OCTOBER, 1927 MONTHLY WEATHER REVIEW 453
SOME INUNDATIONS ATTENDING TROPICAL CYCLONES
By I. R. TANNEHILL hurricanes that Droduce the swells and tides is limited
IWeathar Bureau, Oalveaton. Tex., November 1,18271
The rise of water on the coast in advance of a tropical
cyclone a t sea is frequently termed a “tidal wave.” It is sometimes, with greater propriety, referred to as a
“storm wave.” The use of the word “wave” in either
case is opposed by many meteorologists who contend
that the rise of water, in any event, is relatively slow until
the center of the cyclone approaches closely. It is a subject discussed only superficially in meteorological
textbooks. Cline (1) assembled tidal data and explained
many of the henomena of tides and waves in tropical
in advance of the cyclone is not a wave in any sense of that term.
The tide, due to astronomical causes, rises only 2 or
3 feet in the open oceans. There it is of little significance;
its rise and fall are gradual. When it strikes the coast its
range is frequently 10 to 12 feet. In certain bays and channels, where the wave encounters the shores and a
risng ocean bed, retardation causes a tide of 25 to 50
feet above low water. In the estuaries of many rivers vast sand flats are nearly
dry a t low water, and the tide rises wit,h such rapidity
that the wave assumes the shape of a wall of water called a “bore.” (2) Tides in the St. Lawrence, ranging from
3 to 4 feet in the gulf, penetrate 450 miles up the river,
with a range of 9 to 15 feet, and the rate of propagation
up to Quebec, 350 miles from the mouth, is 83 miles per hour. At the mouths of the Ganges and Amazon, the
bore is a t times very formidable and may, a t spring
equinoctial tide, with moon at perigee and favorable
winds, reach a height of 30 to 50 feet, advancing as a wall of water or series of waves (6).
Thus the astronomic tide, a gradual rise of 2 to 3 feet
in the open oceans, is transformed by contact with the
coasts and in some situations comes as a wall of water.
Its range in certain localities is reatly magnified. The
varied as the contours of the coast lines and the slopes
of the ocean beds.
The rise of water with the cyclone is likewise relatively small in the open ocean. Cline (1) estimates it, as less than 5 feet in the greatest st,onns. Certainly that due to
reduction of atmospheric pressure is not more than 2 or
3 feet. There are near the center of the cyclone tremen-
dous seas, described by many mariners as rising 30 to 40
feet, not like waves, but huge pyramids of water. The
level of the ocean it,self, however, is raised relatively
little. As the storm tide encounters the shallows and
indentations of the coast, it is retarded and the water
rises in many instances 10 to 15 feet and probably in
exceptional cases considerably more.
It is the object of this paper to discuss a number of
inundations attending cyclones and to show that the rise
of water exhibits many if not all of the characteristics of
the astronomic tide in like situations, including bores and similar phenomena, and that the direction of movement
of the cyclone with reference to the coast h e , the contour
of the coast, and rate of the cyclone’s approach and
passage are in certain instances responsible for great
storm waves that have destroyed hundred of millions of
dollars worth of property and hundreds of thousands of lives in the aggregate.
The need of such a study is remarked by Cline, who
says: “Concrete information regarding the winds in
cyclones, conc P uding that the rise of water along the coast
characteristics of this rise and B all of the water are as
76817-27-2
and, therefore, these are subjects that must be investi-
gated in this connection.” He does not consider the currents set up by the storm circulation except to note
their existence in citing the drift of gas and whistling
buoys. We know that the most powerful agent for trans- ferring quantities of water from one place to another is
the current, and strong evidence will be presented to
show that the currents about the cyclone are sufficient to explain the storm wave.
