This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.
Recent advancement in European environmental legislation (Groundwater directive 2000/60/EC) requires a more integrated management approach for Groundwater Dependent Ecosystems (GDEs) which are located on interfaces between surface and groundwater. GDEs provide a number of environmental services among which are barrier function between terrestrial and aquatic ecosystems, flux control, moderation of surface water temperature, attenuation of certain pollutants by mixing or biodegradation, environment for nutrient cycling etc. (see Danielopol et al., 2004; Boulton et al., 2008; Tomlinson and Boulton, 2008). Processes such as (bio)geochemical transformation, precipitation, sorption, degradation and simple transport differ between aquifers and surface waters, thus the role of GDEs' interfaces needs to be considered. Interface's processes are difficult to assess because they depend on both micro large scales groundwater patterns (e.g. Ward et al., 1998). In this context, the use of natural tracers on various spatio-temporal scales should permit to evaluate surface water (SW) groundwater (GW) biocenoses' interactions and to propose schemes of functioning in relation to hydrological conditions and interface structures.
Interactions between SW and GW are governed by hydrological and geometrical drivers at various scales. At the reach scales (i.e. from tens of meters up to kilometers), surface systems tend to lose or gain water to/from the ground as a function of the difference between channel and adjacent water table stage (e.g. USGS, 1999). Flow-through channels receive groundwater through the upgradient bank and a portion of the channel bottom, and lose water through the downgradient bank (Figure 1).
Figure 1. Schematic representation of water exchanges between river and groundwater on the reach scale.
On a more local scale, upwelling and downwelling may be governed by discontinuities such as obstacles which protrude from the river bed (e.g. log jam), changes in the direction of flow, or pool-and-riffle sequences (Brunke and Gonser, 1997) (Figure 2).
SW-GW exchange occurs in the hyporheic or hypolentic (Winter, 2001) zones (HZ). HZ corresponds to the space below the stream/lake bed and adjacent banks where some proportion of SW and GW meet (Boulton et al., 2010). This results in a geochemically, physically and biologically mixed environment.
Figure 2. Conceptual scheme of hyporheic zone (HZ) functioning at the streambed surface (adapted from Environment Agency, 2009 and Boulton et al., 2010). Flows are controlled by streambed topography, distribution and size of sediments which influence hydraulic conductivity and extent of vertical hydraulic gradients. These patterns impact water fluxes and mass transport from and to the HZ.
Methodological approaches for HZ monitoring
Assuming geochemical and hydrological difference between SW and GW monitoring techniques for HZ investigations can be divided into hydrological, physical, chemical and physico-chemical (Table 1).
Hydrological methods such as water level measurements into aquifer and river (USGS, 1999; Kalbus et al., 2006), or differential flow gauging between upstream and downstream of a river section (Harvey and Wagner, 2000) give an idea about instant flow directions. By flow gauging and through a simple mass balance, the three type of SW-GW interactions (Figure 1) can be determined.. In addition, stream hydrographs (Smakhtin, 2001) can be processed and analyzed to characterize the magnitude and timing of groundwater discharge to streams. Baseflow separation techniques use the time-series record of stream flow to derive a baseflow hydrograph.
Physical methods such asdirect measurement of water flux across the SW-GW interface has been proposed and tested by Lee (1977) by using bag-typed seepage meters. Later this method was significantly developed by series of studies (Rosenberry and LaBaugh, 1998). Seepage technique is rather precise for both purposes but is limited by its spatial resolution which is some cm2. In combination with hydrograph separation and discharge gauging these methods permit to localize preferential downwelling of upwelling zones in SW. hydraulic conductivity evaluations through grain size analysis, permeameter tests, or pumping tests) (Fetter, 2001; Rosenberry and Pitlick, 2009 and measurements of hydraulic gradient may also permit to evaluate fluxes in HZ (Kalbus et al., 2006). Scale for this group of methods covers cm2 (riverbed texture analysis) to km2 (pumping tests) area (Hatch et al., 2006). Thermal and electromagnetic air photography covers large areas but is expensive and less detailed on a reach scale.
