Simulation of Density and Flow Dynamics in a Lagoon Aquifer Environment and Implications for Nutrient Delivery From Land to Sea

Sascha Müller,Torben O Sonnenborg,Peter Engesgaard,Søren Jessen,Rena Meyer

doi:10.3389/frwa.2021.773859

Abstract

The near coastal zone, hosting the saltwater-freshwater interface, is an important zone that nutrients from terrestrial freshwaters have to pass to reach marine environments. This zone functions as a highly reactive biogeochemical reactor, for which nutrient cycling and budget is controlled by the water circulation within and across that interface. This study addresses the seasonal variation in water circulation, salinity pattern and the temporal seawater-freshwater exchange dynamics at the saltwater-wedge. This is achieved by linking geophysical exploration and numerical modeling to hydrochemical and hydraulic head observations from a lagoon site at the west coast of Denmark. The hydrochemical data from earlier studies suggests that increased inland recharge during winter drives a saltwater-wedge regression (seaward movement) whereas low recharge during summer causes a wedge transgression. Transient variable density model simulations reproduce only the hydraulic head dynamics in response to recharge dynamics, while the salinity distribution across the saltwater wedge cannot be reproduced with accuracy. A dynamic wedge is only simulated in the shallow part of the aquifer (&lt;5 m), while the deeper parts are rather unaffected by fluctuations in freshwater inputs. Fluctuating salinity concentrations in the lagoon cause the development of a temporary intertidal salinity cell. This leads to a reversed density pattern in the underlying aquifer and the development of a freshwater containing discharge tube, which is confined by an overlying and underlying zone of saltwater. This process can explain observed trends in the in-situ data, despite an offset in absolute concentrations. Geophysical data indicates the presence of a deeper low hydraulic conductive unit, which coincides with the stagnant parts of the simulated saltwater-wedge. Thus, exchange fluxes refreshing the deeper low permeable areas are reduced. Consequently, this study suggests a very significant seasonal water circulation within the coastal aquifer near the seawater-freshwater interface, which is governed by the hydrogeological setting and the incoming freshwater fluxes, where nutrient delivery is limited to a small corridor of the shallow part of the aquifer.

Highlights

Lagoons cover 13% worldwide and 5.3% of European coastal areas (Barnes, 1980; Kjerfve, 1994)
A range of fresh groundwater fluxes discharging into the transect area from east were estimate based on (i) local hydraulic head gradient between the outermost piezometer of the piezometer transect J1d and the lagoon stage or ii) the regional hydraulic head gradients between inland wells (Supplementary Table S2) and the lagoon
The Scenario A model did not contain a peat unit on top of the sand, and it used the dynamic lagoon and groundwater flow boundary conditions (Table 1). This scenario did not reproduce the temporal dynamics of hydraulic head or salinity concentration changes in piezometer J9 (R2 = 0.05, Root Mean Square Error (RMSE) = 0.16 m, and Nash Sutcliff Efficiency (NSE) = −0.09, Figure 7E; Table 1)

Summary

Introduction

Lagoons cover 13% worldwide and 5.3% of European coastal areas (Barnes, 1980; Kjerfve, 1994). The ISC causes an inverse density vs depth distribution and creates a freshwater discharge tube to form between the ISC and the underlying permanent (but not necessarily stationary) saltwater wedge (Michael et al, 2005; Vandenbohede and Lebbe, 2006; Robinson et al, 2007; Heiss and Michael, 2014). This directs freshwater discharge to saline surface waters (lagoons) as SGD. SGD is controlled by density driven circulation patterns, wave setup, tidal pumping (Robinson et al, 2007; Tamborski et al, 2015), subsurface smallscale heterogeneities (Taniguchi et al, 2002), seasonal changes in terrestrial hydraulic gradient and annual recharge cycling (Michael et al, 2005; Burnett et al, 2006)

Results

Discussion

Conclusion