Groundwater Solute Concentrations Research Articles

Tracing biogeochemical processes in stream water and groundwater using non-linear statistics

Summary Stream water and groundwater solute concentration are subject to a multitude of biogeochemical processes that act at different scales and are often characterized by non-linear relationships and feedback loops. Different multivariate statistical methods were applied to investigate the interplay of different processes. The data set from the Lehstenbach catchment in South Germany comprised 2641 stream water and groundwater samples from 38 different sites in the catchment, where 13 different solutes had been determined. According to the correlation dimension analysis, the number of dominant processes was four. The first four components determined via principal component analysis comprised 88% of the total variance, whereas the non-linear isometric feature mapping explained 92% with the first four components. These components were ascribed to prevailing biogeochemical processes and were used to investigate spatial and temporal patterns. Redox processes and contamination by road salt explained 35% of the variance each. Another 13% were ascribed to near-surface runoff in the acidified topsoil, and 9% to the impact of contaminated filter gravel in some of the groundwater wells. The redox component exhibited clear seasonal patterns at most stream water and groundwater sampling sites, with the most reduced conditions in late summer, immediately before the onset of re-wetting. There was clear evidence that redox processes, especially denitrification, play an important role even in the oxic aquifer. During discharge peaks, stream water exhibited higher values of the near-surface runoff component. However, the associated lower values of the redox component pointed to near-surface runoff in the riparian wetlands as the predominating runoff generation process rather than to a contribution of upslope soil water. A series of major rain storms in fall 1998 altered groundwater and stream water solute concentration for months: stream water and groundwater became more oxic and more acidified, and showed a higher impact of road salt contamination at some sites. Overall, the spatial and temporal patterns of the predominating components gave a consistent picture and helped considerably to better understand the interplay between biogeochemical and hydrological processes. The chosen approach is a promising tool for investigating apparently complex hydrological and biogeochemical systems.

A comprehensive study of nutrient losses, soil properties and groundwater concentrations in a degraded peatland used as an intensive meadow – Implications for re-wetting

Summary Most of the once ecologically valuable and widespread peatlands in Western Europe as well as in Mecklenburg-Vorpommern (North-Eastern Germany) have been drained in the course of the intensification of agricultural land use. Peat degradation and mineralisation frequently caused high nitrate–nitrogen ( NO 3 - – N ) losses, while re-wetting may lead to elevated phosphorus concentrations in the porewater. High concentrations of NO 3 - – N and other solutes in a ditch draining a catchment (85 ha) dominated by degraded peatland under intensive grassland use gave reason to investigate the relationship between the spatial and temporal variability of the shallow groundwater quality on the one hand and soil properties, topography and hydrological dynamics on the other hand. Therefore, in addition to the ditch, three transects of dipwells were sampled and soil samples were taken along the transects at a high spatial resolution. Soil organic carbon Corg and total nitrogen Nt contents varied considerably from 0.8% to 40.9% and 0.08% to 2.87%, respectively. A trend surface analysis (TSA) of these soil properties showed a strong trend depending on the ground elevation and the distance to the adjacent drainage ditch, reflecting both a transition from mineral to organic soils and differences in peat formation and degradation connected to the topography and, consequently, to the depth to the groundwater table. Semivariogram analysis was performed using the normally distributed residuals of the TSA and showed for each transect a strong spatial dependence and short (5–26 m) ranges. If possible, soil sampling in peatlands should be conducted at short distance sampling intervals. The patterns of groundwater solute concentrations were complex. They were spatially and temporally very variable between as well as within the transects, thus proving that a few groundwater samples are insufficient for a representative characterisation of the peatland status and for an evaluation of possible environmental impacts or potential consequences of re-wetting. Groundwater concentrations ranged from 0 to 65.4 mg l−1 NO 3 - – N and 0–1.95 mg l−1 Pt, while during the same period, ditch water concentrations of 0–15.9 mg l−1 NO 3 - – N and 0.05 to 0.44 mg l−1 Pt, respectively, were measured. Factor analysis of groundwater concentrations, topographical data, hydrological conditions and soil properties led to the identification of five factors explaining up to 80% of the observed variance of the data: (i) topography and soil chemical properties, (ii) biogeochemical processes, (iii) drainage effects, (iv) climatic effects and (v) dilution effects.

