U.S. Department of Defense


Date of this Version



Man-made Lakes: Their Problems and Environmental Effects, 1991


U.S. Government work


It must be recognized that, when a reservoir is constructed on a stream, all or most of the sediment transported into the reservoir by the stream will be deposited there. With few exceptions, there is no practical method for reducing or eliminating the sediment inflow; thus it is necessary to anticipate and to provide for resulting problems. The major problem is, of course, the accumulative loss of storage capacity. Other items of importance include the distribution of the sediment with respect to various storage increments, the effect on the chemical or physical quality of the water, ecological effects, and the possible degradation of the channel downstream.

The prediction of the quantity of sediment that will be transported into the reservoir is largely a matter of experienced judgment. During the past 20 years, there have been many stations established for the measurement of suspended sediment discharge, but the cost of operating these stations is such that they are restricted in number and location to a relatively few index areas or to specific locations where they may be operated for only a few years. Because of the normal variations in the hydrologic cycle a station must usually have a record of 10-30 years (depending on the physical and hydrologic characteristics of the area) in order to provide a dependable average. Since it is seldom that the need for sediment data at a specific location can be predicted far in advance, it is also seldom that adequate records are available.

In the Missouri River division of the U.S. Army Corps of Engineers we maintain certain index stations continuously. In regions where the sediment discharge is reasonably consistent we operate stations at one location for 5-10 years; then, providing for an overlap of 1 or 2 years, we move to other locations to establish index values for the region. As soon as the need for data at a specific location is known and if the time available for study is adequate, we establish a station at that location.

Finally, we make an extensive ground reconnaissance of the drainage area above the site. Information from this reconnaissance is correlated with all available data from stations measuring suspended sediment discharge from comparable areas and is integrated to derive an estimate of the average annual sediment inflow to be anticipated. It is necessary to add to this estimate some quantity to account for sediments moving along the stream bed and not measured in suspended load sampling. In streams having a coarse gravel bed, this load can be computed with reasonable adequacy by bed load formulas; however, in sand bed streams it is believed that an estimate based on judgment is equally or perhaps more accurate. In small reservoirs the estimate may be reduced to account for sediment that might be transported through the pool, but, where the drainage area contributing to the project is greater than about 250 km', the reservoir will normally be large enough to retain all the in flowing sediment. If the project is constructed under the auspices of the federal government, sufficient storage is provided to retain the anticipated sediment load for a 100-year period without encroaching on the primary project purposes.

In some instances it is possible that the storage required for sediment might be reduced by upstream control measures. It has been demonstrated, for example, that the sediment contribution from very small drainage areas (1-5 km') can be reduced by about 85% by intensive soil conservation procedures. In another instance, sediment discharge reductions due to improved land management and conservation on a group of drainages varying in size from several hundred to 5000 km' were indicated to be 10-35%. This latter study, however, covered a period of only 3 years under the improved regimen and could well have been influenced by favorable hydrologic circumstances.