Abstract:
Springs are natural outlets through which the groundwater emerges at the ground surface as concentrated discharge from an aquifer. Discharge rate from a spring depends on precipitation and other hydrometeorological factors, the geology and geomorphology of the area, the aquifer geometry and its characteristics. Fluctuation of spring flow are due to variation in precipitation, hydrolmeterological parameters like temperature, changes in land use. Springs with a low discharge in variability are normally found in volcanic and sandstone formations. Springs located in mountains usually exhibit fluctuations.
Further,. atmospheric processes induce randomness into the precipitation and temperature series affecting recharge to springs and subsequent spring flow. Land use change affect infiltration characteristics over the catchment area of the spring. A spring flow time series consist of trend, periodicity autoregressive and random residual components. As such, in order to improve the modelling efforts to predict spring flow, a time series analysis of spring flow is necessary. For time series analysis of spring flow, a fairly long time series of flow data is necessary. As a long duration flow data is not available in any Indian spring, the 70 - year of monthly data of spring flow data from Sulkovy Prameny Spring emanating from sandstone strata in Czeckoslovakia has been used. It was found that first order autoregressive model though simulate the field flow data, but it cannot take care of month to month dependence especially between rising and falling limb of spring hydrograph. Thomas Fiering Model is used to generate data which obviates this difficulty and the model simulate the data reasonably well except for cases when flow data differs sharply between month to month.