Hydrogeophysical tracking of three‐dimensional tracer migration: The concept and application of apparent petrophysical relations

Direct estimation of groundwater solute concentrations from geophysical tomograms has been only moderately successful because (1) reconstructed tomograms are often highly uncertain and subject to inversion artifacts, (2) the range of subsurface conditions represented in data sets is incomplete because of the paucity of colocated well or core data and aquifer heterogeneity, and (3) geophysical methods exhibit spatially variable sensitivity. We show that electrical resistivity tomography (ERT) can be used to estimate groundwater solute concentrations if a relation between concentration and inverted resistivity is used to deal quantitatively with these issues. We use numerical simulation of solute transport and electrical current flow to develop these relations, which we call “apparent” petrophysical relations. They provide the connection between concentration, or local resistivity, and inverted resistivity, which is measured at the field scale based on ERT for media containing ionic solute. The apparent petrophysical relations are applied to tomograms of electrical resistivity obtained from field measurements of resistance from cross‐well ERT to create maps of tracer concentration. On the basis of synthetic and field cases we demonstrate that tracer mass and concentration estimates obtained using these apparent petrophysical relations are far better than those obtained using direct application of Archie's law applied to three‐dimensional tomograms from ERT, which gives severe underestimates.

Chemistry and late Quaternary evolution of ground and surface waters in the area of Yabulai Mountains, western Inner Mongolia, China

In the extensive area of western Inner Mongolia, China, the water demand of local residents often depends mainly on shallow aquifers, although scientific investigations of the quality and formation of the groundwater are still lacking. In this study the chemistry and isotopic composition of groundwater and lake water samples collected at 22 sites in the area of Yabulai Mountains ( Fig. 1) in western Inner Mongolia were analysed. Chemical water analysis included the determination of Na +, K +, Ca 2+, Mg 2+, Li +, NH 4 +, Al 3+, Co 2+, CO 3 2−, HCO 3 −, Cl −, SO 4 2−, NO 3 −, F −, PO 4 3−, HBO 2 and SiO 2, and 18O and tritium isotopes. Solute concentrations in groundwater show significant differences within the study area, while the water of the three lakes is dominated by sodium and chloride. This study reveals the importance of spatial variation of groundwater chemistry in an arid environment that has a relatively homogeneous lithological basement. The heterogeneity of groundwater chemistry suggests that the shallow groundwater in western Inner Mongolia is mainly recharged by infiltration of local rainfall. The relatively high tritium content indicates that the water in the shallow aquifer of the study area is generally not older than 100 years. Former shorelines in the lake basin of Shugui ( Fig. 1) suggest the presence of a much larger lake in the past under a wetter climate than present hyper-arid conditions, presumably because of increased East Asian monsoon intensity during the middle Holocene. In the vicinity of commune administrations that have higher population density, the TDS of ground water is also higher, probably owing to water pollution.

On the Nonlocality of Reversible-Time Particle Tracking Methods

This article presents a novel methodology for time reversal in groundwater transport simulations. A variance minimization procedure leads to a stochastic method, analogous to the Random Walk Particle Tracking Method (RWPTM), capable of tracking groundwater solute concentration profiles over time. This method, called the Reversible-Time Particle Tracking Method (RTPTM), can backtrack in time a certain solute plume, of arbitrary shape, and identify the most probable spill location. The correspondence to the "classical" RWPTM is established for forward-time simulations, and the method is applied to one-dimensional transport of both conservative and nonconservative solutes in spatially invariant and variant groundwater flow fields with great success.

