3.3.1.2 AquiferRecharge Estimation
Generally, recharge
is a process where infiltration from the terrestrial land surface or from surface water
crosses the water table (see De Vries and Simmers (
2002) and Scanlon et al. (
2002) for a conceptual description of recharge). Recharge
can be diffuse over large areas when caused by precipitation or irrigation
. It can also be concentrated at specific locations such as tectonic fractures or preferential infiltration landforms (sinkholes in karstic systems, e.g., Andreo et al.
2008) or surface water bodies such as rivers or lakes. Recharge generally follows downward fluxes, but it can also lead to lateral fluxes in the case of interactions with neighboring aquifers or with surface water bodies such as rivers or lakes (De Vries and Simmers
2002). In the latter case, the flow direction can change depending on the season or the location, as it is mainly controlled by the water head gradient, i.e. the difference in altitude between the water table in the aquifer
and in the river or the lake (Sophocleous
2002).
There are several methods to characterize the recharge
of an aquifer
(see Scanlon et al.
2002; Healy
2010). For discussion purposes, we group recharge characterization into direct measurement (physical or chemical
), empirical and analytical or modeling methods.
Local measurements can be obtained using lysimeters. In order to obtain accurate measures of the recharge
rate the base of the lysimeter should not be deeper than the root zone. By using 28 lysimeters (0.61 m diameter, 1 m long) in the Masser Recharge
Site (central Pennsylvania, USA), Heppner et al. (
2007) report that recharge averaged 32 % of the annual rainfall (ranging from 21 % to 52 %) between 1995 and 1999. Along with Seneviratne et al. (
2012), they also discuss the main sources of uncertainty
linked to the recharge estimation using lysimeter data.
Tracers (mainly chloride and environmental
isotopes) can also be used to estimate local recharge
. These methods are based on the analysis of the tracer concentration evolution between the input (in the rain water) and the output (springs
, rivers or water table). For example, Marei et al. (
2010) estimated the spatial distribution of recharge over the western side of the Jordan Rift Valley using chloride mass-balances. Tracer methods typically require several assumptions to account for anthropogenic perturbations, variability
of climate, groundwater
-rock chemical
reactions, and difficulties inherent to output flux monitoring
.
Empirical methods are based on linear correlations fitting between climate and recharge
, and are typically calculated on an annual time scale. Although these correlations are generally specific to the climatic conditions of the locations calculated, they are relatively quick to perform. As an example, Kessler (
1967) developed a way to optimize the calculation of recharge in carbonate aquifers, assuming that “
the amount of precipitation falling in the first four months of the year (that is, preceding the development of the vegetation and prior to the large losses due to evaporation) is determinative”. In order to consider the influence of the initial climatic context, a correction factor, derived from the amount of precipitation of the last four months of the previous year, is then applied. Finally, infiltration rates are proposed in order to estimate the recharge at the monthly time scale. This method has been applied to the Hungarian mountains, and later in a southern Spanish karstic aquifer
where obtained results were realistic compared to other approaches (Andreo et al.
2008).
Modeling of aquifer
recharge
is typically most widely applied for large systems (see De Vries and Simmers (
2002) and Scanlon et al. (
2002) for extended reviews). There is a great variability
in the approaches depending on the kind of data available to describe the aquifer dynamics. Simple hydrological balance methods such as those proposed by Thornthwaite (
1948) or Dingman (
2002) can be used to estimate infiltration out of the root zone and its availability for recharge. Therefore, the focus is on water that is not intercepted by the vegetation
or consumed by evapotranspiration, nor lost as overland runoff. Typically, non-vegetated regions will have higher values than vegetated ones (Gee et al.
1994). The calculation of the distribution of infiltration (recharge) and other sinks is made using a combination of geomorphological, soil and lithology variables. There are methods that use spatially distributed information
through GIS analysis (Mardhel et al.
2004) or though external computer codes designed to calculate recharge in space and time (e.g., Westenbroek et al.
2012). Typically models
are applied at the daily time step, which can then be aggregated to longer time periods for groundwater
analysis and modeling. Models
simulating flow processes in the unsaturated zone can also be used to estimate recharge.
A variety of approaches for estimating recharge
exist, ranging from soil-water storage
-routing to numerical solutions to the Richards equation (see Scanlon et al. (
2002) for an extended review); however their results can vary substantially. Fourteen different methods applied to the same arid setting in Nevada, USA led to recharge estimates ranging from 1 to 100 mm/year (Flint et al.
