Anons~ 1955 MONTHLYWEATHER REVIEW 181
SURFACE STREAMLINES ASSOCIATED WITH THE TORRENTIAL RAINS
OF AUGUST 18-19, 1955, IN THE NORTHEASTERN UNITED STATES
CONRAD P. MOOK
U. S. Weather Bureau, Washington National Airpod, Washington, D. C.
This report deals with a preliminary effort toward ap-
plying a technique advocated for use in the Tropics to a
middle-latitude situation wherein the weather of the
Tropics seemed for a time to have been lifted bodily to
middle latitudes. Namias and Dunn [l] have shown how
the floods, which occurred in the wake of the passage of
the remnants of hurricane Diane in August 1955, occurred
in a homogeneous tropical air mass with its associated
high moisture content. The analysis of Chapman and
Sloan [2] indicates the lack of fronts in the area. The
temporary rejuvenation of the dying hurricane and the
sudden eastward turn of the storm as it reached the 40th
parallel of latitude must therefore be studied using tech-
niques applicable under such circumstances, namely the
techniques used successfully in tropical meteorology.
C. E. Palmer [3] has proposed that the analysis of
tropical situations be undertaken using the streamline
analysis technique of Bjerknes and collaborators [4]. Us-
ing this technique Palmer was able to locate, on day-to-
day low-level weather maps in the Tropics, lines of con-
vergencewhichwere associated with convective cloud
patterns. His lines of convergence, however,weredefi-
nitely not lines separating air masses of differing
densities, though, like fronts, they occurred in regions of
horizontal velocity convergence. However, he pointed
out that such velocity convergence must also occur in a
region where there is convergence in the streamlines along
an asymptote.
In the present instance a preliminary examination of
the surface wind field on the map for 0730 EST of August
18 suggests a marked velocity convergence in a narrow
zone extending northward from the dying storm center
I located in western Virginia, the zone of convergence curv-
ing thence eastward through Pennsylvania into central
Connecticut.One may note in particular that southerly
winds were being observed at Allentown, Pa., LaGuardia
Field, N. Y., New Haven, Conn., and Block Island, R. I.,
at the same time that easterly winds were being reported
a short distance farther north a t Scranton, Pa., Pough-
keepsie, N, Y., and a t Hartford (Bradley Field), Conn.
The chart for 0730 EST in figure 1 shows a heavy line
delineating this zone of convergence between the southerly
and easterly winds. The heavy line is extended westward
to show how it joins with a similar zone of convergence
extending northward from the storm center separating
easterly from northerly flow.
The observation that this zone was later to be a line
north of which the devastating flood rains occurred (and
in fact had already begun ’) and a line which roughly out-
lined the subsequent path of the low pressure center,
which had been identified previously as hurricane Diane,
prompted this more thorough analysis of the streamline
patterns shown in figure 1. The analyses werecon-
structed a t 6-hour intervals beginning with the chart for
0730 EST of August 18.
The method of analysis used in the preparation of the
charts shown in figure 1 was that which is customarily
used in the analysis of streamlines in the Tropics; namely,
the construction of isogons. Numerous short lines are
then sketched on each isogon parallel to the wind direc-
tion. These lines are then used as additional “winds” in
the construction of the streamlines. An example of this
technique is given by Riehl [5]. Riehl’s example includes
both a cyclonic indraft point and a hyperbolic point.
The necessity for a hyperbolic point to accompany a
cyclonic indraft point has been described recently by
Sherman and LaSeur [SI. Similar reasoning applies in
dealing with an anticyclonic “outdraft” wind system as
sketched in the Adirondack Mountain area of New York
State on the 0730 EST chart in figure 1. The accompany-
ing hyperbolic point is located in south central Massa-
chusetts.2
‘At the time of this chart some of the heaviest hourly amounts wereoccurring in
northeast of Scranton, and Norfolk, Conn., both within the two dominant rainfall
areas subsequently to be struck by the flood rains, such as Honesdale, Pa., situated
areas associated with the floods. At Honesdale 0.72 and 1.08 inches of rainfall fell dur-
ing the hours ending at 0700 and 0800 EST, respectively. During the hourending at
0700 EST, 0.80 inrh of rain fell at Norfolk, C o n . However, in addition to this, a rain-
fall pattern which developed independently of the two rainfall a r e a s of Pennsylvania
and Connecticut was beginning in the Boston, Mass., area where heavy rain was occur-
ring at 0730 EST. There 0.95 inch of rain fell during the hour which ended at 08134 EST.
Outside of these immediate areas, heavy rains were occurring in the Blue Ridge Moun-
tains of Virginia, but otherwise hourly rainfall amounts were of considerably lesser
magnitude.
3Although the location of a hyperbolic point that was associated with Diane cannot
be firmly established without extendiog the analysis to a larger area, snch a point may he
the one found on the chart for 1330 EST over eastern Lake Ontario as shown in figure 1.
It is not shown on the chart for 0730 EST but it is probable that it existed along a north-
westward extension of the asymptote of divergence located over Lake Ontario.
