DEPARTMENT OF COMMERCE
CHARLESSAWYER, Secretary
WEATHER BUREAU
F. W.REICHELDERFER, Chief
MONTHLY WEATHER REVIEW
Editor, JAMES E. CASKEY, Jr,
Volume 79 Number 4 APRIL 1951 Closed June 15,1951 Issued July 15,1951
ON THE DESICCATION OF’ A CLOUD BANK BY A PROPAGATED
PRESSURE WAVE
MORRISTEPPER
U. S. Weather Bureau, Washington, D. C.
[Manuscript received February 19, 19511
ABSTRACT
On December 6, 1949, an accelerated microbarograph in Washington, D. C., produced a unique pressure trace at
the time that a peculiar cloud bank passed overhead. The rear (western) edge of this cloud bank ended very abruptly
and had the appearance of an Antarctic ice barrier. A study of the synoptic condit)ions of the day revealed that this
pressure variation was produced by an expansion-type pressure wave which travelled eastward from the Midwest
and was propagated between two inversion surfaces with a speed far in excess of the wind speed in that layer. The
cloud bank also travelled in this same atmospheric layer but was advected by the prevailing winds. The conclusion
is reached that the pressure wave which progressively overtook the cloud bank was associated with a rapid drop of the
isentropic surfaces which in turn produced the adiabatic heating necessary to desiccate the cloud partially in the Mid-
west and completely over Washington. Confirmation of this conclusion is found in solar radiation records.
INTRODUCTION
This study was motivated by two events which occurred
simultaneously in Washington, D. C., on December 6,
1949. On the afternoon of that day a cloud bank ap-
peared over the city. At about 1615 EST the cloud
ended abruptly-like the pictures of an Antarctic ice cliff.
This feature is shown in figure 1. This picture was taken
facing eastward, so that the sun in the westtern skies
brilliantly illuminated the edge of the cloud bank and
showed it to be very thin. The author estimated it a t
the time as no more than 300-500 feet thick. This phe-
nomenon traveled very rapidly eastward. The second
phenomenon, which occurred a t the same time that the
western edge of the cloud bank passed overhead, was a
pressure pulse which produced the barogram shown in
figure 2. The microbarograph which recorded this trace
is one with accelerated gears which records continuously
on an attached Esterline-Angus Recorder. By expanding
the time scale, finer detail in the variation of the pressure
may be studied. This trace shows that with the passage
1 Paper presented at the 10Sth National Meeting of the American Metcorological
Society, New York City, January 30, 1951.
949984-51-1
of the “chopped off” portion of the cloud bank, the pres-
sure dropped very suddenly and in a short time recovered
suddenly. The simultaneity of the peculiar behavior
of the cloud bank and the pressure trace suggested the
possibility of some relationship between them. This
study was a search for this relationship.
BAROGRAM INVESTIGATION
Thc history of the pressure pulse was studied through
an investigation of the barograms for all the stations in
the general neighborhood of Washington, D. C. Portions
of t’hese barograms are reproduced near their respective
stations in figure 3 . Care was taken in copying these
traces to keep all the sheets oriented in exactly the same
manner so that the pressure profiles of the stations and
their slopes might be compared. The peculiar V-shaped
pulse which was noted on the accelerated microbarograph
a t Washington could be traced back as far as western
Pennsylvania, but not below the latitude of Washington,
D. C. Furthermore, as western Pennsylvania was ap-
proached, the “first” part of the V-shaped pulse (i. e.
the pressure fall) became the more pronounced while the
rise diminished in amplitude. Continuing westward still
61
62 MONTHLY WEATHER REVIEW APRIL 1951
farther, one notices that this pronounced fall could be
followed to western Indiana. I n the Ohio-Indiana area,
this fall was part of an over-all, but gentler pressure fall.
