A Multiscale Investigation of Ground Water Flow at Clear Lake, Iowa

Abstract

Ground water flow was investigated at Clear Lake, a 1468-ha glacial lake in north-central Iowa, as part of a comprehensive water quality study. A multiscale approach, consisting of seepage meters (and a potentiomanometer), Darcy's law, and an analytic element (AE) model, was used to estimate ground water inflow to and outflow from the lake. Estimates from the three methods disagreed. Seepage meters recorded a median-specific discharge of 0.25 mum/s, which produced a lake inflow rate between 90,750 and 138,200 m3/d, but no detectable outflow. A wave-induced Bernoulli effect probably compromised both inflow and outflow measurements. Darcy's law was applied to 11 zones around the lake, producing inflow and outflow values of 10,500 and 5000 m3/d, respectively. The AE model, GFLOW, coupled with the parameter estimation model, UCODE, simulated ground water flow in a 700-km2 region using 31 hydraulic head and base flow measurements as calibration targets. The model produced ground water inflow and outflow rates of 14,300 and 9200 m3/d, respectively. Although not a substitute for field data, the model's ability to simulate ground water flow to the lake and the region, estimate uncertainty for model parameters, and calculate a lake stage and associated lake water balance makes it a powerful tool for water quality management and an attractive alternative to the traditional methods of ground water/lake investigation.

Comparison of Darcian flux calculations and seepage meter measurements in a sandy streambed in North Carolina, United States

“…Darcian flux calculations have been used in many field studies of groundwater interaction with streams [ Cey et al , 1998; Conant , 2004; Chen et al , 2009; Kennedy et al , 2007, 2008, 2009a, 2009b], whereas other studies have employed seepage meters [ Lee and Hynes , 1978; Libelo and McIntyre , 1994; Blanchfield and Ridgway , 1996; Cey et al , 1998; Landon et al , 2001; Murdoch and Kelly , 2003; Craig , 2005; Rosenberry , 2008; Rosenberry and Pitlick , 2009]. The apparent simplicity of the two methods makes them attractive tools, but agreement between them has been only poor to modest in the few cases in which they have been applied at the same time and/or place in streams [ Cey et al , 1998; Rosenberry and Pitlick , 2009], lakes [ Lee and Cherry , 1978; Woessner and Sullivan , 1984; Shaw et al , 1990; Kishel and Gerla , 2002; Simpkins , 2006] and coastal zones [ Robinson et al , 1998; Bokuniewicz et al , 2004; Mulligan and Charette , 2006].…”

“…Several studies have used both methods to quantify groundwater exchange with lakes and inputs of submarine groundwater discharge (SGD) to coastal zones, with results generally showing poor overall agreement between the two methods [e.g., Lee and Cherry , 1978; Woessner and Sullivan , 1984; Robinson et al , 1998; Shaw et al , 1990; Kishel and Gerla , 2002; Bokuniewicz et al , 2004; Mulligan and Charette , 2006; Simpkins , 2006]. For example, Simpkins [2006] conducted 341 seepage meter measurements at 21 sites and 33 Darcian flux calculations at 11 sites (in nests of 3 piezometers) over a period of approximately two years, and found that groundwater seepage into Clear Lake (Iowa, USA) was between 90,750 and 138,200 m 3 /d based on the seepage meter measurements and between −5,000 (outflow from lake) and 10,500 m 3 /d based on the Darcian flux calculations (though the author does not report whether high v dacry was measured at the same time as high v meter ). Simpkins [2006] attributes at least part of the difference between the two methods to biased‐high seepage meter values, because the seepage bags were not placed in protective shelters.…”

Analytic Element Ground Water Modeling as a Research Program (1980 to 2006)

“…A new generation of researchers are in place as professors in academia. AEM does outperform numerical methods in some areas, such as including local detail in large regional models (Simpkins 2006), and as numerical laboratories for study of macro‐scale dispersion (Jankovic et al 2006).…”

Ground Water Modeling Applications Using the Analytic Element Method

“…Rather than abandoning such a model and replacing it with a more complex model as more data and insight become available, the initial model is gradually upgraded to a more complex and realistic representation of the hydrological regime (Wittman et al 1996; Dripps et al 2006; Johnson and Mifflin 2006). The focus on parsimonious designs also made stepwise analytic element models well suited for nonlinear parameter estimation techniques for model calibration (Power and Barnes 1993; Simpkins 2006), prediction (Hunt et al 2000b; Kelson et al 2002), and designing future data collection efforts (Hunt 2002). In some cases, the end result of the stepwise process is an analytic element model sufficient to answer the questions that initiated the modeling.…”