Aside from wave motion imparted to the ocean surface,
the cyclone has three important effects:
(a) The winds of the right rear quadrant of the cyclone,
in the Northern Hemisphere, with a counterclockwise rotation, are combined with the movement of translation
and are the most powerful and of the greatest duration
of any of the winds of the cyclone. The contrast of the strength of these winds with those in the other quadrants
is probably even greater than indicated by Cline, be-
cause his observations are not instantaneous. The data
he has assembled are from land stations. As the cyclone
moves inland, it loses intensity. Before t,he winds of the
right rear quadrant reach the observing station the cyclone has weakened and these winds are then less power-
ful than they were at the time the front of the cyclone
was passing over the station. (b ) The waters are raised by the winds and elevated
by the reduction of atmospheric pressure near the center
of the cyclone and form a whirling disk or mound which is redeveloped continually as the cyclone moves forward.
(c) As a vast whirlwind of great power, the cyclone communicates to the waters a turning movement which is quite pronounced near the storm center. As proof of
the power of these currents, Cline cites the fact that two
gas and whistling buoys in August, 1915, and three in
September, 1919, were carried 2 to 8 miles parallel to the coast. He states that Trinity Shoals gas and whistling
buoy, weighing 21,000 pounds, anchored in 42 feet of
water with 6,500-pound sinker, and 252 feet of anchor
chain weighing 3,520 pounds, was carried 8 to 10 miles
westward. These currents scour away the sand from the coast in great quantities. In 1900, a t Galveston, the waterfront
moved in se.vera1 blocks permanently due to the scouring action of the cyclone currents. Where street cam once
ran along the beach, the site of the tracks is now well out in the Gulf. These currents moved from northeast to
southwest or from right to left across the front of the
cyclone. After the cyclone moves inland a reverse cur-
rent is set up from southwest to northeast along Galveston
Island. There have been several instances of persons clinging
to floating wreckage and being carried, after the center of
the storm had passed inland, many miles from southwest
to northeast along the coast. Eliot (3) discusses these currents in connection with
cyclones in the Bay of Bengal and notes a strong westerly
set a t the head of the bay characteristic of storms in this
position and indicative of storm formation or approach toward the north end of the bay. The excessive drift
noted in many instances by Eliot, Piddington (4), and
others from the logs of vessels in cylcones is su$cient
proof in itself of strong currents about the center of the
cyclone. The fact that the speed of translation of a cyclone is
relatively small and its winds violent are evidence that a
strong circulatory motion of the waters must be set up.
454 MONTHLY WEATHER REVIEW OCTOBER, 1927
The very clear evidence that such a current does actually exist makes it all the more remarkable t,hat no one has
considered t,he effect, of an obstruction t.o t,his flow of
water.
When any headland or bay shore is so situated as to
impede the progress of this current from right to left in
advance of the cyclone, the waters pile up against the
obstruction and accumulate in bays and i n k s facing
into the current either directly or at an angle.
As the center of the cyclone approaches the coast, the whirling motion of t.he wn ters become more vigorous
with more violent winds. The resistance of shallows near the coast causes the waters in this rotating disk to
pile up on the right side of the center of the storm just
before it Inoves inland. All the waters to the, left of
the center are now carried around the rear of the miter and piled up to the right. The tide falls with great,
rapidity to t,he left of the center because the return hrnncli
of the current in front of the c.entcr which has supplied water to the left is now hindered by the shore and shoals.
If there is a bay or inlet to the right of the track of the
cyclone center, the, waters of this revolving mound are
precipitated toward the mouth of the bay or inlet. At that, point waters are already Rcc.umulated from hours
of resistance to the right-t'o-left current prior t,o the arrival of the center.
The center of the storni now inoves inland? leaving
the accumulated waters to be driven by t.he powerful winds to the rear of the center against the accumulated
water a t the entrance of the bay, inlet, or river mouth and a storm wave results, driven forward by the most
powerful winds of the hurricane.
With a gently sloping bed, this wave is ret.arded and
piled up by resistance. In some favorable cases it takes
the form of a wall of wate.r that sweeps everything before.
it. It is important to note that the displac.ement of this central mass of water to the right takes place before the storm cente,r move,s inland and a t the time shoal water
is reached. Therefore this pitching of the central mass to the right and its movement wit'h the mar winds can
take place in time for the rising wat,ers to be cnught in the shift of wind to southeast m c l south and thus go
forward t.0 the right of the cei1te.r.