Validation of the dual pumping technique for level-determined groundwater sampling in a contaminated aquifer

The dual pumping technique (DPT) was introduced recently as a new and inexpensive technique to measure the level-determined solute concentrations in groundwater. Using two pumps, one placed near the groundwater table and one placed near the bottom of a fully screened well, this technique allows to determine vertical concentration profiles for solutes in groundwater given that additional information about the influx distribution into the well is known. Until now, however, the DPT was applied only in an aquifer with a thickness of about 40 m and the validation was weakened by the lack of a reliable reference system. The present study aims to investigate the applicability of the DPT for shallow (thickness <10 m) and unconfined aquifers and to validate the results of the DPT with a better reference system. For this purpose the DPT was applied in Menziken, Switzerland to a gravel aquifer contaminated with petroleum hydrocarbons. Solute concentrations measured at that site with an established drive point sampling technique (Ram technique) were taken as a reference to demonstrate the applicability of the DPT. Results for eight different solutes showed a reasonable agreement between both techniques. An algorithm was developed that allowed the computation of a single solute concentration profile incorporating measured data from both pumps. It was possible to demonstrate that this alternative algorithm can improve the quality of the solute concentration profiles obtained by the DPT. This indicates that the DPT is a useful tool for determining the vertical concentration profiles in groundwater for conditions similar to those in Menziken (unconfined gravel aquifer, screen length less than 10 m).

A Phenomenological Model of the Increase in Solute Concentration in Ground Water Due to Evaporation

We present a new phenomenological model of the evaporation of ground water containing a polluting material in the dissolved form. Only the one-dimensional case is treated. It is assumed that there exists a sharp evaporation front separating between the dry and the water-saturated soil. The water-saturated soil is assumed to occupy a semi-infinite domain x > X(t), where x is a vertical coordinate directed downward and X(t) is a position of the evaporation front. The mathematical description is based on four linear diffusion equations coupled through four boundary conditions, one of which is nonlinear, on the free moving evaporation front. We use a similarity solution of the governing equations and analyze it qualitatively showing that the solute concentration increases in the upward vertical direction and reaches its maximum on the evaporation front. The dependence of the solute concentration at the evaporation front and of the velocity of the front on the initial solute concentration and the temperature of the ground surface are computed. It is shown that for not high values of the initial solute concentration that are below the concentration value cd at which a deposition of the pollutant sets in, the solute concentration on the evaporation front can reach values that are above the deposition value cd. These results point to a possible mechanism of pollutant deposition in ground water caused by the evaporation.

Modelling of nitrate leaching from arable land into unsaturated soil and chalk 2. Model confirmation and application to agricultural and sewage sludge management

A layered deterministic N-leaching model, IMPACT, has been calibrated using data from two study sites on the unconfined Chalk aquilfer of East Anglia, UK. The model predicts nitrogen species movement resulting from the application of sewage sludges and fertilizers to arable land for different vegetation-soil-hydrogeological conditions. One site received sludge in the form of digested sewage cake (DSC) for the first time during the study period, whilst the other site had over 15 years history of liquid undigested sludge (LUS) applications at 3 year intervals. Site data included: 3-monthly concentration profiles at 0.3 m intervals to depths of up to 6 m for N-species and chloride; unsaturated potential measurements; water level and saturated groundwater solute concentrations, fertilizer and sludge input; daily recharge, and soil/chalk type and moisture content. The observed average movement rate for nitrate peaks in the Lower Chalk, measured at one site, was 0.2 m year −2. Leachate peaks were not observed annually but approximately every third year, being associated with large sludge applications and ploughing of grass crops. Significant correlation between observed and modelled nitrate profiles in soil and chalk were obtained which demonstrated applications. The relationship between crop demand, application times of fertilizers and sludge, nitrate availability and recharge was shown strongly to control the shape of nitrate profiles in the soil and chalk and the quantity of nitrate leached tochalk. The change in hydrogeological conditions at the soil-chalk contact and associated potential for denitrification was also shown to exert a significant control on the shape of the nitrate profile. Following calibration, different arable crop and sludge application regimes were examined for a 6 year period and ranked according to their nitrate leaching risk. Of the modelled cereal farming scenarios, the crop/sludge regime giving the least nitrate leaching was a late autumn surface spread application of DSC followed by winter cereals, while highest nitrate leaching was generated by an autumn injection of LUS followed by spring cereals. Field and modelled results may be used in the development of sludge disposal policies to arable land particularly with regard to sludge types, application times, and following crop types and fertilizer requirements. Overall, observed and model data demonstrate the importance of examining nitrate leaching as a continuum from the soil through the chalk to the water table.