Isochrones of the passage of the V-shaped pulse to the
east and the marked pressure fall to the west were drawn
as indicated in figure 3. For the sake of uniformity, the
time of onset of the fall was chosen throughout as the
time designator for the isochrone. The two significant
features of this isochrone pattern are:
(1) The pressure disturbance (which hereafter shall
refer to both the V-shaped pulse to the east and the pres-
sure drop to the west) can be followed only through the
northern portion of the area studied. The southern por-
tion shows gradual prolonged falls (particularly in the
west) but no indications of a disturbance such as expe-
rienced farther north.
(2) The speed of the disturbance was approximately:
43 m. p. h. during 0800-1100 EST (in Indiana)
77 m. p. h. during 1100-1400 EST (in Ohio)
77 m. p. h. during 1400-1700 EST (in southern Penn-
sylvania and Maryland)
It should be indicated in passing that due to current
practices of pressure recording by means of barographs
the isochrones can be considered accurate only to about
CLOUD INVESTIGATION
A study of the Surface Weather Observations (W. B.
Form 1130) for the stations in the area, revealed the
following progression of cloud formations. The early part
of the day was characterized either by clear skies or by the
presence of thin, high clouds above 20,000 feet. Later in
the day, each station reported the appearance of low
clouds with bases varying from 5,000-10,000 feet. Finally,
and this has only incidental significance for our purposes,
most of the stations reported lower overcast. Our pri-
mary concern will be with the initial appearance of the
one-half hour. FICTJRE 5.-Barogram, December G, 1949, Washington, D. C
APSIL 1951 MONTHLY WEATHER KEVIEW 63
FIGURE 3.-Barograms, December 6 , 1949, showing isochrones of pressure drop.
64 M O N T H L Y W E A T H E R APRIL 1951
low cloud deck. The time of this initial appearance was
plotted on a map and again isochrones were drawn (fig. 4).
The principal features of this isochrone pattern are:
(1) The isochrone pattern indicates a uniform progres-
sion of the onset of the cloud bank in a general west to
east direction.
(2) The average speed of the isochrones is approxi-
mately:
40 m. p. h. between 0500 and 0800 EST (* in eastern
50 m. p. h. between 0800 and 1100 EST (in central
35 m. p. h. between 1100 and 1400 EST (in western
35 m. p. h. between 1400 and 1700 EST (in south
Indiana)
Ohio)
Pennsylvania and western Maryland)
central Pennsylvania and eastern Maryland)
Again, it should be kept in mind that weather observa-
tion practices cannot permit the consideration of the
positions of these isochrones accurate to more than about
one-half hour.
COMPARISON OF ISOCHRONE PATTERNS
A comparison of the pressure disturbance isochrone
pattern with the onset of the cloud deck isochrone pattern
shown in figures 3 and 4 reveals:
(1) East of Indiana the leading edge of the cloud bank
travelled significantly more slowly than the pressure
disturbance.
(2) The leading edge of the cloud bank preceded the
pressure disturbance in time over most of the area.
(3) The pressure disturbance, travelling faster than the
leading edge of the cloud bank, overtook the latter across
a line roughly bisecting the State of Pennsylvania from
the northwest corner to the southeast corner.
HOURLY SYNOPTIC MAPS
Prom the transmitted teletype data it was possible to
plot hourly synoptic maps and analyze them. Four such
maps drawn for 2-hourly intervals beginning with 1030
EST are shown in figure 5. The main synoptic feature
on these maps is an open wave moving slowly through
central Iowa and northern Illinois. The associated warm
front stretches generally southeastward from the center
of the wave. To the east, there is a pressure ridge which
is also moving eastward. The position of the leading
edge of the cloud bank is given by a line of open circles
while the position of the pressure pulse is given by a
thick line.
As may be expected, the lcading edgc of the lower cloud
bank seems to coincide with the limit of the circulation
around the wave. Wc may then attribute the cloudiness
FIGURE B.--$urface synoptic maps for 1030, 1230, 1430, and 1630 EST, December 6, 1949. Synoptic position of leading edge of cloud bank is indicated by line of open circles and the
synoptic position of the pressure pulse by a thick solid line.
