Popular Science Monthly/Volume 69/September 1906/The Protection of the Alluvial Basin of the Mississippi




THE hydrographic or drainage basin of the Mississippi River (Fig. 1) is equivalent in area to one third of the United States. Thirty-two states and territories contribute water to the volume of the river; eight of these divisions send water to no other system. The discharge of rivers is not in any proportionate way related to the size of their drainage basins. The potent factors which determine the volume of discharge are the precipitation of rain over the basin and the character of the soil. The upper Ganges has a basin less than one seventh that of the Mississippi. It equals the latter river, however, in the volume of its discharge. The Hoang Ho, with a basin area fifty per cent, only of that of the Mississippi, discharges more than twice as much water into the sea. If the discharge of the Mississippi proportionally to the size of its basin equaled that of the Po, the volume of the discharge of the former would be multiplied by six. The Danube more nearly equals the Mississippi in the ratio of discharge to size of

Fig. 1. Hydrographic Basin of the Mississippi, with Rainfall Types. Missouri type has the maximum rainfall in April, May and June; Tennessee, in February and March; Lake, a late spring maximum in June, and an early fall maximum in September.

Fig. 2 The Lower Mississippi, with Tributaries, alluvial basin, and higher lands (dotted). CG, Cape Girardeau; C, Cairo: C", Columbus; M, Memphis; H, Helena; A, Arkansas City: G, Greenville; V, Vicksburg; N, Natchez; BR, Baton Rouge; NO, New Orleans. basin. It, too, has a portion of its drainage area protected from the prevailing winds and approximates to the rainfall condition of the Mississippi. The Ganges and the Hoang Ho are in regions of copious seasonal rainfalls.

The portion of the Mississippi Basin subject to inundation includes land on both sides of the river south of Cape Girardeau, Mo., and aggregates 29,700 square miles, or an area equivalent to the state of South Carolina. Through this alluvial basin (Fig. 2) the river winds in a rather tortuous path, the distance by river (1,700 miles) being nearly three times as long as a straight line drawn from Cape Girardeau to the Gulf of Mexico (GOO miles).

It is characteristic of many flood plains that the land immediately adjacent to the river is higher than the more distant parts of the plain (Fig. 3). These higher parts are called natural levees by some writers. If, therefore, the river rises beyond the limit of its banks, there is likely to be a general inundation of the alluvial basin. This feature of flood-plain form is further illustrated by the tributaries of the lower river. From an ordinary valley slope a river flows fairly direct into the trunk stream. If a tributary crosses a flood plain of some size, the stream must force an entrance against the rise of the back slope of this plain and finally must breach the banks of the river.

Fig. 3. Cross Section of a River Valley.

Fig. 4. Hydrograph of the Mississippi River from June, 1902, to May, 1905.

It is apparently easier for the tributaries of the lower Mississippi not to overcome this rise, but to flow down stream in the back swamp lands or bottoms until some more formidable barrier forces them to empty their waters into the main channel. In this manner the St. Francis (Fig. 2) flows for 100 miles and is forced into the Mississippi just above Helena, where the main river, after crossing the alluvial basin, touches the higher land on the west. The Yazoo River is forced in at Vicksburg, after flowing along the flood plain for 200 miles. Other tributaries show this same characteristic. The Atchafalaya pursues its course to the Gulf as an independent stream. The height of the river bank over the back swamp districts varies from 10 to 25 feet.

The large area of the drainage basin of the Mississippi would yield an unmanageable amount of water to the lower river during the stage of flood if the excess of discharge at that time resulted from a uniform rainfall condition. The very size of the basin with the tributaries of the main stream reaching far into rainfall areas of different types and seasons is beneficial to the control of floods. Reference to the map (Fig. 1) and the appended explanation may aid one in understanding the condition of rainfall over the basin. In addition, it is well to bear in mind that the condition of the ground affecting the amount of run-off of water is an important factor in the amount of discharge. The Ohio basin has its heaviest rains in January, February and March. Its largest tributaries, the Cumberland and the Tennessee, rise in regions of copious winter rainfall and add enormous volumes of water to the Ohio. The basin of the Ohio is less than one half that of the Missouri, yet it furnishes over twice as much water to the Mississippi. The melting of the snow and the frozen condition of the ground which increases the percentage of run-off at the time of the early spring rains swell the volume of the Ohio, while the late spring rains over the Missouri basin fall on an absorptive soil. Only 15 per cent, of the rainfall is drained from the latter basin and 24 per cent, from the former. The percentage over the Ohio during the flood months because of the conditions stated above is probably much higher.

