Variable-density Flow Model Research Articles

Efficient Stochastic Simulation of Seawater Intrusion, With Mixing, in Confined Coastal Aquifers

We present a high-efficiency method for simulating seawater intrusion (SWI), with mixing, in confined coastal aquifers based on uncoupled equations in the through-flow region of the aquifer. The flow field is calculated analytically and the tracer transport numerically, via spatial splitting along the principal directions (PD) of transport. Advection-dispersion processes along streamlines are simulated with the very efficient matched artificial dispersivity (MAD) method of Syriopoulou and Koussis and the system of discretized transverse-dispersion equations is solved with the Thomas algorithm. These concepts are embedded in the 2D-MADPD-SWI model, yielding comparable solutions to those of the uncoupled SWI equations with the state-of-the-art FEFLOW code, but faster, while 2D-MADPD-SWI achieves an at least hundredfold faster solution than a variable-density flow model. We demonstrate the utility of the 2D-MADPD-SWI model in stochastic Monte Carlo simulations by assessing the uncertainty on the advance of the 1,500 ppm TDS line (limit of tolerable salinity for irrigation) due to randomly variable hydraulic conductivity and freshwater flow rate.

Numerical modeling of seawater intrusion in karstic aquifer, Northwestern Coast of Egypt

The growing population and increasing of the urbanization activities at the Egyptian coastal areas create a great demand on groundwater. El Negila which is located on the Northwestern coast of Egypt is of great environmental, economic, and social importance, so the sustainable management and preservation of the groundwater quantity and quality is imperative. Such task can be addressed through modeling tools that take into account both the human impacts and the effects of climate changes on the hydrogeological system. A variable-density flow model (SEAWAT) had been used to investigate the extent of the seawater intrusion the study area, and to introduce an effective management policy for sustainable development of the groundwater system. An attempt had been made for model calibration; based on adjusting the hydraulic head of the hydrogeological system with respect to the lineaments frequency in the study area. The calibrated numerical model showed the limited extent of the seawater intrusion in the study area under current recharge and abstraction conditions, in addition to that, the model was validated using hydrochemical and stable isotopic data. The aquifer response to different stressing pumping scenario has been tested and predicted up to the year 2022. Furthermore, the impact of rising the seawater level as a result of rapid climatic change was assessed, with pumping stress of the second scenario. It was concluded that the tableland area acts as recharge source and providing an opportunity of recharging the karstic aquifer in the study area especially through the lineaments. The extent of seawater intrusion in the study area for the karstic aquifer is limited, and would increase under the induced activities of pumping as well as seawater rise conditions. A range of withdrawal rate from the study area was recommended to avoid more invasions from seawater.

Seawater intrusion in karstic, coastal aquifers: Current challenges and future scenarios in the Taranto area (southern Italy)

Mediterranean areas are characterized by complex hydrogeological systems, where management of freshwater resources, mostly stored in karstic, coastal aquifers, is necessary and requires the application of numerical tools to detect and prevent deterioration of groundwater, mostly caused by overexploitation.In the Taranto area (southern Italy), the deep, karstic aquifer is the only source of freshwater and satisfies the main human activities. Preserving quantity and quality of this system through management policies is so necessary and such task can be addressed through modeling tools which take into account human impacts and the effects of climate changes.A variable-density flow model was developed with SEAWAT to depict the “current” status of the saltwater intrusion, namely the status simulated over an average hydrogeological year. Considering the goals of this analysis and the scale at which the model was built, the equivalent porous medium approach was adopted to represent the deep aquifer.The effects that different flow boundary conditions along the coast have on the transport model were assessed. Furthermore, salinity stratification occurs within a strip spreading between 4km and 7km from the coast in the deep aquifer. The model predicts a similar phenomenon for some submarine freshwater springs and modeling outcomes were positively compared with measurements found in the literature.Two scenarios were simulated to assess the effects of decreased rainfall and increased pumping on saline intrusion. Major differences in the concentration field with respect to the “current” status were found where the hydraulic conductivity of the deep aquifer is higher and such differences are higher when Dirichlet flow boundary conditions are assigned. Furthermore, the Dirichlet boundary condition along the coast for transport modeling influences the concentration field in different scenarios at shallow depths; as such, concentration values simulated under stressed conditions are lower than those simulated under undisturbed conditions.

