis of nlotion, provided tlle final inflow
speeds greatly exceedetl the translational a ~l d rotatio~lal
speeds. Thus, clue to the high rate of inflow, the wind
wonld blow back toward the approaching center for R
brief period before the center passed, to be followed, south
of the center, by strong flow in the direction of general
storm morement.
The ~~~s t e :~c l j ~ state of the storm during this f i d stage
is evident from the variability of d:xmage patterns shorv1l
r,l56OS-69"--2
i l l figure 8. For this re:tson 110 one hypotheticxl model
C :U ~ be, expected to provide an exact "fit" to the observed
1)atterns. Also, as has been mentioned, the rolling and
hilly topogr>xplly in this section may have had a decided
influence 011 resulting damage by producing distortion of
the wind f l o w .
h flow pattern of quite different character is illustrated
in figure 11, which shows a, distinct fan-shaped or "crom's-
foot" distribution of felled trees. This pattern occurred
at mile 26, and a similar one occurred a mile earlier. Over
216 MONTHLY WEATHER REVIEW JUXE 1959
FIGURE ll.--Fan-shaped damage pattern.
level terrain one would undoubtedly conclude that such a REFERENCES
(downdraft) and t,ranslation. In both of these cases, Beitackrift, rol. 42, Jan. 192.5, pp. 42-52.
however, the pattern occurred 0x1 hills (wooded portion 2, G. Reynolds, CoIllmon 7vind Damage I’attern in Relation
in fig. 11), leaving the possibility that it was due to the to the Classical Tornado,” Bulletin of the American Meteor-
terrain. olo{]ical Society, vol. 38, No. 1, Jan. 1987, pp. 1-5.
patter’’ ’vas produced by a combination a Source 1. .J. Letzn~ann, ”Fortschreitende Luftwirbel, “Veteorozogi.Ww