VOL. 25, NO. 17, December 23 2005 (Paush 08, 2062 B.S. )

Sediment Movement: Kosi River

- By DR. AB THAPA

Kosi floods have been regarded in the past as the synonymous with unlimited destruction and devastation, ruins of towns and villages, laying vast areas of agricultural lands bare with sand deposits. Prof. C.V. Gole and S.V. Chitale have written about the Kosi floods “ the Kosi River is known as the river of sorrow of North Bihar. After Indus, Brahmaputra rivers, the third largest river emptying from Himalayas, the Kosi has shifted from east to west over 70 miles in the last 200 years, spreading devastation and laying an area of approximately 3000 sq. miles bare with sand deposits.” The Kosi carries with it an enormous sediment load. It unloads sediments into the main drainage channel and builds up inland delta.

Major Rennel had perhaps been the earliest person to conduct investigations of the Kosi River, and thereafter Mr. James Ferguson F.R.S. further investigated it. After this Mr. F.A. Shillingfields ( 1893) had made observations of the Kosi for about 25 years. In the 20 th century many scientists and engineers were involved in the study of the Kosi. The main purpose of the study was initially to control the Kosi floods.

Oscillation of Kosi River Channel

Past studies have confirmed that the channel of the Kosi River had been oscillating from east to west between two boundaries far away from one another. . Chatra had always been the starting point of the channel. However, the location of the place where it emptied into another major river varied considerably. According to first theory the Kosi River flowed eastward across the present course of the Mahananda, and eventually joined the Brahmaputra River. The bed of the Kosi oscillated over a vast tract of land from Brahmaputra to near the mouth of the Gandak River in the west. The oscillation was being repeated at a long interval of time. According to another theory the Kosi River channel oscillated only up to Purnea and joined the Ganges River opposite Rajmahal.

Mr. F.A. Shillingfield, referring to the Kosi oscillations, had said that in contradiction to the slow western movement, the eastward movement would take place in one great sweep. In 1941, Sir C.C. Inglis, Director of Central Water and Hydro-Dynamic Research Station;, Poona, identified the problem of oscillation of the Kosi channel, as being due to an excessive charge of the sand that Kosi waters carry. The Kosi River carries a very high load of sediments, which vary between 2000 and 5000 mg per litre. It is regarded that the maximum concentration may even be as high as 10,000 mg per liter. Leopold and Maddock (1954) considering Kosi behavior had stated that a river like Kosi tends to shift laterally at a rate dependent on the rate of accumulation of material being deposited. As one course becomes higher than possible adjacent paths, the river would shift.

Ganges Changes and Kosi

It is noticed that beyond the mouth of the Kosi there is a sudden turn of the Ganges towards nearly southwardly direction in the vicinity of Rajmahal. Upstream of this, the Ganges is flowing from west to east. On the north, the Kosi Delta prevents the Ganges from moving towards north while the Rajmahal hills prevent the movement of the Ganges straight to the south Thus, to the south of the Kosi delta the present position of the Ganges is fixed. According to the study of Prof. Gole et al, if Kosi is left to itself , the delta building process would continue and the river would shift from east to west and back to east and so on, over the cone. One such sweep of the Kosi from east to west had taken place about 250 years ago. With an increased rate of sediment yield, as being seen now, the period of swing would be becoming smaller in the future. The limit of the delta cone with every swing is expected to extend beyond previous limits, both on the east and west boundaries as well as on the south. Prof. Gole et al have concluded that the Kosi, if left to itself, would ultimately push the Kosi Delta into the Ganges when major changes in the Ganges itself could be expected to occur.

Theory of Sediment Flow

A river is a flowing body of water in an open channel. The characteristics of the flowing water are the domain of the hydraulic engineering, whereas the dimensions and pattern of the open channel are geomorphic problems. The flowing water not only moves through, but it transports with it a variety of materials. The total sediment load of the channel is comprised of a chemical or dissolved load, a suspended load, fine particles that are kept in suspension by the turbulence of the flowing water, and a bedload composed of coarser particles that are moved on or near the bed of the stream. The sediments are moved by the transport capacity of water that is in the channel.

It can be assumed that the quantity of water that fills or nearly fills the channels ( i.e. the bankfull discharge) is the dominant discharge that is responsible for establishing the characteristics of the channel. The flow velocity and the manner in which the water flows through the channel are important considerations in determining the morphology of that channel.

Water in open channels is subjected to two principal forces: the force of gravity which propels the water downslope, and frictional forces between the water molecules themselves and between the water and the channel boundaries which together resist the downslope movement.

It is relatively easy to obtain information on the water discharge through an open channel, but it is far more difficult to sample the sediment load of a stream. This is due to the fact that there are two distinct types of sediment movement – suspended sediment load and bed load. Suspended sediment is easy to describe and to sample though its character changes as velocity increases. The movement of pure bedload is much more difficult to estimate than is that of suspended load because attempts to measure it severely interfere with the movement itself and because bed movement is influenced by a complex association of bed variables, including grain size, sorting, shape, roundness and larger-scale bed environment configuration.

Hans Einstein (1940), the son of the famous physicist Albert Einstein had calculated a theoretical bedload function expressing bedload movement with respect to various debris sizes applied to Big Sand Creek, Mississippi River. For many years the transport function for the bed load had been based upon the classical Du Boy’s relation. Shield gave the transport function which is a modified form of Du Boy’s expression.

The Braided Channel

A braided river like Kosi is relatively wide with poorly defined unstable banks, and is characterized by a steep shallow water course with multiple channel divisions around alluvial islands. Braiding is one of many patterns which can maintain quasi-equilibrium among the variables of discharge, sediment load, and transporting ability. Lane (1957) concluded that, generally, the two primary causes that may be responsible for the braided condition are: (1) the stream may be supplied with more sediment than it can carry resulting in deposition of a part of the load, and (2) steep slopes, which produce a wide shallow channel where bars and islands form readily.

Either of the above factors alone, or both in concert, could be responsible for a braided pattern. If the channel is overloaded with sediment, deposition occurs, the bed aggrades, and the slope of the channel increases in an effort to maintain a graded condition. As the channel steepens, the velocity increases, multiple channels develop and cause the overall channel system to widen. The multiple channels, which form when bars of sediment accumulate within the main channel, are generally unstable and change position with both time and stage.

Another cause of braiding is if the banks are easily erodable. Where the banks are easily eroded, the stream widens at high flow and at low flow bars form which become stable. Thus islands are formed.

In Conclusion

In general, a braided channel of a river, like Kosi, has a steep slope, a large bed-material load in comparison with its suspended load, and a relatively small amount of silts and clays in the bed and banks. The braided streams have the characteristics that it is unstable, changes rapidly its alignment, and carries a large quantity of sediments. It is very wide and shallow even at flood flow. It is in general unpredictable.

(Dr. Thapa writes on water resources)