Eolian transport, saline lake basins, and groundwater solutes

Eolian processes associated with saline lakes are shown to be important in determining solute concentration in groundwater in arid and semiarid areas. Steady state mass balance analyses of chloride in the groundwater at Double Lakes, a saline lake basin in the southern High Plains of Texas, United States, suggest that approximately 4.5 × 105 kg of chloride is removed from the relatively small (4.7 km2) basin floor each year by deflation. This mass enters the groundwater down the wind gradient from the lake, degrading the water quality. The estimates of mass transport were independently determined by evaluation of solutes in the unsaturated zone and by solute mass balance calculations of groundwater flux. Transport of salts from the lake was confirmed over a short term (2 years) by strategically placed dust collectors. Results consistent with those at Double Lake were obtained from dune surfaces collected upwind and downwind from a sabkha near the city of Abu Dhabi in the United Arab Emirates. The eolian transport process provides an explanation of the degraded groundwater quality associated with the 30–40 saline lake basins on the southern half of the southern High Plains of Texas and New Mexico and in many other arid and semiarid areas.

Spatial variability of groundwater solute concentrations at the water table under arable land and coniferous forest. Part 3: Field data for a coniferous forest and statistical analysis

AbstractThe spatial variability of groundwater solute concentrations at the water table beneath a fine to medium sandy soil (Gleyic Podzol) under a 45‐years old pine stand was investigated. Samples were taken within the first decimeter below the groundwater table at horizontal intervals of 0.5 m along a 30 m transect in between two rows of trees. Nitrate, sulfate, and aluminum concentrations and pH were determined and analyzed by geostatistical and spectral analysis methods.Semivariance analysis shows spatial dependence in the data (except for nitrate) up to ranges of between 1.7 and 3 m. Spectral analysis indicates periodic variations with a period of 4 ‐ 5 m. The periods for sulfate and aluminum are synchronous and 180° out‐of‐phase with the pH curve. The sulfate and aluminum maxima and pH minima correspond to areas within edges of the tree canopies (canopy drip). Differences in the acid input into the weakly buffering soil caused by the interception deposition of atmospheric constituents at the canopy result in a regular pattern in the concentrations of certain solutes in the uppermost groundwater. The variogram and spectrogram of the nitrate concentration indicate the presence of a trend. The cospectrogram for nitrate and pH has a peak that corresponds to synchronous short‐range (about 2 m) variations, i.e. a small‐scale correlation between nitrate concentration and pH.

Spatial variability of groundwater solute concentrations at the water table under arable land and coniferous forest. Part 2: Field data for arable land and statistical analysis