APRIL 1951 M O N T H L Y W E A T H 65
t,o overrunning or convergcnce which precedes a cyclonc.
On the other hand, the period of relatively clear skies may
be associated with the anticyclonic circulation around
the ridge.
The hourly synoptic maps graphically illustrate how
the pressure disturbance travelled through the surface
system.As the dist*urbance travelled away from the,
circulation around the wave and closer lo the region dom-
inated by the pressure ridge, its charact'er changed from
a pressure drop to a V-shaped pulse as identified in
Pennsylvania, Maryland, and Washington, D. C. An
explanation for this relationship will be given in a sub-
sequent section.
It should be noted that throughout, the surface position
of the warm front designates the southern limit for tjhe
pressure pulse isochrones.
WINDS ALOFT
It is, of course, of considerable importance to investigate
whether or not advective processes could account for the
propagation of either the pressure disturbance, the cloud
bank movement, or both. Table 1, the wind data for
1000 EST and 1600 EST, shows that winds below 3 km.
could have advected the cloud bank but not the pressure
disturbance, while above 3 km. the winds were strong
enough to have advected either or both. However, evi-
dence will be presented in the next sect'ion to indicate
that the disturbance was probably propagated below 3 km.
and as such could not have been the result of an advective
process.
CROSS SECTION
From the atmospheric soundings taken a t Joliet, Toledo,
Pittsburgh, and Washington, it was possible to construct
a vertical cross section through the atmosphere for the
time 1000 EST (fig. 6). At this time t.he pressure pulse
was in a position indicated by the arrow and the leading
edge of the lower cloud bank in the position indicated
by the upright triangle. The wind data for Fort Wayne,
Toledo, Columbus, Akron, Elkins, and Washingt'on were
superposed on the section. The winds were rotated in
the plotting to conform to the orientation of the cross
section. The important features of the cross section are:
(1) Iscntropes: Above 4 km. the isentropic surfaces are
prstty much horizontal and the atmosphere in this region
represent>s quasi-horizontal homogeneity. Below 4 km.
and cast of Toledo t'he same situation prevails. The
isentropic surfaces are again roughly horizontal. How-
ever, between the Toledo and Joliet soundings there is a
marked change in the isentropic surfaces below 4 km.
Interpreting this space cross section as reflecting the
changes that would appear on a time cross section we may
conclude that the potential temperatures indicate that
subsequent warming has tjaken place. It should be noted
t h a t t h e isentropic surfaces were drawn to represent
extremely mpid dropping of the isentropic surfaces in the
area of the pressure pulse only, rather than a gradual
descent from the values a t Toledo to the values a t Joliet.
The reason for this analysis will be clear later.
(2 ) Inversions: The thick vertical lines represent ele-
vations for which the soundings showed inversions of tem-
TABLE l.--Winds aloft (m p h ) at 1000 EST and 1600 EST, December 6, 1949
Altitude Surf. 1 500 1 1000 I 1500 1 2000 I 2500 1 3000 1 4000 5000 I 6000 1 7000 I 8000 19000meters
Station \ \ 1000 EST
Evansville, Ind.. ..........
Indianapolis, Ind Fort Wayne, Ind- . -. .- .. _.
Akron, Ohio ................
Cincinnati, Ohio- ..........
Toledo, Odio.. Columbus Ohio
Harrisbure. Pa-
Curwensville. Pa.".-
...........
............
.............
......
............
Philadelpiiia, Pa. ..........
Elkins, W. Va .............
Norfolk, Va.." ............
Richmond, Va. ............
Roanoke, Va..". ..........
Washington, D. C ..........
SSW 16
SE 11
S S E 13 s A
SSE 9 SSE 5
NW 13 SE 9
WNW 18 W 16
WNW 11 WSW 9
WSW 9 W 9
WSW 16
WSW 58 SW 29
S W 26
I I I
WSW 56 ........................................................................................
