USGS Fact Sheet 041-01
Download this fact sheet as a PDF file
Download a free copy of
The 1951 Floods in Kansas Revisited
"Measured in terms of human suffering, tremendous losses in property, and extensive
disruption of business activities throughout the flooded area, it was the greatest catastrophe
within the history of the region. Measured in terms of river stages and discharges, and of
extent of the areas inundated, it was the greatest flood in the Kansas River Basin of which
there is reliable record." (Veatch, 1952)
-Kyle E. Juracek, Charles A. Perry, and James E. Putnam
July 2001 marks the 50th anniversary of the largest floods to occur in Kansas during the 20th
century. The 1951 floods, exceeded only in recorded history by the legendary flood of 1844,
primarily affected the Kansas, Marais des Cygnes, Neosho, and Verdigris River Basins in
eastern Kansas and the Osage and Missouri River Basins in Missouri. According to the American
Red Cross, 19 people were killed, directly or indirectly, and about 1,100 people were injured
by the 1951 floods in Kansas and Missouri (U.S. Geological Survey, 1952). The most damaging
flooding in 1951, and the event that received the most media attention, occurred along the
Kansas River where the cities of Manhattan, Topeka, Lawrence, and Kansas City sustained
extensive damage (fig. 1).
Total damage from the floods was unprecedented. From the headwaters of the Kansas River to
the mouth of the Missouri River at St. Louis, about 2 million acres were flooded, 45,000 homes
were damaged or destroyed, and 17 major bridges, some of them weighted with locomotives in an
attempt to hold them, were washed away. By October of 1951, estimates of the total damage
ranged as high as $2.5 billion (Davis, 1953) (about $17 billion in 2000 dollars).
Within the affected areas, transportation was disrupted as highways and railroads were closed
from days to weeks. Damage to municipal water supplies and sewage-treatment works was also
extensive. In Kansas, 33 water-supply systems were shut down, requiring that water be brought
to the affected communities by tank trucks. At Topeka, the water works were kept in operation
thanks to the efforts of as many as 5,000 men at a time that maintained a floodwall during the
flood (U.S. Geological Survey, 1952). One of the more unusual damage reports came from Le Roy,
Kansas, where the Neosho River had washed caskets from graves at the Le Roy Cemetery (Christy,
The flood caused significant changes to the affected river and stream channels and the
adjacent flood plains. Along the Kansas River, the flooding resulted in substantial bank
erosion and channel widening. On the adjoining flood plain, which was submerged to depths of
15 to 20 feet in the vicinity of Lawrence and Topeka, the land surface was scoured to depths
of as much as 15 feet in some places and covered by deposits of sand and silt to thicknesses
of as much as 4 feet in other places (McCrae, 1954) (fig. 2). Similar changes were
noted in the other affected basins.
The July 1951 floods in Kansas were caused by a storm of unusual size and intensity for the
Great Plains. Above-normal precipitation during May and June 1951 caused some major flooding
and established conditions favorable for maximum runoff from subsequent precipitation. These
conditions included high streamflows, high ground-water levels, and a minimum capacity for the
soil to absorb any additional rainfall (U.S. Geological Survey, 1952).
Then came the great storm of July 9-13, 1951, that was centered near the common divide of the
Kansas and Neosho River Basins (fig. 3). Precipitation began during the afternoon of
July 9 and continued through the morning of July 10. Following a brief respite, the
precipitation began again the evening of July 10 and continued through July 12. Each day was
characterized by excessive rainfall during the late afternoon and night with little or no
rainfall during the early and mid-afternoon hours. By midnight July 13, almost unprecedented
total amounts of rain had fallen since the beginning of the storm. Four areas of particularly
excessive rainfall, centered about 27 miles southwest of Manhattan, 36 miles south-southwest
of Manhattan, 15 miles southwest of Emporia, and 30 miles west-southwest of Topeka, had total
storm amounts of more than 16 inches (fig. 3) (U.S. Geological Survey, 1952).
In 1951, the U.S. Geological Survey (USGS) operated a network of 96 streamflow-gaging stations
in Kansas (fig. 3). Of those 96 stations, 36 recorded the highest flows since the time records
began through the year 2000 (fig. 3). Most of the record-high flows recorded during
the 1951 floods occurred in July, although for a few stations the high flow was recorded in
May or June.
The magnitude of the 1951 flood can be put into perspective by comparing the highest flows
recorded in that year with the highest flows recorded for the entire period of station
operation. For example, information on the highest annual streamflow has been collected for
the Kansas River near Lecompton since 1891. As shown in figure 4, the 1951 flood at Lecompton (with an estimated
high flow of 483,000 cubic feet per second) was substantially larger than the high flows
recorded for the 1903 and 1993 floods. Table 1 lists selected currently
(2001) operated USGS streamflow-gaging stations with record flows recorded during 1951. None
of the stations shown in table 1 have had flows that exceeded those
recorded in 1951.