AbstractThe spatial variability of groundwater solute concentrations at the water table beneath a fine to medium sandy soil (Gleyic Podzol) under arable land was investigated. Samples were taken in March (within a period of high seepage) within the first decimeter below the groundwater table at horizontal intervals of 0.5 m along a 30 m transect perpendicular to the direction of plowing. At the time of sampling the field was fallow, the preceding crop was maize. Nitrate, sulfate, potassium, and calcium concentrations, as well as electrical conductivity were determined and analyzed by geostatistical and spectral analysis methods.Semivariance analysis shows spatial dependence between measurements (except for potassium) within a range of about 6 m. Spectral and cospectral analysis show a pattern of periodic variations with a period of 12 m for nitrate, sulfate, calcium, and electrical conductivity. This pattern is related to the manner in which the mineral fertilizer is spread (width of a single application = 12 m). The periodic variation of sulfate is 180° out‐of‐phase and has a phase shift of 0.4 m with respect to nitrate, calcium, and electrical conductivity, which is probably due to separate application. The concentration of potassium shows a spatial trend, presumably due to a trend in the amount of potassium fertilizer applicated from one side of the field to the other.

Spatial variability of groundwater solute concentrations at the water table under arable land and coniferous forest. Part 1: Methods for quantifying spatial variability (geostatistics, time series analysis, Fourier transform smoothing)

AbstractThe solute load from soil into the groundwater plays an important role in considerations about the causes of groundwater pollution. Solute concentrations in the uppermost groundwater can be used to characterize the solute input into the groundwater in a particular area. Such interpretation, however, depends on a knowledge of the spatial variability and patterns of these concentrations.Geostatistical methods (semivariance and autocorrelation), spectral and cospectral analysis, and Fourier transform smoothing of data were used to quantify spatial relationships of the solute concentrations determined along 30 m transects. Electrical conductivity and calcium concentration data of the uppermost groundwater in a sandy soil under arable land were used for testing the procedures mentioned above. The range of the semi‐variogram was found to be 6 m. Spectral analysis and Fourier transform smoothing yielded a periodical variation of 12 m length for this data.

Aqueous geochemistry and diagenesis in the eastern Snake River Plain aquifer system, Idaho

Water budget and isotopic analyses of water in the eastern Snake River Plain aquifer system confirm that most, if not all, of the water is local meteoric in origin. Solute mass-balance arguments suggest that ∼5 × 10 9 moles of calcite and 2.6 × 10 9 moles of silica are precipitated annually in the aquifer. Isotopic evaluations of calcite and petrographic observation of silica support the low-temperature origin of these deposits. Approximately 2.8 × 10 9 moles of chloride, 4.5 × 10 9 moles of sodium, 1.4 × 10 9 moles of sulfate, and 2 × 10 9 moles of magnesium are removed annually from the aquifer framework by solution. Proposed weathering reactions are shown to be consistent with mass balance, carbon isotopes, observed mineralogy, and chemical thermodynamics. Large quantities of sodium, chloride, and sulfate are being removed from the system relative to their abundances in the rock. Sedimentary interbeds, which are estimated to compose 2 )/yr or less than half the average of the North American continent. This contrasts with the rate for the eastern Snake River basin, 34 (Mg/km 2 )/yr, which is almost identical to the average for the North American continent. Identification and quantification of reactions controlling solute concentrations in ground water in the eastern plain indicate that the aquifer is not an “inert bathtub” that simply stores and transmits water and solutes but is undergoing active diagenesis and is both a source and sink for solutes.

Management model of a groundwater system with a transient pollutant source

A management model is presented for determining the maximum concentration of a transient pollutant source such that space‐dependent groundwater quality requirements are met. The method is illustrated using a one‐dimensional groundwater system with a single pollutant source at an adjacent stream. The chloride concentration of the source is to be managed to meet variable groundwater quality constraints. Management of the pollutant source is demonstrated for a single‐period as well as for repeated‐period pollutant discharges. The Crank‐Nicolson numerical approximation of the convective‐dispersive equation is used in the management model. The pollutant source concentration is treated as a parameter in the resulting system of linear equations. Taking advantage of the block structure of the matrix, the concentrations throughout the system are defined and manipulated as functions of this parameter for each time step. The parameter is maximized by comparison of groundwater quality limits with the groundwater solute concentrations at the corresponding nodes. The minimum value of the parameter over the entire time frame is the maximum concentration allowable in the source water over the management period or periods.