One of the techniques which are used to increase the water yield of springs during droughts, is the drilling of wells upstream or downstream the springs (BONI, 1968; CELICO, 1982; BONI & PETITTA, 1994). This technique implies low well efficiency, high drawdowns, high cost of withdrawals and it can lead to spring drainage. In this work a finite difference mathematical model of the Scirca spring (Umbria - Marche Apennines) was developed, which allows to simulate the effects of pumping wells located at different distances from the spring; the model was built by means of the modflow code (MCDONALD & HARBAUGH, 1984), using the porous equivalent
approach (ANDERSON & WOESSNER, 1992; ANGELINI & DRAGONI, 1997; SCANLON ET ALII, 2003).
The only input parameter for the model is the aquifer recharge, which tends to nil during the dry season, when effective rainfall is very low. The implemented model is able to simulate both the spring discharge during recession, and the variations of daily discharges, continuously measured from 1996 to 2011. Due to the lack of piezometric data and of direct measurements of hydrogeological parameters (hydraulic conductivity and effective porosity), it has been necessary to
run different simulations in order to obtain a set of Trasmissivity and Specific Yield values able to reproduce the average recession curve of Scirca spring. Afterwards a process of “inverse modelling” (DOHERTY, 2000; CARRERA ET ALII, 2005; HILL & TIEDEMAN, 2007) was performed to estimate, by means of calibration, the daily recharge of the hydro-geological system of Scirca spring throughout
several years. In order to investigate the degree of correlation between daily rainfall and simulated recharge, the two time series data sets had undergone MA Analysis; this kind of analysis allowed to “clean up” the data from the influence of short precipitations close to each other, and to highlight a periodic, similar and in phase trend of rain and recharge. An absolute maximum of recharge in the
fall-winter period and a relative maximum in the spring season, were detected. Finally, the efficiency of the management schemes was evaluated by simulating the reaction of the spring, in terms of discharge, to different pumping scenarios; in every scheme the total withdrawal (spring + well) during summer was much larger than the natural spring discharge. The wells were located at different distances from the spring (maximum 2850 m, minimum 100 m), the pumping time span was fixed to 90 days, and 3 pumping rate were used (60, 90, 120 l/s). Results show that the maximum discharge for which the drainage of the spring is avoided and the minimum vital flow (MVF) is guaranteed is 90 l/s. The higher water volumes extracted during summer are balanced by a lowering of the maximum natural discharges. Simulations indicate that, by drilling pumping wells
far from the spring, it is possible to optimize the efficiency of the whole system in terms of total withdrawal, drilling and management costs, and to reduce the environmental impact. According to the model, for the Scirca system, the optimum distance of a pumping well from the spring is 2850 m, whereas the optimum pumping scheme is a discharge of 90 l/s, running three months a year
(June, July and August). The mathematical model shows that the Scirca hydrogeological system needs 625 days to restore its “undisturbed” state, with a tolerance of 0.5 l/s. The model highlights the possibility to force the system to supply a lower amount of water in winter, with the aim to increase the summer yield. Such a management scheme can be of use to better meet the water demand during dry seasons.