Response to Comment by H. Lough, Department of Civil Engineering, University of Canterbury, Christchurch, New Zealand, on the Paper " Stream Depletion Predictions using Pumping Test Data from A Heterogeneous Stream-Aquifer System (A Case Study from the Gr

• Published in: Journal of hydrology (Amsterdam) Vol. 313
• Main Authors: Kollet, S J, Zlotnik, V A
• Format: Journal Article
• Language: English
• Published: United States 20.12.2004
• Subjects: AQUIFERS; CIVIL ENGINEERING; DRAWDOWN; ENVIRONMENTAL SCIENCES; FRACTURES; GEOSCIENCES; POROSITY; PUMPING; SIMULATION; STREAMS
• ISSN: 0022-1694; 1879-2707

We thank H. Lough for her interest in our data set and the attempt to re-analyze our results (Kollet and Zlotnik, 2003) using the recent model by Hunt (2003). We welcome others to share our unique data set of the pumping test from the Prairie Creek site, Nebraska, USA. Nevertheless we believe that this particular attempt failed, because H. Lough selected a wrong model of semi-confined aquifer conditions for the interpretation of the pumping test data, which was collected in an unconfined aquifer. H. Lough based her selection on the three distinct drawdown segments observed during the test. It is well known that geologically distinct aquifers can yield a three-segment drawdown response under pumping conditions (e.g., Streltsova, 1988). Examples include unconfined aquifers (e.g., Neuman, 1972; Moench, 1997), aquifers with double porosity or fractures (e.g., Barenblatt et al., 1960; Boulton and Streltsova-Adams, 1978), and (semi-) confined aquifers in contact with aquitards (e.g. Cooley and Case, 1973; Moench, 1985). At the Prairie Creek site the aquifer is unconfined. The interpretation of the pumping test data collected at the site using type curves that are valid for an aquifer-aquitard system is a mistake. In fact, this approach illustrates a typical problem associated with inverse modeling: drastically different models can closely reproduce a system response and yield some parameter estimates, although the models do not represent the real system adequately. Here, the improper model yields some parameter estimates for an aquitard, although the aquitard does not exist at the Prairie Creek test site. We must also unequivocally state that the model by Hunt (2003) is clearly formulated and correct for stream-aquifer-aquitard systems within the stated limitations (pumping wells screened only in the lowest stratigraphic layer, etc.). However, the Hunt (1999) or BZT (Butler et al., 2001) models should be used for interpreting pumping tests near streams in non-leaky aquifers as outlined in our study (Kollet and Zlotnik, 2003). The purpose of the comment by H. Lough is to examine three drawdown segments and results from Kollet and Zlotnik (2003) using a newer analytical model of stream-aquifer interactions by Hunt (2003). We will address the key issues of this comment in this paper.