Medium to long term
Differential flow gauging
Ten of m
Harvey and Wagner, 2000
Channel/aquifer water level measurements
m to km
Variable(depends on length of monitoring record)
Kalbus et al., 2006
K investigations (grain size analysis, permeameter tests, or pumping tests) (used in combination with the previous techniques)
cm to km
Thermal IR, electro-magnetic remote sensing
m to km
Brodie et al, 2007
Bag-type seepage meters
cm to m
Artificial tracer test (in combination with numerical modelling)
m to km
Triska et al., 1993
δ18O; δD monitoring
Depends on number of sampling points; 1 point: cm-m
Aseltyne et al., 2006
Radon 222 monitoring (in combination with temperature )
Depend on number of sampling points; 1 point: cm-m
Hohen and Cirkpa, 2006
Dissolved Oxygen monitoring
Soulsby et al., 2009
Major ions monitoring
Kalbus et al., 2006
Metals monitoring (e.g. Fe)
Gandy et al., 2007
Soulsby et al., 2009; Vogt et al., 2010
Hatch et al., 2006
Table 1. Summary of methods allowing estimation of SW-GW interactions. Long term means decades-centuries; Medium-term means season-years scale; short-term corresponds to daily or monthly time scales.
Chemical methods are preferable when environment demonstrates sufficient difference between GW and SW. Groundwater flowpaths can be assessed by using an artificial tracer. Groundwater dynamics (velocity, dispersion) and aquifer properties (e.g. matrix and fissures porosities) may be estimated knowing distance between injection and discharge zones and monitoring restitution curves. (e.g. Malozweski and Zuber, 1985). The shortcoming is the prior knowledge of groundwater flow direction, besides those case studies where it is a research question. The above-mentioned method is punctual and may be completed by natural tracer tests (Kalbus et al., 2006) by monitoring major elements (e.g. Na+, K+, NO3- etc.) (Cook and Herczeg, 2000; Ladouche et al., 2001; Carey and Quinton, 2005), trace elements (Fe2+, Cd2+, Mn2+...etc) (Morrice et al., 1997; Hart et al., 1999), isotopic composition of water (δ18O; δD) (Coplen et al., 2000; Hinkle et al, 2001). The methods differ in their resolution, sampled volume, and time scale and usually the choice between them is a sort of trade-off with impact on the results (Vogt et al., 2010; Cox et al., 2007; Dahm et al., 2003; Cook and Herczeg, 2000; Kendall and McDonnel, 1998 and references therein).
Finally, Physico-chemical methods provide alternatoive choices for HZ investigations and are increasingly used. Indeed, they belong to easy and relatively inexpensive techniques that allow collection of continuous data series with the help of simple and reliable measurement instruments.. Regarding the GDE's topics, these environmental tracers may provide information on hydrological exchanges but may also be considered as indicators of environmental (abiotic) conditions for the living parts of ecosystems (e.g. Van der Kamp, 1995; Malcolm et al., 2009). In the following, a review of these tools and examples of their applications will be synthetized.
Temperature as indicator
Error: Reference source not foundGroundwater temperatures are relatively stable throughout the year (Younger, 2007). In contrast, river, stream or lake heat pattern heavily depends on daily and seasonal air temperature variations. Consequently the monitoring of temperature spatio-temporal patterns into HZ can be used to allocate zones of groundwater recharge or discharge (Error: Reference source not found). Schematically ,Gaining river reaches or lake shores are characterized by relatively stable temperatures (Figure 4), whereas losing reaches demonstrate vastly variable heat balance behavior (Shimada et al., 1993).
Figure 3. Conceptual scheme of the use of Temperature to assess SW-GW water exchanges (from USGS, 1999).
Figure 4: A) Loosing reach where warmer water from the river enter the GW and becomes attenuated; B) Gaining reach with colder GW streaming into the river maintaining colder SW conditions.