The answer is yes. The occurrence of the 1951 flood helped initiate the construction of
numerous flood-control reservoirs and levees that have helped to reduce the inundation by
subsequent floods in Kansas. Thus, although a flow of a magnitude comparable to the 1951
flood is certainly possible, the associated flooding would likely be less due to storage of
floodwaters in the reservoirs. Damage caused by flooding will vary by location depending on
the amount of development in the flood plain. Given the right combination of circumstances
and conditions, a flow of equal or greater magnitude is possible. For example, a major flood
could result if excessive rainfall occurred at a time when the basin was already saturated
and the reservoirs were already full, and (or) if much of the rainfall fell downstream from
the reservoirs. Major floods occur occasionally, and the risk of an extraordinary flood like
those of 1844 and 1951 will always be with us.
Even though the floods of 1951 were of epic proportion, there was at least one other flood in
eastern Kansas that was larger. On the Kansas River, the largest flood in recorded history
occurred in 1844; however, little damage resulted from this flood as it happened before
permanent settlement of the region (Flora, 1952). Other significant floods on the Kansas River
occurred in 1903 and 1993. These floods, like the 1951 flood, occurred after the flood plains
had been extensively developed and thus caused substantial damage.
The flood of 1844 is considered the "maximum" flood on the Kansas River. The 1785 flood on the
Mississippi River at St. Louis, Missouri, was approximately 1 foot higher than the 1844 flood
(Reed and others, 1993), but accounts are sketchy. Undocumented accounts hint that the 1785
flood also occurred on the Missouri and Kansas Rivers, but no reliable records exist on its
magnitude. Reliable data are available for the floods of 1844, 1903, 1951, and 1993, and they
can be compared according to relative depth of water and the amount of flow
(fig. 5, table 2).
Relative flood depths for 1844, 1903, 1951 and 1993 can be traced along the Kansas River from
where it is formed by the confluence of the Smoky Hill and Republican Rivers near Ogden,
Kansas, downstream to the Missouri River at Kansas City, and onto the Mississippi River at
St. Louis, Missouri. Figure 5 shows the relative depth of
water for the different floods at Ogden, Topeka, and Lecompton in Kansas, and Kansas City and
St. Louis in Missouri.
From Ogden to Lecompton, Kansas, the 1844 flood along the Kansas River was approximately 5
feet deeper than the 1951 flood. Once the 1951 flood reached Kansas City, water depths were
only about 2 feet less than in 1844. Along the Kansas River, the 1951 flood depths were
greater than in 1903 and 1993. Along the Missouri River, the 1993 flood depths were greater
than in 1844, 1903, and 1951. Upstream from Kansas City, flood depths in 1993 were affected
by the flood-control reservoir system which substantially reduced the high flows. However, at
locations where levees were built to protect urban areas, the 1993 flood depth increased
substantially within areas bounded by the levees. Flood depths at Kansas City and St. Louis
in 1993 were increased by the levees that protected much of the flood plain
Table 2 lists the highest flows for each of the floods.
Estimates for the flood of 1844 are available only for gaging stations on the Missouri and
Mississippi Rivers. Along the Kansas River, the highest flows were recorded during the 1951
flood, followed respectively by the 1903 and 1993 floods (table 2).
There are no flow estimates for the 1844 flood along the Kansas River. However, documented
flood depths would have produced flows greater than the 1951 flood. If the flood-control
reservoirs in the Kansas River Basin had not been in place during 1993, the resulting flood
flows would have been greater but still not as great as the floods of 1903 or 1951 (Perry,
1994). During the height of the flood on July 13, 1951, almost 90 percent of the flow in the
Missouri River at Kansas City came from the Kansas River, a tributary that represents only
about 12 percent of the Missouri River's drainage basin.
The hydrologic conditions prior to each of these floods were similar. A lengthy rainy period
prior to the maximum flooding created saturated conditions. Then, a major storm system with
excessive precipitation over a large area occurred that simultaneously drove many tributary
streams and rivers over their banks. Each of the major floods had storm precipitation totals
that were similar, only their duration and location were different. It has been suggested
that "...a small difference in the distribution of the heavy rains on July 10-12, 1951, and
their continuation for one day longer, would in all probability have produced a flood equal
to that of 1844" (Flora, 1952). Had the storm in 1993 occurred over a shorter period of time,
flooding probably would have been more extensive.
The rains will come again. When, where, and how much will determine whether the next flood
will rival the big ones.
When flooding occurs, the USGS mobilizes personnel to collect streamflow data in affected
areas. The USGS was out in force during and after the great floods of 1951, collecting
streamflow data and documenting high flows that occurred.
Currently (2001), the USGS operates more than 150 streamflow-gaging stations on streams and
lakes in Kansas. Although the station equipment has been modernized since 1951, the type of
data collected at the gaging stations is the same now as then. Streamflow information
collected by the USGS during floods is used for reservoir operations, flood warning and
forecasting, design of bridges and flood-control structures, and flood-plain regulation and
The USGS measures streamflow in terms of cubic feet per second. One cubic foot per second
is equal to about 448 gallons per minute, 27,000 gallons per hour, or 646,000 gallons per
day--close to the amount needed to fill an Olympic-sized swimming pool in 1 day.