WSW 36 WSW 43 .............................................................................
WSW 43 ........................................................................................
I
W 36
1
WNW45
I
W 45
I
1%' 81
1
WNW96
I
..........
/
..........
1
.......... ~::1:1:::::: WSW 43 ............................................................................. 7 x 1 ", 7x7 >n
S W 40
S SW 34 SSW 29 S 36 SE 31
WSW63 1
WSW 36 WSW 36 SSW 27 7 W 11 w 22
SSE 18 SW 18 W
S E 13 SSW 20 SW 20 1%'
WNW 22 WNW22 WNW20 ."......
W 16 NW 29 WNW43
.~~ .____._.. W 13 WNW31
W 31 W 20 NW 25
WSW 9 WNW16 WNW36 WSW 22 W 16 WNW22
W 18 WNW27 WNW49
"
4s
WNW31 WNW40 WNW49 WNW56 WNW63 27 1 '" i?
~
3 38 1 ............
Evansville, Ind. ...........
Fort Wayne, Ind.". .......
. Indianapolis. Ind.. ........
Akron, Ohio-..- ...........
Cincinnati Ohio ...........
Columbus.' Ohio.. .........
Toledo, Ohio..: ............
Curwensville, Pa ...........
Harrisburg Pa..-. . -. _. ~.
Philadelphia, Pa ...........
Elkins, W. Va .............
Norfolk, Va ................
Richmond, Va. ............
Roanoke, Va ...............
Washington, D. C .........
SSW 18
S 11 SE 9 SSW 25
ESE 20 SW 7 WSW 5
WNW 9 s 5 WSW 9
s w 5
SW 9
WSW 2
..........
..........
ssw
S W SE
ssw
SE
W WNW
WSW STV S W SW
. "" ~.._
.........
..~ .""
........
..........
...........
...........
S W 9
S W 56
43 S W 45
34 ". -. . -. - -. 18
SW 54 38
S 29
29
...............
...............
.....................
._..
WNW36 W 25 20
WNW34 W 16 9 WNWlR WSW 5
._" SW 18 WSW 38
20 WSW 20
WSW 13 9 WNW31
WNW34 WSW 20 11 W S W 29 SW 22 16 W 27
......................
W N W 5 8
NW 56 NW 56 WNW56 WNW5O
NW 58 NW 44 NW 50 NW 40 WNW47 WNW45 N W 67
..........
WNW67
N W 74
WNW 67 NW 60
WNW 54
NW 67
..................................................................
......................
WNW72 WNW76
WNW 87 WNWXi WNW85 WNW81 WNW81 WNWi4 WNW154 WNW98 WNWl19 WNWl16 WNW98 WNW81
WNW105 WNW87 W 90
.......................................................
WNW78 1 WNW107 ...................... WNWll2 WNW114
3 ,
1600 E S T
"
...................................................................................................
...................................................................................................
...................................................................................................
SW 27 ........................................................................................
...................................................................................................
..................................................................................................
...........
WNW'40 WN W 49
WNW47
w 47 NW 38 WNW63
W 38 WNW49
..........
WNW49
WNW 52 WNW49
WNW47 W 49
W iYW 60
W 47 WNW 55
.............................................................................
WhTW60
................................. W 90 W X i W 87 WNW52
..................................................................
.............................................................................
.............................................................................
.............................................................................
W 58
.................................................................. W 56
....................................................... W 65
.............................................................................
66 MONTHLP WEATHER REVIEW APRIL 1951
-
-m
-6500
-6OOC
-5500
-5000
-4500
-4OOC
L
Y
-3500
-3OOC
-2500
-2ooc
- I500
- 1000
- 500
-a
-
-33.0.313 -34.0.312 -32.5. 313 -33 0 313 M -/05 -r //.?