The thermal method is based on the principle that heat (energy) in the subsurface is transported by flowing water (advection), as well as by heat conduction via the fluid and solid parts of the soil matrix. The advective flow strongly influences the temperature distribution in the mixing zone between groundwater and surface water. Hence water fluxes between groundwater and surface water can be traced by measuring temperature distributions between the two systems (Stonestrom and Constantz, 2003; Anderson, 2005). Flux estimates are obtained by fitting solutions of the heat flow equation to observed temperature distributions in the soil.
Suzuki (1960); Stallman (1965) and Lapham (1989) describe the 1D, vertical, anisothermal flow of heat through an incompressible fluid through a homogeneous, porous media as:
where k is the thermal conductivity of the soil-water matrix in J.s-1.m-1.K-1, T the temperature at depth z in m and time t in the soil in K (C), cw the specific heat capacity of the fluid in J.kg-1K-1, _w the density of the fluid in kg.m-3, vz the vertical component of the darcian fluid velocity in the soil in m.s-1, c the specific heat capacity of the rock-fluid matrix in J.kg-1.K-1, and _the wet-bulk density (density of the rock-fluid matrix) in kg.m-3. The terms cw_w and c_ represent respectively the volumetric heat capacity of the fluid and the rock-fluid matrix in J m-3.K-1. The first term of the left hand side of equation 1 represents the conductive and the second term the advective part of heat transport. Values for cw and _w can easily be obtained from the literature. In contrast to hydraulic conductivity, the thermal conductivity k has a small range across different sediment textures (Stonestrom and Constantz, 2003).
Figure 5 presents the concept of typical fluid and heat transport between groundwater and surface water. The upper thermal boundary is usually given by the surface water temperature time series (e.g. the temperature at the interface surface water-river/lakebed), while for the lower boundary a quasi-constant groundwater temperature is assumed at a sufficient depth. Under a hydraulic steady-state (constant exchange flux between groundwater and surface water), Equation 1 can be used to simulate the shape of the vertical temperature profile at any given time (denoted by the dashed line in Figure 5). Upward flow of water (i.e. gaining conditions, negative sign of vz) is referred to as groundwater discharge, whereas groundwater recharge is defined as downward movement of surface water (i.e. a losing condition, positive sign of vz).
Figure 5 : Concept of transient simulation of vertical 1D-fluid-heat transport in a river or lake bed; part a indicates the continuously measured surface water temperature used as upper boundary condition; part b indicates the soil column and its changing temperature distribution represented in the numerical models (in Anibas et al., 2009)
Constantz and Stonestrom, 2003; Silliman and Booth, 1993). Conant (2004) as well as Schmidt et al. (2007) obtained spatial patterns of groundwater exfiltration using temperature and piezometer data from a dense monitoring network. They analyzed snapshots of vertical temperature profiles with a model assuming steady-state heat transfer. Anibas et al. (2009), in a river located in Belgium, by comparing the obtained fluxes between groundwater and surface water with fluxes inverted from fitting observed vertical temperature profiles to an analytical solution for 1D steady-state heat flow, showed good agreement of fluxes during and towards the end of the summer and winter. Underestimation and overestimation of transient fluxes by the steady-state solution occurs in fall and spring respectively. It is concluded that for a temperate climate like in Western and Central Europe during certain periods of the year, namely from mid-July, August up to mid-September in the summer season and January, February up to mid- March in the winter season, exchange fluxes between groundwater and surface water can be inverted from a steady-state heat flow model. In other words, during those times the assumption that vertical temperature distributions at the interface between groundwater and surface water are at a quasi-steady-state, is acceptable. Under certain meteorological conditions it might be possible to also get accurate flux estimates outside of those periods. The advantage of this approach is that if the temperature distribution is at a steady-state, only the temperatures at the upper and lower boundaries, as well as the thermal conductivity of the porous medium need to be known to determine the exchange flux or velocity (vz in equation 1), reducing the necessary input data significantly. As this solution assumes that the temperature distribution in the subsurface is not changing over time, no continuous input data is necessary to describe the boundary conditions. A temperature profile consisting of a few measurement points with depth and the temperatures at the upper and lower boundaries are sufficient to verify the fit between steady-state assumption with actual processes.
