TR(BR)-110 : Semi-pervious stream and aquifer interaction

Abstract

For many years hydrologists have been interested in examining the relationship between flow in a stream and flow in an adjacent aquifer when hydraulic connection exists between the two. The exchange of flow between an aquifer and a stream influences the flow pattern within the aquifer and the quantity of flow in the stream. The exchange of flow depends upon the difference between the hydraulic head in the aquifer and the stream-stage. Valuable contribution to the understanding of stream-aquifer relationship have been made by Cooper and Rorabaugh(1963), Morel-Seytoux(1979, 1988), Morel-Seytoux and Daly(1975), Mishra (1987), etc.

In most of the stream-aquifer interaction studies, stream has been assumed to penetrate the entire depth of aquifer. Very few studies (e.g. , Halek and Svec, 1979; Hall and Moench, 1972; Morel-Seytoux and Daly, 1975, Mishra,1987, etc.) have analyzed the interaction of a partially penetrating stream and an aquifer. Even for fully penetrating stream, temporal variation of stream-stage has been taken care of only in few studies ( e.g. , Cooper and Rorabaugh, 1963 and Morel-Seytoux, 1988). The studies of Morel-Seytoux and Daly,1975 ; Morel-Seytoux, 1988; and Mishra, 1987 are applicable foriany variation of stream-stage with time.

In this report, the concept of 'retardation coefficient' or 'substitute length' as proposed by Hantush(1965) and Halek and Svec(1979) respectively has been used to model a semi-pervious stream. The convolution technique has been employed to determine the aquifer response to time-varying stream-stage. The perturbation has been approximated as a train of pulses and accordingly, discrete kernels have been obtained and system response has been computed. The aquifer response in terms of the rate of flow and the cumulative volume of flow at any section and the influent seepage have been studied. The effect of 'substitute length' upon the system response has also been discussed.

A model has been developed using discrete pulse kernel approach for determining the aquifer responses due to the time-varying stage in the semi-pervious stream. The model takes into account any type of variation in the stream-stage. The aquifer responses that have been analyzed, include:

i) Piezometric head at a section ii) rate of influent seepage iii) rate of flow at a section iv) cumulative volume of influent seepage v) cumulative volume of flow at a section vi) volume of return flow during a time-step vii) volume of flow passing through a section during a time step

The following conclusions have been arrived at:

i) The perturbation that travels in the aquifer consequent to a symmetric flood wave is skewed to the right. ii) The semi-pervious bed of the stream offers a resistance to the flow which is represented by 'substitute length'. The effect of semi-perviouness is of the the stream is to reduce the total volume of ;,fluent and effluent seepage. The effective recharge from a stream decreases with increase in R.