I TEMP - POT. TEMR
PLOTTING MODEL
I I MIX.RATIO- REL.HUM. I
-22.3. 308 M
"20 0 309 M
I
-12 5.306 "142. 305
-
53 24.
-11.7 303 .65 25
-78
4
-10.4.304 77 27
-11.3 e300
"-81 -76
B3 32
/4 3 -45
JOT 4 FWA TOL CMH AKR PIT EKN
I
I
FIGURE 6.-Atmospheric cross section, Joliet-Washington, 1000 EST, Dccernber 6,1949. Solid lines are isentropes (" A), dotted lines are lines of equal mixing ratio (gm./kg.), upright
rectangles are temperature inversions. Along the bottom of the chart, arrow represents location of pressure pulse and upright triangle the leading edge of the cloud bank. The
winds (in m. p. h.) are plotted with reference to the directional arrow in the top center. JOT=Joliet, FWA=Port Wayne, TOL=Toledo, CMH=Columbus, AKR=Akron,
PIT=Pittsburgh, EKN=Elkins, DCA=Washinglon.
perature. It should be noted that from Toledo and east-
ward there are two distinct inversion surfaces; the lower
a t approximately 1,000 m. and the upper one a t approxi-
mately 2,000-3,000 m. The lower inversion was associ-
ated with the warm front to the south. At Jolict, only
the lower inversion is present and it has lowered between
300 and 500 m. The upper inversion has been almost
eliminated undoubtedly due to the warming t,hat has been
discussed above. (A remnant of the upper inversion may
still be found over Joliet a t 2 km.)
(3) Moisture and clouds: The cloud bank has been
drawn in the rcgions where t)he relative humidity was
greater than 70 percent. I t should bc noted that the
cloud bank appears t'o travel in between the two inversion
layers. At t'he point where the isentropic surfaces were
drawn to dip down very sharply, thc cloud pattern was
drawn in a manner t'o conform with this adiabatic heating.
Similarly, the isentropic surfaces were lifted over Joliet
which reported higher humidities. This feature will be
cliscussed in greater detail in the subsequent sections.
APRIL 1951 MONTHLY TJrEATHER REVIEW 67
(4) Winds: Below the lower inversion t'he winds are
predominately easterly. Between the two inversions they
change to westerly and are of the same speed as t'hc spced
of propagation of the cloud bank. Above the second
inversion thc winds are very strong, but here nothing
seems to be happening. It is of particular interest that
the winds over Fort Wayne, below the lower inversion
are much faster (easterlies) than thc winds a t Tolctlo,
Columbus, or Akron.
EASTERLY WIND COMPONENTS
This increase in t,hc easterly wind component may bc
seen in figure 7. Here the reported winds a t various
stations have been broken down into their components
and the easterly component plott,ed with respect t'o position
of the passage of the pressure pulse. It can be sccm that
there are definite indications of an increase in the easterly
wind component associated with the passage of the prcssure
pulse. The wind values plotted arc not continuous but
represent discrete wind observations taken a t relatively
long-period int'ervals. Continuous records might have
indicated this behavior much more clearly.
HYDRAULIC EXPERIMENT
In setting up a hypothesis to explain the phenomena
described above, it will be helpful first to consider the fol-
lowing laborat>ory experiment (fig. 8 A, R, C ). Water is
made to flow in a t'rough (60 by 6 by 2 inches) from right
t'o left. By adjusting the opening at the left a constant
level of water is maintained (fig. SA). (The size of the
opening is controlled by sliding a plate up or down.)
The plate is suddenly raised (fig. SB) allowing the water
to accelera,te to the left. This produces a depression of
the water surface which trovels to the right. Thc plate is
suddenly lowered (fig. SC) to its original position and the
I I I I I I I I I I I I
I
-6 -5 -4 -3 -2 -I +I +2 +3 +4 +5 +6HRS
I Terre Haute. Ind.
