On November 3 and 4, 1927, torrential rains fell
over much of New England, causing the most severe floods of which we have
knowledge over extensive areas in Vermont, New Hampshire, Massachusetts,
Connecticut, and Rhode Island. Vermont experienced an exceedingly
heavy rainfall, the area of greatest precipitation centering on the long
ridge of the Green Mountains and extending southward over western Massachusetts
into Connecticut. A smaller area of equal and possibly much greater rainfall
centered on the White Mountains of New Hampshire, with lighter rainfall
over the southern and northern portions of the State. A third area of much
smaller extent, but with recorded intensities of precipitation practically
as great as in Vermont, centered in northern Rhode Island and extended
northward to Worcester, Mass.
The destructiveness of such a storm depends upon
a number of circumstances, important among which are the character of the
soil, the topography, the condition of the ground and of the streams and
ponds, and the rate of precipitation. At the time of this storm the ground
had been thoroughly saturated by heavy rains which fell from the 18th to
the 21st of October. The natural lakes and swamps had been filled, and
most of the rivers had been raised to medium high stages, so that practically
all the surface storage available had been utilized less than two weeks
before the storm.
As a result, the rivers quickly overflowed their
banks, spread over meadows and farm lands in the first bottoms, and filled
many of the valleys from hill to hill. The grades of the streams are so
steep that excessively high velocities were attained, and the rushing waters
washed out bridges, retaining walls, dams, road embankments, buildings,
and farm lands. In many sections of the mountainous country near the headwaters
the flood peaks arrived suddenly and at night, the inhabitants were taken
unawares, and many were unable to reach safety before being drowned in
their homes. The report of the Advisory Committee of Engineers on Flood
Control, State of Vermont, shows that the total number of lives lost in
the State was 84, and of these 55 were in the Winooski River Basin.
Robert M. Ross, commissioner of forestry, State of
Vermont, and chairman of the Vermont Flood Survey, states that the damage
to cities, villages, railroads, and other public utilities was over $14,000,000;
to highways and bridges, $7,000,000; and the total damage more than $28,000,000.
The other States suffered severely but less than Vermont.
Only by careful analyses of data concerning flood
magnitude is it possible to study adequately the problem of flood control;
and flood control is a very necessary part of complete utilization of the
water resources of a region. Many feasible reservoir sites in the devastated
area have not yet been developed because the expense is apparently unwarranted
by the demand for utilization of the stored water for power. However, a
detailed study of these sites as reservoirs for flood prevention combined
with power developments may show that the construction of the projects
would be warranted, if the cost were equitably divided among those who
would be benefited.
Although damages to property resulting from failure
of engineering structures may be evaluated as a basis for determining the
limit of economic cost of such structures, the loss of life cannot be evaluated.
Structures whose safety involves human life should be designed much more
securely than those whose failure would involve damage to property alone.
As a general rule, storms that visit the northeastern
section of the United States approach from a westerly direction and are
of moderate intensity. Occasionally a storm is blown in from the Atlantic
Ocean by an east or northeast wind, and the rainfall in such a storm is
likely to be high. At more rare intervals tropical storms are forced inland,
entering the New England States from the south and proceeding northward.
These storms are usually attended by heavy precipitation. The great storm
of November 1927, and perhaps all storms producing abnormally high rainfall
in the past have been of this nature.
From a study of meteorologic data and storm centers,
it appears that several factors combined their influence to cause conditions
of rare occurrence, which produced a storm of unusual proportions. These
factors were indicated by weather maps issued by the United States Weather
Bureau immediately before and during the storm and were discussed in several
published reports, especially the paper by J. H. Weber and C. F. Brooks,
of Clark University, Worcester, Mass.
Of greatest effect was the steady approach of a tropical
storm from the south, which according to the weather map first appeared
almost directly over Cuba as early as October 29. This storm was not of
unusual severity and did not show much action until November 1, when it
started northward, reaching a point off the coast of South Carolina by
the night of November 2. By the morning of November 3 the storm center
had reached the lower end of Chesapeake Bay.