The process of streamflow measurement at USGS gaging stations has not changed significantly
since 1951. Where possible, direct measurements of flow during the 1951 floods were made from
bridges and boats (fig.
6). However, at most gaging stations in eastern Kansas, the 1951 floods reached such great
depths and high velocities that USGS personnel were unable to reach the gaging stations
located on some bridges and, therefore, were unable to make direct flow measurements. In some
locations, the gaging station was left isolated from the river and, thus, could not be used to
record river stage. For example, in the flood analysis for the Kansas River at Ogden, Kansas,
R.W. Carter of the USGS wrote that "...the river cut a new channel to the right of the bridge
during the flood leaving the gage in the old channel." W.P. Somers of the USGS wrote in the
flood analysis for the South Fork Solomon River at Alton, Kansas, that "...the gage was lost
during the flood of July 12, before the bridge was destroyed." In such cases, "indirect
methods" were used to estimate high flows after floodwaters receded. A detailed description
of streamflow conditions during the 1951 flood is provided in a USGS report titled
"Kansas-Missouri floods of July 1951" (U.S. Geological Survey, 1952).
The USGS is prepared to document future floods. The streamflow-gaging network in Kansas has
grown from 96 stations in 1951 to more than 150 stations in 2001. Streamflow data from the
expanded network improve National Weather Service flood forecasting, provide additional data
for USGS flood-frequency analysis which is used by the Kansas Department of Transportation for
bridge design, and provide information for use by State Emergency Management Agencies and
Mitigation Teams before, during, and after flooding.
Most gaging stations in Kansas are now equipped with telemetry equipment that relays stream
gage-height data from the gaging stations to USGS offices via satellite. Water-resource
managers and the public also have access to this data to make decisions necessary during large
floods. Real-time data for Kansas streams are available on the USGS web site at
The USGS normally determines streamflow by direct measurement. The USGS measures stage
(the height of the water surface, also known as gage height) and streamflow at all gaging
stations on a routine schedule. Typically, measurements of water depth and velocity are made
at approximately 30 locations across the stream. The distance between measurement locations
(width), the speed of the water (velocity), and water depth are multiplied to compute
streamflow (discharge) in cubic feet per second. Many streamflow measurements made over the
range in stage of the stream are plotted against the corresponding stages to define the
stage-discharge relation that is used in conjunction with the continuously recorded stage to
determine streamflow throughout the year.
However, in 1951 the USGS had to rely on indirect measurements to determine high flows
after the floodwater receded. Indirect measurement involves the use of field-surveyed
high-water marks, information on channel characteristics, and a hydraulic flow model to
- Christy, Wanda, 1987, Coffey County-volume 1, A glimpse into its past, present,
and future!: Burlington, Kansas, Coffey County Today, 92 p.
- Davis, K.S., 1953, River on the rampage: Garden City, New York, Doubleday and Co.,
- Flora, S.D., 1952, The great flood of 1844 along the Kansas and Marais des Cygnes
Rivers: The Kansas Historical Quarterly, v. XX, no. 2, May 1952, p. 73-81.
- McCrae, R.O., 1954, Geomorphic effects of the 1951 Kansas River flood: Lawrence,
Kansas, University of Kansas, master's thesis, 68 p.
- Perry, C.A., 1994, Effects of reservoirs on flood discharges in the Kansas and
Missouri River Basins, 1993: U.S. Geological Survey Circular 1120-E, 20 p.
- Reed, H.L., Perkins, T.J., and Gray, G.L., Jr., 1993, Water resources data,
Missouri, water year 1992: U.S. Geological Survey Water-Data Report MO-92-1, p. 142.
- U.S. Geological Survey, 1952, Kansas-Missouri floods of July 1951: U.S. Geological
Survey Water-Supply Paper 1139, 239 p.
- Veatch, N.T., 1952, The Kansas flood of 1951: Journal of the American Water Works
Association, v. 44, no. 9, September 1952, p. 765-774.
For more information on the 1951 flood and to view a collection of 1951 flood photographs
and flood hydrographs for selected USGS streamflow-gaging stations visit the USGS Web site at:
Other Flood Links of Interest
- Kansas Flood Watch
- Current Streamflow Data
- FS 024-00 Significant Floods in
the United States During the 20th Century--USGS Measures a Century of Floods
- WRIR 00-4079 Estimation of Peak
Flow Streamflows for Unregulated Rural Streams in Kansas - HTML 64-9 KB;
PDF - 6.86 MB
- WSP 2502 Summary of Significant
Floods in the United States, Puerto Rico, and the Virgin Islands, 1970 Through
U.S. Geological Survey
4821 Quail Crest Place