The evidences that this action tnkes p1ac.e are. prac-
tically umquest,ioned, once we have realized t,he power of
t.his current. In September, 1900, the c.yc.lone moved t,o the west of
Galvestson Bay, the center passing to the left of t,hc city
shortly after S p. m. Quoting from the report of Dr. 1. M. Cline:
The water rose at it st,eady rate from 3 p. 111. until about 7.38
1). m., when there was B sudden rise of atmiit. 4 feet in ax many
wconds. I was staiiding at, niy front door, which was part,l>- opeii, watching the wat,er which mas flowing with great. rapitlit,y from east to west. The water at this tixiie was aliout tc ir~ches deep in my residence. arid the sudden rise of 4 feet. hroiight it. shove my
waist before I could change my position.
He notes the rapid current from eastlwest or from right to left in front of the cyc.lone. The c,enter was then
sweeping over a rising Gulf bed to the southwestward
and the shore line was rapidly cutting off the return current in front. His residence was located a t Twenty- Streets, several blocks from what was then
This wave penetrated
It was undoubtedly the front Approximately 6,000 persons lost
fifth the shore and P ine, and many buildings stood between his
residence and the open water. this section of the city.
of the storm wave.
their lives as this wave advanced.
To produce a storm wave of this kind, the cyclone must
move in a direction nearly normal to the coast line. Thus its rotary currents are developed to maximum strength when the storm strikes the shallows near the coast. The wave is then best developed when a head-
land, island, or other obstruction arrests the rotary
movement and there is a bay or inlet to the right for development of the wave.
The highest tides on the coast of the Gulf of Mexico have been developed under such circumstances.
At Corpus Christi, in 1919, the storm center passed to the left of the bay, and the waters reached a height of
16 feet on the left bay shore. The storm moved nearly
normally to the coast line. At Indianola, in 1886, the cyclone moved to the left
of Matagorda Bay, in a direction nearly normal to the coast line, and the city was swept away. It was located
on the left side of the bay, considered from the point of approach of the storm. The city has never been rebuilt.
At Galveston in 1900 and 1915 the high water was produced by storins moving to the left of Galveston Bay.
Galveston is located to the left of the bay entrance and on the northeast end of the island.
In 1915 a severe storm moved inland over southern
Louisiana, in a direction nearly normal to the coast line, and the highest tides of record were measured in the left end of the sounds to the right of the Mississippi
Delta, and to the right of the storm center. In July, 1916, the cyclone moved inland to the left of
Mobile Bay, nearly normal to the coast line; and Mobile, a t the upper end of the bay and to the left, had the
highest tide of record, 11.6 feet.
At Tampa, in October, 1921, a cyclone recurved in the eastern Gulf and passed to the left, nearly normal to
the coast line. The waters in Tampa Bay, which had been a t normal height, rose to 10.5 feet above low water, by far the highest of record.
The tracks of these storms are shown in Figure 1.
The Corpus Christi storm of 1919, the Galveston
storm of 1900, and the Louisiana storm of 1915 crossed very closely to the same point, approximately 27 north
and 89% west. One was moving westward, another
west-northwestward, and the third northwestward. The
tide a t Burrwood was between 2 and 3 feet approximately
ill each instance, evidently somewhat higher in the case of the storm that moved northwestward. However, the
water was then banking up on the right of headlands
which interfered with the flow from right to left about the centor. A t S a. 111. of September 29, 1915, there was
it tide of 3.7 feet at Burrwood and at Mobile 2.5 feet.
This difference was consistent as the stonn approached
nnd the water was not so high to the left of Burrwood, showing that the right-to-left flow was being hindered
and the water piled up as i t passed around the Mississippi
Delta. Practically all of these tide records show the water
piling up to the right of obstructions to its flow. These are separate and distinct effects from that caused by the drive of water from the right rear quadrant of the
cyclone described by Cline. In the Bay of Bengal the conditions along shore are
more favorable for tidal waves, especially a t the head of
the bay. The cyclones are perhaps as a rule more
intense than West India hurricanes. The astronomical tide is developed to a much greater extent than in the
Gulf of Mexico.