As predicted by the United States Weather Bureau
at Washington, on the evening of November 2, under normal conditions the
storm would have continued up the coast with moderately heavy rains and
with light showers in central New England, causing little if any disturbance.
However, out to the northeast an extensive area of exceedingly high pressure
prevented the storm from proceeding in that direction. At the same time
a high-pressure area had moved in from the northwest to a position north
of New York State. Thus the tropical storm was caught between the two cold
areas of high pressure and was forced to pass over them, causing torrential
rainfall.
The effectiveness of this barrier is indicated by
the great range in temperature on its sides. A difference of 19° occurred
in about 80 miles between warm Amherst, Mass., and cold Albany, N.Y., on
the west, and between warm Brattleboro, Vt., and cold Northfield, Vt.,
on the north. Converging winds were observed at midnight November 3, of
a velocity of 50 miles an hour from the southeast at Providence, R. I.,
and of 8 miles an hour from the north at Worcester, Mass. During the hour
after these observations rain fell at Worcester at the rate of nearly 2
inches an hour.
The path of least resistance, as indicated by the
pressures shown on Figure 7, was directly over western Massachusetts and
Vermont. The great stream of warm moisture-laden air was not only forced
over the Berkshire Hills and Green Mountains, whose altitudes range from
1,500 to 3,000 feet, but in addition it was thrust upward over the barrier
of cold, heavy air moving down from the north, which was just as effective
as the mountains in forcing the warm air upward.
The warm moist air was cooled so greatly as it reached
the higher altitudes that much of its moisture carrying capacity was lost,
and it therefore yielded tremendous quantities of rain. As the air became
cooled it moved on, giving place to more moist air from the unfailing supply,
and this in turn yielded its quota of moisture. As the meteorologic conditions
remained practically unchanged over a period of 24 hours, it is not difficult
to account for the excessive rainfall which was reported in the path of
the storm.
The magnitude of the northward movement of the moist
tropical air was indicated by the low-pressure area which extended from
the Appalachian Mountains on the west to a point a considerable distance
east of the Atlantic coast. This immense body of moisture-bearing air was
moving at a high velocity through the funnel-shaped exit over New England.
At Boston, Mass., the highest aerologic observation, nearly 2 miles above
the surface, indicated a wind blowing from the south at a velocity of 45
miles an hour, while at 6,500 feet the rate was 51 miles an hour. Both
at Worcester, Mass., and at Washington, D. C. the highest clouds visible
during the storm were moving rapidly from the south.
The storm of November, 1927, was general over the
northeastern section of the United States, but the greatest effect was
concentrated over western New England. Records from Vermont and New Hampshire
indicate that the precipitation increased with the altitude. Unfortunately,
no rainfall records were obtained from the high areas of the Green Mountains
in Vermont and the White Mountains in New Hampshire. Were such records
available, they would undoubtedly show a much greater rainfall than any
record obtained during the storm. Records were obtained from only two areas
receiving more than 9 inches of rain. One was a long, narrow area along
the summit of the Green Mountains in central Vermont, and the other was
a small area in southwestern Rhode Island. The total area in which there
was a precipitation of 9 inches or more was about 500 square miles.
Records of precipitation have been collected by the
United States Weather Bureau, by X. H. Goodnough, chief engineer of the
Department of Public Health of the Commonwealth of Massachusetts, and by
many private persons. The rainfall records were well distributed over the
territory affected by the storm, with the exception of the areas near the
tops of the mountain ranges.
The following records have been taken from a paper
by Goodnough. They include the records of the numerous rain gages of the
United States Weather Bureau in all parts of New England and the adjacent
sections of New York, the records of the Department of Public Health in
Massachusetts, and the results of observations by many power companies,
water departments and companies, and private observers.
|