However, the applicability of the steady-state model is questionable because river temperatures typically exhibit a diurnal cycle so that the streambed temperature is transient. Under infiltrating conditions, the steady-state contribution to temperature profiles is theorically not appliable for the quantification of seepage rates. The amplitude of diurnal temperature fluctuations must be significantly higher than the accuracy of the temperature sensor. If the seepage flux is constant, retrieval of the diurnal signal is simple. In highly dynamic cases, separating the diurnal contribution subject to time-varying auto-correlated parameters from the trend contribution, which is also assumed to be auto-correlated, may become delicate.
Because diurnal temperature variations are periodic and more or less sinusoidal, Stallman's (1965) analytical expression of the convection-conduction equation for sinusoidal boundary conditions is a good starting point for the analysis of temperature data. Hatch et al. (2006) applied a band-pass filter to pairs of temperature time series obtained at different depths within the sediment. Mean flow rates were estimated from the amplitude and phase angle determined for each day using a semi-automated computer program. Kerry et al. (2007) extracted sinusoidal components with periods of one day from temperature time series at various depths using Dynamic Harmonic Regression (Young et al., 1999). In this approach, the amplitude and phase angle were estimated as continuous, auto-correlated time variables. Following the analysis of amplitude and phase-angle information of Stallman (1965), Kerry et al. (2007) obtained the seepage rate as a continuous time variable rather than as a set of daily values. Details for the application of this methods can be found in Young et al. (1999) ; Kerry et al. (2007), and Vogt et al. (2010).
As this approach requires time series of temperature evolution at various depth in the bed, the recent technological innovations such as continuous logging sensors (Hoehn and Cirpka, 2006; Vogt et al., 2010) for in situ monitoring over prolonged periods and allow to obtain a detailed view of the variation of river bed temperature over depth and time. Their incorporation into 3D numerical models (Poole, 2010) will probably permit a better understanding of groundwater-surface water interactions in a near future.
Electrical conductivity as indicator
Electrical conductivity (EC) is another option of environmental tracing techniques in the field of SW-GW interplay and basically can be used in the same way as temperature. Fluctuations of EC results from variation of total dissolved solids (Appelo and Postma, 2007). In aquatic ecosystems EC variability may be caused by carbon turnover by aquatic biota on seasonal scale, and such factors as photosynthesis and respiration on diurnal basis (Odum, 1956). Rainfall events and snowmelt introduce a dilution. These changes also affect inorganic carbon cycle equilibrium, provoking phase changes (either precipitation or dissolution) of calcium and magnesium carbonates. Assuming a constant EC in the subsurface waters SW-GW interactions may be evaluated through deconvolutions of the chemical variability at the GDE's interface. As it has been observed by Hatch et al. (2006) daily EC variations are predominantly distinct at low river water level and high temperatures. Diurnal phosyntetical processes responsible for CO2 fluctuations, keep EC highest in the early morning, and lowest in the afternoon. Seasonal variations usually introduce discrepancies between the river and groundwater data sets and can be mathematically removed. Electrical conductivity has the advantage and propagates quicker through the media than temperature concurrently showing less smooth trend (Cirpka et al., 2007).
Combined approach and description of the test site
Simultaneous application of EC and temperature. improves assessment of SW-GW interaction (Cox et al., 2006). Thermal patterns of the HZ can be used to derive hydraulic conductivities. Vogt et al. (2010) have as well noticed that combined information is usually more characteristic since EC doesn't undergo retardation while propagating through the HZ. More stable signal from EC varies on a diurnal basis due to variations in bicarbonate and hardness, on several days basis as a result of extended precipitation, and on seasonal basis reflecting winter base flow conditions dominated by the groundwater flow (in majority of the cases). Different infiltration regimes can be registered by analyzing diurnal and seasonal EC patterns.