Assessment of Saltwater Intrusion and Discharges to a Wetland with a 3D Transient Variable Density Flow Model: The Coastal Plane Oropesa-Torreblanca Aquifer, Spain

The proper characterization of coastal aquifers requires modeling variable density flow effects. However, most models estimate processes as saline intrusion based on 2D models with constant density and are rarely calibrated to honor salinity measurements. These facts limit the model predictions reliability, affecting the estimated hydrodynamic parameters, external stresses, and other model outputs that can be critical for planning or management decisions. This paper describes the re-assessment of a coastal aquifer model (Oropesa-Torreblanca, eastern Spain) subjected to moderate-to-high saline intrusion with a transient 3D variable density flow model. Previous models were based on 2D low-resolution grids without variable density effects. The new model honors the observed trends of both piezometric and salinity data. Results show the importance of the variable density effects having on critical outputs as sea intrusion and the discharges to a local wetland of high environmental value. The widespread intrusion process and its current stabilization are confirmed but, compared to previous models, the annual average intrusion is 156% higher, discharge to the wetland increases 30%, and the inflows from neighboring formations increase 22%. The more accurate aquifer models, as well as the new discharges and intrusion estimations, are important contributions for future water and environmental planning decisions in the area.

Groundwater resource sustainability in the Wadi Watir delta, Gulf of Aqaba, Sinai, Egypt

The Wadi Watir delta, in the arid Sinai Peninsula, Egypt, contains an alluvial aquifer underlain by impermeable Precambrian basement rock. The scarcity of rainfall during the last decade, combined with high pumping rates, resulted in degradation of water quality in the main supply wells along the mountain front, which has resulted in reduced groundwater pumping. Additionally, seawater intrusion along the coast has increased salinity in some wells. A three-dimensional (3D) groundwater flow model (MODFLOW) was calibrated using groundwater-level changes and pumping rates from 1982 to 2009; the groundwater recharge rate was estimated to be 1.58 × 106 m3/year. A variable-density flow model (SEAWAT) was used to evaluate seawater intrusion for different pumping rates and well-field locations. Water chemistry and stable isotope data were used to calculate seawater mixing with groundwater along the coast. Geochemical modeling (NETPATH) determined the sources and mixing of different groundwaters from the mountainous recharge areas and within the delta aquifers; results showed that the groundwater salinity is controlled by dissolution of minerals and salts in the aquifers along flow paths and mixing of chemically different waters, including upwelling of saline groundwater and seawater intrusion. Future groundwater pumping must be closely monitored to limit these effects.

Modeling flow and reactive transport to explain mineral zoning in the Atacama salt flat aquifer, Chile

Geochemical processes are typically neglected when modeling groundwater flow in highly saline systems. In the Atacama salt flat, northern Chile, field data indicate that the main aquifer presents a mineralogical zoning consisting of carbonates, sulfates and halite, which is a result of precipitation and/or dissolution reactions. When precipitation and/or dissolution of minerals are active processes in the aquifer, physical properties that control groundwater flow, such as porosity and permeability, would also change. Until now, groundwater flow in the Atacama salt flat has been simulated with conventional constant- or variable-density flow models. This study, for the first time, investigated the importance of incorporating geochemical processes in modeling the Atacama salt flat aquifer. In particular, the impact of mineral precipitation/dissolution on aquifer’s physical properties and flow patterns was studied. To this end, the SHEMAT code was utilized to develop a two-dimensional numerical groundwater flow model in the eastern border of the Atacama salt flat. The SHEMAT code solves the classic groundwater flow equation coupled to the reactive transport of ions and mineral precipitation, with variable density, viscosity, porosity and permeability. When geochemical processes were included in the simulations, the model reproduced reasonably well the mineral zoning evidenced by field observations. Mineral precipitation was responsible of a strong permeability reduction in areas of the aquifer associated with halite and calcite formation. Permeability alterations were responsible of changes in the equivalent freshwater head that resulted in significant flow pattern variations compared to variable-density flow models. These results highlight the importance of considering geochemical processes that modify the aquifer’s physical properties when modeling groundwater flow in highly saline aquifers such as the Atacama salt flat aquifer.