These conditions combine to produce more severe storm waves in India, especially when the time of high
water approximately coincides with the arrival of the
OCTOBER, 1927 MONTHLY WEATHER REVIEW 455
storm wave. On the coasts of India the storm wave frequently arrives as a sudden rise of water, sometimes as
an advancing wall and sometimes as a bore. The only authentic case of record of a bore produced from a West India hurricane was in September, 1926, a t Miami.
There the highest water occurred with the shift of wind,
and in the Miami River the tide came in the form of a
bore that left a mass of wreckage from boats that had
sought safe anchorage.
Eliot cites several great cyclones in the Bay of Bengal, some of which were attended by pronounced storm
waves. The movement of the storm, the shape of the
coast line, and other conditions were practically the same as that shown in the case of record water heights
in the Gulf of Mexico.
The Calcutta cyclone of 1864.-It crossed the coast, line moving in a direction nearly a t right angles to the coast,
near Contai, to the left of the mouth of the Hooghly. The barometer fell to 28.025 inches and the calm lasted
a t Contai from 9:45 to 11:OO a. m. The storm wave
arrived a t the mouth of the Hooghly a little after 10
a. m., high water being due a t about noon as the moon
was nearly full.
There was an enormous accumulation of water a t the northwest angle of the bay, the left side. In the case of
a storm there is for some time, as the storm center ap-
proaches so as to give a stonn wave, a large accuniula- tion of water, according to Eliot, and this head of water finally gives rise to a sudden and overpowering advance
of the accumulated mass of water up the river and an almost equally rapid inundation of the lowlands near the
seashore. The storm wave is estimated to have risen 40
feet in the Calcutta cyclone. The loss of life from drown-
ing was estimated a t 50,000 and from disease as a result
of the storm, 30,000. Eliot recites the account of an
eyewitness, to the effect that the natives on the coast,
whence he was traveling toward Midnapore, informed
him that a high bore was to be expected a t half past 12.
“The water,” his informant relates, “all a t once suddenly
rose as if by magic and steadily rolled towards us.”
The Backergunge cyclone of lS76.-The center moved
across the bay toward the northeast, passing to the left
of Chittagong, and near Bakergunge on the left shore. An enormous storm wave was dnven over the islands and lowlands near the mouth of the Megna. There was an
unusually high tide, followed very shortly by the storm
wave. The pressure of the advancing wave prevented the tidal and river water from flowing off. The storm wave
was retarded and finally overpowered the downflowing.
waters and rushed with irresistible force over the islands
and low-lying coastal areas covering them to a depth from
10 to 30 or 40 feet. It was estimated that 100,000
Lives were lost from drowning and sub!equently 100,000
more died of disease as a result of the inundation. The
exact time of the passage of the center of the cyclone
and the beginning of the storm wave are not given. Both occurred shortly after midnight with a spring tide.
The False Point cyclone of 1885.-The center passed
over False Point Lighthouse, the barometer falling to
27.135 inches. The reading was taken by a trained
observer a t a land observatory with a properly verified barometer. The wind a t the lighthouse at 6:30 a. m.
hauled from northeast to northwest, continued to blow a hurricane for a few minutes, then suddenly lulled. The
calm lasted till 6:50 a. m., when the wind came with re-
doubled fury from the south-southwest. The storm
wave came up a t 6: 20 a. m. and swept over False Point harbor, destroying all the houses ashore. It rolled in a
wide unbroken wave in a northeasterly direction, sub-
merging villages and carrying away before it, with irre-
sistible force, houses, cattle, human beings, e k . The measured height of the wave a t False Point was 32 feet.
Piddington (4) gives descriptions of inundations that
have visited Coringa on the Coromandel coast of India.