2002). A comparison of different methods to estimate recharge in another arid setting in the northern Sandveld area, Western Cape, South Africa also showed variability
(Conrad et al.
2004), with estimates ranging from 0.2 % to 8 % of annual rainfall as recharge. Therefore, adequate description of how recharge was calculated for the IGM-scape is critical for acceptance by others.
3.3.1.3 Aquifer Interactions with Surface Water
Groundwater
and surface water
interaction is driven by hydraulic gradients (Gilfedder et al.
2012). The discharge
of a river is often separated into two components, a fast and short response signal to rainfall corresponding to superficial and interflow sources and a slower response corresponding to aquifer
drainage. Several techniques ranging from applying analytical methods for base flow separation to hydrographs (Gustard and Demuth
2009) to detailed hydrodynamic modeling or geochemical hydrograph separation (mainly using chloride concentration or stable isotopes of water) can be used to estimate the contribution of aquifer drainage to river discharge.
Commonly, aquifer
water levels are highly sensitive to surface-water state
, and can vary depending on the season (Allen et al.
2003). During high flows, river water typically recharges the aquifer and moves laterally away from the channel, causing groundwater
levels to rise (Scibek et al.
2007); within a relatively short period after peak discharge
, the groundwater flow direction is reversed. This is generally the case for river-aquifer interactions in natural conditions in humid climates. However, this relation can change in response to external stressors such as pumping
. In some cases, water extracted from pumping wells can be almost exclusively derived from the surface water
sources (e.g., Scibek et al.
2007). In some settings, extreme drought conditions and/or excessive pumping can lead to a complete river drying up. The potential for such adverse effects led regulators in several countries (e.g. France, Spain) to consider aquifer withdrawals close to rivers to be water withdrawals from the river itself. Even without complete drying, groundwater abstraction can affect ecological communities (Bradley et al.
2014; Chaps.
12 and
13).
In addition to well
recognized surface water
resources such as streams, rivers, and lakes, groundwater
can also play a critical role for some wetlands (Chaps.
12 and
13). Groundwater
contributes to the good ecological status of these water bodies through its effects on their physical and chemical
characteristics. In terms of the IGM-scape, this importance has been recognized at the European level, where the EU Water Framework Directive (WFD) stipulates that groundwater abstraction must not unacceptably degrade ecological status of dependent wetlands. This relation to the groundwater system can be critically important even if the inflow from the aquifer
represents a marginal part of the water supply
ing a wetland
, and water levels in aquifers can represent the main environmental
driver for wetland services (Gasca and Ross
2009).
Wetlands can also contribute significantly to the quality of the groundwater flowing through it, through soil characteristics that facilitate low oxidation-reduction conditions, filtration properties, and interaction with hydrophytic vegetation. This can be important for the retention and the recycling of some pollutants for groundwater, such as nitrates and pesticides, which can be important to consider in IGM approaches.
In the case of coastal aquifers, groundwater
level decline due to pumping
is one of the main causes of seawater intrusion, defined here as the landward subsurface incursion of seawater. Other factors such as land-use changes, climate variations or sea-level fluctuations also control the timing and magnitude of intrusion. Werner et al. (
2013) provides a comprehensive review on the diversity of the challenges associated with seawater intrusion issues. Many diverse processes can influence IGM efforts. Dynamic hydrological conditions must be assessed taking into account density-salinity relationships. Together with the slow dynamics of the processes involved, it raises significant challenges for groundwater managers charged with determining optimal groundwater use. Effective groundwater management
of coastal aquifers requires characterization of the position and thickness of the mixing zone between freshwater and intruding seawater (the seawater wedge toe) and monitoring
that combines head measurements, geophysical methods, and environmental
tracers. Simple measures such as head measurements in an observation well
can be confounded by groundwater density effects caused by salinity and fluctuations at the toe in an observation well (Shalev et al.
2009). Geophysical methods typically can detect the large electrical resistivity contrast between seawater and freshwater, allowing 1D vertical or lateral to 3D characterizations (e.g., Poulsen et al.
2010). Even simple ion analysis of coastal groundwater can document seawater intrusion occurrence. High total dissolved solids in groundwater can also be caused by rock dissolution, connate saline water and irrigation
return flow (e.g., Bouchaou et al. (
2008).