The sequence of floods from the tributaries is first, the Ohio, then the upper Mississippi, followed by the Missouri and the western streams. Great floods are not annual occurrences. Recently the years marked by excessive floods have been the years of 1893, 1897 and 1903. The floods of the years 1898 and 1901 did not fall much short of the records of the seasons previous.

The hydrographs of floods (Fig. 4) and the profiles of rivers at different seasons (Fig. 5) show that the floods proceed down stream somewhat as a wave. Fig. 5. Profiles of the Mississippi River at Flood Stage (BC), at low water stage (DE), and at an intermediate stage. The highest point or crest marks the extreme danger limit due to height of flood. The hydrograph also indicates that the crest flattens somewhat in its down-river progress. The reason for this may be seen in the river profiles which are drawn from three stages of the river's annual fluctuations. The down-stream slope (AC) of the flood wave has increased over the normal slope of the river (GE). This results in an increase of speed of the waters on this slope. On the other hand, the up-stream slope (BA) is a less one than the normal slope (DG), and a decrease of speed over normal flow results. At certain times the slope (BA) may be against gravity, and a further retardation will be experienced. This would occur when a tributary added a large volume of water to the Mississippi, as in the case of the Ohio during the spring freshets. The tendency of this increase of speed on the down-river side of a flood wave and the decrease of velocity on the up-river side is to reduce the size of the flood wave by draining off the excess of water faster than it can accumulate. If the river is long enough to allow this process to so act, a flood would finally reduce itself to an insignificant rise of water.

The following table may yield additional information concerning flood conditions[2]:

Levees were constructed by the early settlers of Louisiana about the year 1717 near New Orleans. In 1844 the right bank of the river in this state was embanked and many other isolated levees were in existence, especially along the Yazoo basin front. The average height of the Louisiana levees was four feet. Through the years up to 1883, there seems to have been a constant agitation of the question of levee building. Riparian states, districts and owners formed committees and boards, taxed the protected lands or the products of these lands and in general managed to put their constituents under a heavy burden of debt. In 1858 a tax of 10 cents per acre was demanded on all lands which were freed from inundation, and such lands as bordered the river were often subjected to a tax of 25 cents per acre. In 1865 a board was constituted with a revenue derived from a tax on cotton of 1 cent per pound. In the year 1882, a flood greater than any previously experienced overflowed the entire basin and destroyed most of the levees then existing. The Mississippi River Commission, created a few years before, now entered upon its work at an opportune time. With the landowners disheartened, their labors resulting in little gain, their money invested in levees swept away, the allotments of the commission were enough to revive the courage of the riparian proprietors. There has been a steady gain in the protection of the alluvial basin since the creation of this commission. At present the levee system comprises about 1,500 miles of structure and is 71 per cent, completed.