+I O j E &
F i n d l o y , 0 2
-0 a 5 0 Clevelond, Ohio
Wheeling, W. Vo.
-+lo
"IO w
Altoono, Po. 0
Washington, D.C.
-20 -I0 j
FKIJRE 7,"Easterly wind component (m.p.h.) in relatlon to the passage of
the pressure pulse.
FIGURE S.-Hydraulic experiment. (A) Strady st,ate conditions; the water is flowing
from right to left and the level of the water is maintained constant. (B) The plate a t
the left is lifted producing a drop of the water surface. The wave moves to the right.
(C ) The plate at the left is lowered to its original position producing an elevation of
the water surface. This wave also moves to the right. (The bottom surface of the
trough is perfectly smooth. The apparent roughness is due to paint along the edge of
the glass.)
water which has been accelerating piles up and raises the
water surface. This elevation of the water surface also
travels to the right behind the depression.
By analogy to gas flow,we may refer to the traveling
depression of the water surface as an expansion type wave
and t,he subsequent elevation of the water surface as a
compression type wave. It is emphasized that both the
expansion and the compression type waves travel in a direc-
tion opposite to that of the flow of the water.
68 MONTHLY WEATHER REVIEW APRIL 1951
THE HYPOTHESIS
Freeman has indicated the conditions under which the
flow of air under an inversion surface may be considered
analogous to the flow of water in an open channel. In this
analogy, the inversion surface corresponds to the free sur-
face of the water, and variations of the height of the in-
version surface behave in a manner similar to water waves.
In particular, waves on the inversion surface would be
propagated independent of the flow of the air particles
themselves.
We postulate then, that some mechanism accelerat,ed the
air below the warm frontal inversion. As in the case of
the water in the laboratory experiment the inversion sur-
face dropped suddenly, producing an expansion-type wave
which moved eastward. The second phase of the pressure
pulse is associated with the compression-type wave pro-
duced as the accelerating air is slowed down by convcrg-
ence with the air ahead of it. It would be expected from
hydrostatic reasoning alone, that the expansion-type wave
would be reflected as a pressure drop at thc surface and the
compression-type as a pressure rise. This combination of
pressure fall followed by a pressure rise when superposed
on a falling pressure field would produce the type of t,race
characteristic of the Midwest. When this pulse reachcd
the flat pressure area in the East it appeared primarily as a
V-shaped pulse.
Now, associated with the rapid fall of the inversion sur-
face there would be rapid heating and consequent desicca-
tion of part or all of the cloud. We should then expect
that when the pressure wave traverses an area where the
cloud is relatively thick that its power to desiccate the
cloud completely would be less than in those places where
the cloud was relatively thin.
Returning to figure 1 which shows the picture of the
cloud bank over Washington, we may state that the pres-
sure wave had just caught up to the cloud and because the
cloud was very thin, the pressure wave was able to desic-
cate the cloud completely.
SOLAR RADIATION
While the physical int’erpretation given above explains
the peculiar phenomenon that passed Washington, D. C.
on December 6, 1949, it also raises the obvious question,
did this phenomenon happen only a t Washington or did
it occur elsewhere? The answer is that the basic phe-
nomenon did occur elsewhere but on a different scale.
The evidence for this statement lies in solar radiation data.
We shall assume here t’hat sudden changes in transmis-
sion are probably caused by sudden changes in the thick-
ness of the intervening cloud layer. While it is conceded
that changes in transmission may also be caused by
changes in the water content and/or drop size in the cloud
layer, it will be safe to assume that the sudden changes in -
Flow”, Journal 01 Meteorology, vol. 5, No. 4, August 1948, pp. 138-146.
2 J. C . Freeman, Jr., “An Analogy Between Equatorial Easterlies and Supersonic Gas
transmission that will be shown were probably caused by
decrease in thickness of the cloud layer.
Figure 9 shows the radiation traces from Indianapolis,
Columbus, Put-in-Bay, Cleveland, and Washington.