Coringa is located to the right of the delta of the Goda-
very River. According to these accounts, in December,
1789, during a cyclone, when the high tide was at its
highest point and the northwest wind, blowing with
fury, accumulated the waters at the head of the bay, t>he unfortunate inhabitants saw with terror three mon-
strous waves coming in from the sea. The first, sweep- ing everything in its passage, brought several feet of
water into th,e town. The second inundated all the low
country and the third overwhelmed everything.
In 1839 more than 20,000 persons are said to have
perished a t Coringa in a storm wave.
h r t i a l tracks of 7 tro ical cyclones that created record storm tides on the coast of the Gulf of Mesico. Alfrnoved so BS to approach themast appmrimatelg at right anKles. Record tides were produced in b3ys and inlets immediately to the right of the Dolnt where the center crossed the coat
A number of these inundations have occurred on the
low coastal lands of China. One in 1881, at Haifong, in
a typhoon, caused the death of 300,000 persons. As an illustration of the power of the cyclone in driv-
ing water around the center: In Ockober, 1910, a cyclone
described a loop in the eastern Gulf and finally passed
out over southern Florida on the lsth. On that day
the tide fell to 9 feet below mean low in the Hillsboro River at Tampa, while on the right of the center, south
of Cape Romano, the keys and islands were swept by
great waves from the Gulf that reached a great dist,ance inland.
The above accounts seem to bear out the stat.ement
that the rise of water in the storm inundation is by no
means gradual. In many cases the direction of move-
ment of the cyclone, the shape of the c.oast line, the oc-
currence of normal high tide at time of inundation, from
astronomical causes, and other influences, c,ombinc to
produce a sudden rise of water which sweeps forward
like a great wave and causes immense destruction and
great loss of life.
The survivors escaped by climbing trees. (5)
CONCLUSIONS
If the rise of water to the r’ h t of the center of n cyclone,
of the winds of the right rear quadrant of the cyclone,
there remains no satisfactory expla.nation of the sudden rise of waters to the right of the center. These winds to
on moving inland, is due so P ely to the driving forward
456 MONTHLY WEATHER REVIEW OCTOBER, 1927
the right of the center do not change abruptly either in
speed or direction, and even the sudden shift of wind to the op osite or nearly opposite quarter as the center
in many cases. It is difficult to understand how this wind, even though in n sudden aiid violent onslaught, such as occurs in the cyclone, can in so brief a space
drive forward such a mass of water. The storni wave sometimes precedes the shift of wind a t the rear of the center, aiid with this explanation it mist be assumed
that the wave outruns the wind which produces it. Cline then assumes that the rise is gradual and tbat
his esplanation of the tide, if accepted as offering the c1ea31 o y
there is no “storni wave” or “tidal wave.”
causes of high water, do not include the cnuses of a
storm wave. Yet the testimony of observers and the fact that hundreds of thousands of persons have been
drowned in these overflows, seem to be indubitsable evidence thnt the rise is sudden and overwlielniing.
But Cline does not assume that his is a full and coni- plete explanation of the tidnl phenomena of the hurricane.
Reasons have been advanced for believing that the
waters take on tt rotary motion, similar to the winds in the cyclone acting upon the water. These currents will
be conirnunicated to great depth, setting enormous masses of water in motion, as evidenced by the niovement
of buoys anchored in water 40 feet or more in depth.
The power of this great rotating mass of water is fearful to contemplate when it is obstructed by the const line and its accumulations are driven by the cyclonic winds.
Near the center the accumulation of water on the
right front is relieved by a swift1 flowing current along
passes w iE not account for the suddenness of the rise
shore. This current is sudden r y inipeded and later
reversed as the center of the cyclone moves inland and
the rear winds come upon it. With great pressure suddenly thrown agajlist this relieving current as the
center leaves the rotating miiss, there is cause for a more
rapid accuniulatiori on the right of the center. All the waters of the rotary disk tend to pile up on the right of
the ceilter against tlie coast line. Far from the center this is a slow process, but near the center, the shorter
the diameter of the whirl nnd the greater the velocity
of the current the more sudden and violent will be the onslaught.