The height of the levees is a varying one. If the levee is built on the immediate banks of the stream (Fig. 3), which is the highest part of the flood plain, a levee less high might suffice as well as one built on the back-slope. Where the original four-foot levee stood in Louisiana is a levee about 16 feet high. The top is not much higher than the top of the original levee, but, being situated at some distance from the banks, its base is 10 feet lower. The immediate banks of the river are so subject to caving that they do not make, in all cases, a safe foundation for a levee. It has been considered safer and wiser in many instances to build a more stable foundation. Furthermore, if the levees are built upon the higher and immediate parts of the flood plain, the levees would be nearer each other. The nearer they approach one another, the higher they must be. Any detraction from the horizontal expansion of the waters must be evidenced in a vertical expansion. With the levees placed further apart, a less height is possible. The thing that determines the height to which they shall be built is the flood. There is an endeavor to place the grade of the levees at from 2 to 4 feet above the gauge measurements of the highest floods. Thus for a while the provisional grade was 2 to 3 feet above the 1897 flood. Later the 1903 flood set a new mark and a grade 2 to 2.5 feet above the 1903 high-water line is suggested. At Lake Providence, the 1903 water mark was 2 feet above the 1897; at Greenville, it was 2.4 feet above. Accordingly, the high water of 1903 reached approximately to the tops of the levees suggested after the 1897 flood, and a new height, 2 to 3 feet above the former, is now demanded. This latter height is about 5 feet above the provisional grade of five years ago. As the levees approach completion, higher and higher grades must inevitably result. As long as there is relief of the waters by incompleted levees and crevasses, the necessary height can not be determined. When the remaining 29 per cent, of levee is constructed and the system withstands one of the greater floods without a crevasse, the excessive flood will determine the height of the levees. Even then certain factors of flood conditions may so unite as to cause a flood which will overtop the system. The statement of the character and condition of the system under the strain of the 1903 flood indicates that there is much left to be done. In the upper part of the basin, the levees were reported too low and of insufficient dimensions. The high water reached to the top for one half of the entire length of the lower St. Francis district and for many miles was above the tops, being restrained from spreading over the basin by capping the levees with planks, dirt and sand-bags. In the district below, it is reported that because of the settlement of the embankment and because about 20 miles of the line in the lower end of the district had not been raised to the provisional grade, considerable work was required to prevent the water from overflowing the levees. Again in Louisiana, the topping of the levees by planks and sand-filled bags was necessary over a distance of 71 miles of the line in order to prevent a wash-over. An engineer reports that the most vulnerable feature is the instability of the foundation along much of the levee line.

The efficiency of the levee system is the test of all the labor and the justification of the large expenditures. The increasing efficiency may be measured in many ways. A comparison of the number of crevasses and the total number of miles of destroyed levees with the records of previous floods is the one in general use. In 1882, 284 crevasses were recorded and 59 miles of levee were destroyed. This record has been gradually improved. In 1890, but 23 crevasses were reported and 4.25 miles of levee destroyed. In 1897, the number of crevasses increased to 49, with a loss of 8.3 miles of levee. In the 1903 flood, 9 crevasses of importance were recorded and 5 of these caused a loss of 2.1 miles of the levees. The loss of levees by caving banks was a little less than 1 per cent, of the entire contents. Last year (1904) in a period of quieter flood the percentage was over 2.5. The number of square miles of overflowed area in 1903 was.5 the mileage for 1897. There is no doubt but that the levee system as it approaches completion is being made stronger and safer. Yet each crevasse or natural break spreading the confined waters over larger areas releases the tension on the banks and to some extent prevents others from occurring. To enclose the water which has spread naturally over 29,700 square miles between two walls less than 5 miles apart and covering about one tenth of its former area is no easy task. Till the present system is completed, the possibility of a flood will be uncertain that always grave dangers may be incurred to life and property within the limits of the alluvial basin of the river.

The commissioners in their general report furnish the best proof of the increased confidence in the levees by citing the progress and growth of the Yazoo Basin since their board was created. The population of this district was 94,672 in 1880, and 195,346 in 1900. The present valuation of the basin is $42,000,000. The number of banks have increased from 2 in 1893 to 51; the mileage of railroads from 225 in 1884 to 816. The cotton production in 1879 was 185,868 bales; in 1903, 426,414. The increase in corn, peas, clover and alfalfa is reported to be even greater than that of cotton. The original timber of the basin is being cleared, and there are now large shipments of lumber, as logs, boards, staves, headings and the like. Flourishing crops are seen to-day where in former years the floods measured from 20 to 25 feet in depth. 'The Yazoo Basin is sprinkled with towns whose sites were the home of the bear and the wild cat ten years ago.'

The engineers of this district in their reports of the flood of 1903 state that about one fourth of the Yazoo Basin was under water during this flood. Two crevasses occurred, letting water into the basin. One of these, three miles below Greenville, Mississippi, was at its greatest width a breach of 3,900 feet. The water flowed back upon Greenville and more than one half of that city was under water. Forces were set to work building a protection levee in the city and moving goods to a place of safety. As far as can be ascertained no lives were lost. Little or no loss in crops was sustained, as the flood came before the planting and the area was largely drained before that season arrived. The losses were mainly in live stock, fences and buildings. In contrast with the security so easily shown, in the reports of the commission, to be the experience of the inhabitants of the Yazoo Basin, the engineers in charge of the levees seem to congratulate themselves that no other crevasses occurred, for many weak places developed that required the utmost care and attention to hold intact. They state as their expectation that crevasses are liable to occur at any high-water season and at any point in the system or 'until all levees are brought up to a sufficient section to withstand the long-continued strain due to the water remaining for weeks near the top.' Although the statement quoted, on the face of it, rather begs the question, we are at liberty to infer therefrom that too much confidence had better not be held in the protective value of much of the present line. So near to disaster do the floods approach oftentimes, that every element which the engineers can control is considered a necessary ally in cooperation for the protection of the levees. During the 1903 flood, the river boats were required to run at a reduced speed along a portion of this basin front. So full was the river and the waters stood so near the top of the levee that it was not considered wise to subject the embankment to the wash of passing steamers. It may be stated in this connection that a storm, arising as the water is nearing the crest of the levees, can not be so summarily dealt with; and it often causes a day or two of apprehension, if indeed the beating waves do not tear their way through the structure.