For comparison the radiation trace for a relatively clear
day at each station is shown by the dashed line.
The general feature of all these traces is that they con-
tain two significant points. The first is a decrease in
radiation and the second is a very sharp increase in radia-
tion. Some of the time scales are in Solar Time and so a
conversion to Eastern Standard Time is required. This
conversion factor is shown where required. The time of
occurrence of each of the two significant features from
each trace was recorded and compared with the isochrone
pattern for the onset of the low cloud bank and the pas-
sage of the pressure pulse. This comparison is shown in
table 2 . The two columns refer to the time of decrease
in radiation and the time of sudden increase in radiation.
The numbers in parentheses refer to the time that the
lowcr cloud bank appeared over the station as reported on
the Surface Weather Observation Sheet. The numbers
in brackets refer to the time of the passage of the pressure
pulse over the station as recorded on the microbarograph
trace.
Considering the errors that may arise in reading the
traces, the difference in location of the instruments at
any one city, and the possible lack of accurate time con-
siderations in taking observations, the agreement in the
time values is remarkable. We may safely associate the
decrease in radiation with the onset of the cloud bank and
the sudden increase in radiation with the passage of the
pressure pulse.
TABLE 2.-Times of sudden. decrease and increase in solar radiation.
Figures in parentheses refer to time of arrival of the cloud bank
and jigures in brackets to time of passage of pressure pulse
Time (EST)
Decrease Increase
Indianapolis, Ind .-..
~~~~~~~~~~
........ ~~.~~~~ .........
1602 [1610] 1455(1455) \\’ashington, D. C-” ........ ~~~~.~~ ......... ~~~~~~~...
1315 113201 *1200 (1203) Cleveland, Ohio ..... ~~~~.~~~~~~ ......... ~~~.~~~ .......
1148 [1215] 0914 (0925) Columhus, Ohio __... ~~.~~~~~-~~~ ........... ~.~~~ .....
1240 [12301 0947(1000) Put-in-Bay, Ohio.. . ....
~
......... ~~~.~.~ ........- ~~-.
0922[0920] ..._ (0500)
scattered Cu. which were reported on the Surface Weather Observations Form beginning
‘The fluctuations in radiation at Cleveland, beginning with 0950 are related to a few
at 0000. Thrsc clouds were reported under “Remarks.” Lower clouds were not reported
under “Sky Conditions” until 1203 and this entry was undoubtedly associated with the
continuous decrease in radiation beginning at 1200. Surface Weather Observations
Forms are not available for Put-in-Bay so that it is not possible to determine whether
of tho cloud bank or scattered clouds unrelated to the cloud bank. The records from
the fluctuating transmission after 0927 was due to scattered clouds at the leading edge
Toledo seem to indicate 0900 as the time when the leading edge of the cloud bank moved
in at Tolcdo. It is for this reason that 0947 was chosen as appropriate for Put-in-Bay.
CONCLUSION
We conclude from the evidence presented in this study
t’hat a pressure wave consisting of a pressure drop followed
APRIL 1951 MONTHLY WEATHER REVIEW 69
._ . I
I
/i
,."
z I
E
14 12
'\
FIGURE 9.-Radiation traces for Washington, D. C., Cleveland, Ohio, Put-in-Bay, Ohio, Columbus, Ohio, and Indianapolis, Ind., December 6, 1949. A relatively clear-day radi-
ation trace is indicated by the dashed line.
70 M O N T H L Y W E A T H E APRIL 1951
by a pressure rise, was propagated on the warm frontal isontropic surface was associated with a partial (in the
surface with a speed far in excess of the winds in the lower Midwest) and total (over Washington) desiccation of t'he
layers. This wave was produced by sudden acceleration cloud layer. We lcave unanswered the question: what
in the winds below the frontal surface. The pressure is the mechanism that produced t'lle sudden acceleration
drop which is but an indication of the sudden fall in the of tho winds below t'he inversion surface?