Variable density bore interaction with block obstacles

A horizontally variable density flow model is used to simulate hydraulic bore interactions with idealised urban obstacles. The 2D non-linear shallow water equations are solved using a second-order Monotonic Upstream-centered Schemes for Conservation Laws-Hancock Godunov-type HLLC approximate Riemann scheme. Validation test results are reported for wave propagation over a hump, a constant-density circular dam break and two 1D dam breaks involving different spatial distributions of solute concentration. Detailed parameter studies are then considered for hydraulic bore interactions with single and multiple-square obstacles under subcritical, critical and supercritical flow conditions. In all cases, reflected and diffracted wave patterns are generated immediately after the bore impacts the obstacle(s). Later, the incident bore reconstitutes itself downstream of the obstacle(s). Variable density flows are also considered, with the upstream volumetric concentrations set to values corresponding to water–sediment mixture densities of 1165 and 1495 kg/m3. It is found that the upstream Froude number, gap spacing between obstacles and upstream to downstream density difference influence the strength of the bore–structure interaction, run-up at the front face of the obstacle(s), and subsequent wave–wave interactions.

Using quantitative indicators to evaluate results from variable-density groundwater flow models

Variable-density transport models are typically tested by comparing model output with the results of three standard test cases: (1) the HYDROCOIN Level 1, Case 5 “salt dome” problem—Organisation for Economic Co-operation and Development (OECD 1988), (2) the Henry (1964) approximate analytic solution for steady-state saltwater intrusion and (3) the Elder (1967) problem for complex natural convection where fluid flow is driven purely by fluid-density differences. The complex flow phenomena that result in many variable-density flow problems often means that the intercode testing (and the evaluation of results from variable-density flow models more generally) is largely limited to a visual inspection of isochlor distributions. Visual inspection can often be quite subjective, prone to errors and may not allow easy detection of discrepancies, especially when they are small. Moreover, a match of certain isochlors at a number of prescribed time intervals does not necessarily enure the model is adequately tested. Recently, the well-studied Elder problem has been the subject of significant discussion in the research literature. Authors such as Diersch and Kolditz (2002) have shown that the solutions obtained to the Elder problem are dependent upon the level of grid discretization used. Simpson and Clement (2003) stated that the results of the Elder problem can only be matched in a qualitative sense because the problem is highly sensitive to discretization. In this paper, a number of quantitative indicators are developed that can be used for a more rigorous quantitative evaluation of results of variable-density flow models. They may also find application in future model benchmarking practice.

Determination of aquifer roof extending under the sea from variable-density flow modelling of groundwater response to tidal loading: case study of the Jahe River Basin, Shandong Province, China

The main task of studies on salt-water intrusion into coastal confined aquifers is to predict the position of the fresh- salt-water interface, which can be determined from the length of the aquifer roof extending under the sea. Records of groundwater level affected by tides can be used to infer hydrological conditions and determine hydraulic parameters of an aquifer extending under the sea. In this paper, a three-dimensional, variable-density groundwater flow model has been developed to determine the equivalent roof length of an aquifer extending under the sea from the tidal-effected data of groundwater level in the Jahe River Basin, Shandong Province, China. The seaward boundary is obtained by converging hydraulic head fluctuations observed in drill holes with calculated values, and the aquifer parameters in the extending zone are estimated. The impacts of aquifer roof length and aquifer parameters on the fluctuation of tidal groundwater are studied. It is concluded that the length of the aquifer roof extending under the sea should correspond with certain aquifer parameters in the extrapolation zone. Therefore, the seaward boundary determined from tidal-effect information is the equivalent boundary in hydrodynamic characteristics rather than the true boundary of the confined aquifer

Three-dimensional physical benchmark experiments to test variable-density flow models

Flow and transport in saturated porous media may be influenced, or even dominated, by phenomena of variable-density flow, if solute concentration gradients are high. The numerical codes used for simulation of these situations have to be tested in order to ensure their reliability. A series of laboratory experiments with well-defined experimental parameters for a typical variable-density problem were performed to provide data for such a test. A Nuclear Magnetic Resonance Imaging technique was used, which was appropriate for measuring the three-dimensional salt concentration distribution in a porous medium, including its evolution with time. Two experiments were conducted with 1 and 10% salt mass fraction as initial salt concentration contrasts. These key experiments were accompanied by supplementary ones to determine experimental parameters and to check the behaviour for intermediate initial salt concentrations. The set-up involved a stable layering of the saltwater below freshwater, affected by recharge and discharge of water at the top. The situation is comparable to the upconing of a saltwater–freshwater interface due to pumping. Numerical simulations were performed with a variable-density flow code and the results compared with those from the laboratory experiments. Finally, it is illustrated how these experimental data can be used as a benchmark test for variable-density flow models.