Temperature appears a very useful indicator in hyporheic exchange when high frequency components can be registered, e.g. diurnal temperature fluctuations of loosing stream waters entering the HZ (Hoehn and Cirpka, 2006; Vogt et al., 2010). The low frequency signals, i.e. seasonal temperature variations, are characterized by slower traveling time and indicate mixing of groundwater components of different age. This type of signatures usually cannot be employed in SW-GW exchange dynamics. On the other hand temperature variations on a few days basis are very useful in estimating water residence time and hyporheic exchange due to their uniqueness of frequency. These signals are difficult to mismatch in their peaks (Hoehn and Cirpka, 2006).
Temperature tied with EC measurements is not the only coupling approach, but can be applied with other tracer techniques like 222Rn or sulphate to trace origin and age of the young groundwater components.
It has been observed that in some site specific conditions seasonal and/or diurnal variations of one of the parameters can be negligible and therefore will deliver no understanding on exchange between surface and groundwater. Laroque et al. (1998) speculated that temperature alone is not a conservative tracer and in such complex environment as karst aquifer is not advantageous as stand-alone technique. Thermal exchange of the groundwater with the surrounding bedrock is of a very high importance in the karstic environment. Thus, combination of different techniques is advantageous. (Laroque et al., 1998; Hoehn and Cirpka, 2006; Cirpka et al., 2007; Vogt et al., 2010).
Hyporheic exchange is also studied within Genesis project. The test site is located in the vicinity of the regulated Luleå River in northern Sweden. Hyporheos appear under threat and is subject to diurnal and seasonal river level fluctuations. The test site has been equipped with series of groundwater wells placed orthogonal to the river on different distance. Continuous monitoring of temperature, water levels and conductivity was established. Measurements were settled in July 2010 and are still ongoing. Temperature and conductivity were registered by combined four-electrode measuring cell. TetraCon 325 has a temperature response t99 < 20s (data from the producer), resolution and accuracy of the temperature sensor ± 0.2 oC. Conductivity is measured with resolution of ±0.1%, and accuracy of ±0.5% in the range 1µS/cm - 2S/cm. Collected time series represent data with 1 hour temporal resolution. Water exchange between the river and the HZ is complicated by rather clogged river bed which is the result of extended low flow conditions during the observational period. We expect increased erosion to occur during the winter and the spring snowmelt when hydropower release is higher (Hoehn and Cirpka, 2006). Nevertheless, there is a very tight connection between the aquifer and the river. Hydraulic heads deviate simultaneously in the river and the nearest the shore well. A bit more complicated pattern occurs while investigating EC and temperature time series. Raw data doesn't deliver so much information and therefore time series analysis is required. Data sets incorporate fault measurements representing maintenance periods, fouling of sensors, times when they are occasionally buried into the sediments, electrical failures, drifts and minor deviations. All these events are not desirable in the time series. Trivial plotting, data transformation and general statistics have to be applied in order to receive preprocessing data. In highly frequent water level variation conditions diurnal temperature and EC are of the highest importance. We would like to exclude seasonal trends resulting from the diverse contribution of weather dependent factors. For this purpose least-square fitting can be employed to help to identify the seasonal drift and spectral filtering to separate it (Jenkin and Watts, 1968; Hipel and McLeod, 1994; Laroque et al., 1998). Cross correlation analysis of the time series between the river signals of temperature and EC and those from the groundwater wells is supposed to reflect the strongest responses (Hohen and Cirpka, 2006). One of the effective ways to detect smaller peaks of the same origin both in the river and the HZ is to apply non-parametric deconvolution. By removing daily component using equally weighted moving average (Laroque et al., 1998) different infiltration regimes can be identified (Vogt et al., 2010) (Figure 6).
Temperature and EC of the Luleå River will be affected by the weather related factors as well as water releases from hydropower functioning. As the next step it will be important to separate these effects and extract the signal caused by hydropower regulation.
Figure 6: Scheme of the planned methodological approach for the study of SW-GW exchanges in the Luleå river.