182 M O N T H L Y W E A Auons~ 1955
FIQURE 1.-Surface streamlines at 6-hour intervals, August 18-19, 1955.
AUQUBT 1955 MONTHLY WEATHER REVIEW 183
The 1330 EST chart in figure 1 shows that for a brief
period a cyclonic circulation developed in the Boston area,
it too being accompanied by the development of a new
hyperbolic point. This cyclonic indraft was associated
with a second peak of hourly precipitation amounts in the
Boston area which culminated in the measurement of
1.20 inches of rain there in the hour ending at 1500 EST
on the 18th. This precipitation cannot be attributed to
topographic lifting and must therefore be associated with
a general vertical motion pattern and upper-level diver-
gencewhich could have influenced the behavior of the
over-all storm circulation, including the turning of the
storm Diane toward the east as it reached the 40th
parallel.
A third peak of precipitation occurred in the Boston
area between 0700 and 1100 EST of the 19th as the storm
moved eastward to the south of New England.
The occurrence of the heavier amounts of precipitation
north of the line of convergence in northeastern Pennsyl-
vania and southwestern New England can be attributed
for the most part to orographic lifting of the lowest 5,000
feet of air, the air being saturated with dew points in the
70’s [7]. In this connection it should, however, be noted
that the orographic effect was augmenting a more general
vertical motion and precipitation pattern which developed
in advance of the storm center and that this more general
vertical motion pattern produced 2 to 3 inches of rain at
sea-level installations along the coast south and east of
the region of topographic lifting. In the lowest 5,000 feet
there was considerable horizontal velocity convergence in
the heavy rain area which is not shown on these streamline
charts in which wind speed is not a factor.
In figure 1 the primary lines of convergence are shaded
more heavily than ordinary streamlines. Note especially
the converging streamlines on the 1930 EST chart for
August 18 in the region extending from northeastern
Pennsylvania into southern New England and the general
tendency for the old hurricane to bring this convergence
line into its spiraling circulation in the manner suggested
by Wexler [8] as the mechanism which accounts for the
spiral-banded structure of hurricanes generally.
The late Dr. Isaac M. Cline [9] noted in his investiga-
tions of rainfall associated with tropical hurricanes that
the temperature in the different parts of the storms which
he studied did not vary greatly. He stated that the
causes of the heavy rains in tropical cyclones could not
be found in the surface temperature distribution (See
page 220 of [9]). Finally, he noted that the region of
greatest precipitation intensity was located “50 to 100
miles in front of the region where the winds of the right
rear quadrant converge with those of the right front
quadrant” and attributed the rainfall to the vertical
motion above the zone of convergence.
Briefly summarizing, the streamline patterns shown in
figure 1 reveal a marked zone of convergencewhich lay
close to the subsequent path of the storm center and
across the region of flood-producing rainfall.
Note added in proof-A paper by Y. Masuda,
M. Takeuchi, and M. Hashimoto (“On the Forecasting
of the Movement of Typhoon”, Papers in Meteorology,
Meteorological Research Institute, Tokyo, vol. 3, No. 4,
1953, pp. 246-251) has just come to the attention of the
author. In the 1953 paper it was suggested that typhoons
move toward a point in the streamlines that corresponds
to Scherhag’s “delta”. If on the maps for 0730 and
1330 EST, August 18 (fig. 1) this point is considered to lie
in southeastern New England, the subsequent movement
of hurricane Diane provides an interesting confirmation
of this hypothesis.-C.P.M.
REFERENCES
1. J. Namias and C. R. Dunn, “The Weather and Circu-
lation of August 1955”Including the Climatological
Background for Hurricanes Connie and Diane,’’
Monthly Weather Review, vol. 83, No. 8, August
1955, pp. 163-170.
2. W. T. Chapman and Y. T. Sloan, “The Paths of
Hurricanes Connie and Diane,” Monthly Weather
Review, vol. 83, No. 8, August 1955, pp. 171-180.
3. C. E. Palmer, “Tropical Meteorology,” Compendium of
Meteorology, American Meteorological Society,
Boston, 1951, pp. 859-880.
4. V. Bjerknes and collaborators, Dynamic Meteorology
and Hydrography, Part II, Kinematics, Carnegie
Institution of Washington, Publication No. 88, Part
11, 1911.
5. H. Riehl, Tropical Meteorology, McGraw-Hill Book
Company, Inc., New York, Toronto, and London,
1954, 392 pp.
6. L. Sherman and N. E. LaSeur, “Reconnaissance of the
‘Complete Hurricane’,” Bulletin of the American
Meteorological Society, vol. 36, No. 4, April 1955, pp.
7. U. S. Weather Bureau, “Hurricane Rains and Floods of
August 1955, Carolinas to New England, Part 1-
Meteorological Analysis,” Technical Paper No. 26.
(In preparation.)
8. H. Wexler, “Structure of Hurricanes as Determined
by Radar,” Annals of the New York Academy of
Sciences, vol. XLVIII, Art. 8, September 1947, pp.
9. I, M. Cline, Tropical Cyclones, The Macmillan Com-
152-158.
821-824.
pany, New York, 1926, 301 pp.