A harsher note is sounded by an observer of the Weather Bureau[3] than is struck in the reports of the commission. There were favorable and mitigating circumstances which decreased the volume of the 1903 flood. Two factors materially modified the destructive feature of the flood; one of these was an early occurrence and the other a shorter duration. Both these factors spring from the same cause. Notwithstanding these compensating qualities this writer reports that the water during a period of two weeks was higher at Greenville and Arkansas City than it had ever before been known to be. The destructive work of the flood is summed up as follows: Some loss of stock in the basin; 115 houses evacuated in Greenville; 200 acres of fine farming land badly washed and left covered with sand; suspension of traffic on the Yazoo and Mississippi Valley Bailroad from March 27 to April 17, and on the Riverside Division from March 27 to May 7; 1,460 square miles of land overflowed in this basin, one half of which was farm land; 60,000 people lived in the overflowed district and were, therefore, inconvenienced; this number of people represents about one third of the inhabitants of the basin. For some time previous to the coming of the flood, the dwellers in the basin were preparing for the flood season. Mounds were built for temporary refuge. Stationary platforms were constructed to the same end. Rafts were also made. The mules, horses and the feed were in many instances transferred to places of safety, often to the lofts of the barns. Farm implements and machinery were put beyond the reach of the water. That the warning of the Weather Bureau was so extensively heeded explains why there was no loss of life and little loss of stock.

These three reports of the same thing are not so contradictory as they sound. Each observer is looking for the things that sustain him and his point of view, and is not directly interested in the things that are foreign. One writer tries to establish the security of the basin of the Yazoo against danger to life and property, because that is what the board was created to do; a second writer tries to show how weak the levee is, in order to press home the need of funds—and he makes imminent danger to the basin area a means; the third shows that without the services of the branch he represents, the loss of property and life would be multiplied. The first man is right to some extent, and he is sustained by the second, who sees how near to each other danger and safety sometimes approach—and they are aided by the third. I doubt not but that the Weather Bureau may make as just a claim for the credit of the progress in the Yazoo Basin as the River Commission.

If the increasingly better reports influence a larger population and larger expenditures in holdings within the alluvial basin of the Mississippi, and the hopes of the engineers become realized to the extent of normal safety, then, perhaps, the levee system can be called efficient. Twenty years may be too short a time to consider the effect of the system upon population, and at the same time we must remember that but two thirds of the levee lines are completed, yet in this time the commissioners report an increase of population over the Yazoo Basin of over 100 per cent. It seems as if the people were becoming confident that there is 'security and permanence of protection' in the work that is in progress. Yet just so far as this confidence is expressed in settlement within the area liable to overflow, so much further must the levees protect beyond peradventure of disaster. In an increase of 100 per cent, in population and a decrease of 50 per cent, in mileage of overflow, if the terms are commensurate, there is no gain; if the terms are incommensurate, there is as good a chance for a loss as a gain. Just meeting the limit of strain, or preventing a break only by excessive vigilance and energy, or saving from disaster by some mitigating circumstance is not the end to be aimed at; but to be as reasonably sure as it is given man to be that an overflow can not occur must be the plan.

  1. Compiled, largely, from the reports of the Mississippi River Commission.
  2. High water, 1903, and low water, 1895, are reckoned from the Memphis datum. The numbers are, roughly, seven feet too high for Gulf levels as the base.
  3. Bull. M, 'The Floods of the Spring of 1903 in the Mississippi Watershed,' H. C. Frankenfield.