When n bay, hilet, or river mouth lies immediately
to the right of the point where the cyclone crosses the coast, this mass of water drives forward into the sloping
bed and nnrrowing channel, to be retarded and heaped
up. It finally spills over and sweeps foiward. These
places are frequently harbors for ships and tlie locations
of cities with a considerable population.
.4CK NO WLE DG M ENT
The author claims nothing original in the way of ob-
servation. He his consulted the writings and observa- tions of Weather Bureau officials recorded mostly in the MONTHLI WEATHER REVIEW aiid numerous additional sources, but clietly the works of Cline, Eliot, and Pid-
tlington.
LITERATURE CITEI)
I j Tropical Cyclones, by I. A I . Cline.
12) A Practical hlnriiinl of Tides and Waves bv W. H. Wheeler. (3) Crclonic Storms in the Bay of Bengal, by John Eliot.
(4) Sailor’s Hornbook, by R. Piddington.
(5) Monthly Weather Review, October, 1910. (6) Reclus, Elisee, The Ocean.
THE RELATION OF SPRING TEMPERATURES TO APPLE YIELDS
By W. A. MATTICE
[Weather Bureau, Washington, October 28.19271
Apples, while not of such universal need as corn, wheat, and other important food crops, are still of suffi- cient value to the human race to have rather large areas
of certain States devoted to their cultivation. By ex-
amining tt dot map prepared by the United States Department of Agriculture, it will bc found that the
heaviest centers of apple prodiiction lie in two States, New Tork and Washington. No other State has the
concentration that is found iii these, and between them they produce a large proportion of the Nation’s apple
The cultivation of apples requires soinewliat diff ?rent
conditions of soil, climttte, etc., tlisn most crops. While different varieties of apples require longer or shorter
growing seasons, in most caqes the local conditions or topography must be f avornble if ti high-producing or-
chard is to be m:tintainecl. One of the most import:tiit
risks that confront the apple grower is the liability of
damage froin late spring frosts. Most npple-producing areas of the United States are exposed to this injury and
serious losses occur, but the frost hazard in some sec- tions is comparatively small, particularly in the North- east. Fruit trees respond readily to relatively short
periods of warin weather in spring, and when there are
rather long periods of warmth, premature blooming is practically certain. In cases of this kind, the late frosts
cause greater damage than when the trees are in a less
advanced stage, even though the temperatures may be lower in the latter case.
It has long been well known that the location of an orchard is n vital factor in determining its success. A
supply.
north slope in some cases has been found to be slightly
more favorable than other exposures, due to the retarding effect on blooming and thus reducing the liability of dam-
age by frosts. Orchards in pockets are esposed to harm through air drainage, which will often cause extensive in- jury to bloom or newly set fruit, while CL neighborin
orchard on a higher elevation may not be herniecf Spring frost.;, the nrnount of precipitation, the summer
temperatures, C t C ., RTC eleinents over which the orchard-
ist has 110 control, hiit weather influences can often be controlled or modified through improved orchard manage-
ment. The iiiland river valieys of Washington are peculiarly
dapted for ttpplc culture, with their comparatively mild
climate and long summers. The southern shore of Lake Ontario in New Pork State is another region wkich has
been largely devoted to the cultivation of deciduous
fruits, with the great body of water actingas a deterrent for spring frosts ant1 otherwise moderating the climate.
In other States, Virginin and West Virginia are probably
the only ones showing such a concentration of fruit orchards, with the Shenandoah Valley famous for ih3
orchards, and especially for the apple-blossom festival which takes place there every year. Conditions in Vir-
ginia, again, are such as to promote apple growing on a large scale, with the great valley affording an extensive
area sheltered from many climatic severities. During the suninier of 1936 n survey of the apple-
producing sections of Virginia, West Virginia, and Penn- sylvania was macle by the several State experiment sta-
tions and the Department of Agriculture with the coop