Groundwater最新文献

Groundwater level observations are used as decision variables for aquifer management, often in conjunction with models to provide predictions for operational forecasting. In this study, we compare different model classes for this task: a spatially explicit 3D groundwater flow model (MODFLOW), an eigenmodel, a transfer-function model, and three machine learning models, namely, multi-layer perceptron models, long short-term memory models, and random forest models. The models differ widely in their complexity, input requirements, calibration effort, and run-times. They are tested on four groundwater level time series from the Wairau Aquifer in New Zealand to investigate the potential of the data-driven approaches to outperform the MODFLOW model in predicting individual target wells. Further, we wish to reveal whether the MODFLOW model has advantages in predicting all four wells simultaneously because it can use the available information in a physics-based, integrated manner, or whether structural limitations spoil this effect. Our results demonstrate that data-driven models with low input requirements and short run-times are competitive candidates for local groundwater level predictions even for system states that lie outside the calibration data range. There is no “single best” model that performs best in all cases, which motivates ensemble forecasting with different model classes using Bayesian model averaging. The obtained Bayesian model weights clearly favor MODFLOW when targeting all wells simultaneously, even though the competing approaches had the chance to fine-tune for each tested well individually. This is a remarkable result that strengthens the argument for physics-based approaches even for seemingly “simple” groundwater level prediction tasks.

The study of hydraulic head changes over time is a common task for groundwater hydrologists. Groundwater signatures are numerical metrics, or statistical aggregates, that quantify the behavior observed in hydraulic head hydrographs. Signatures can be helpful in a number of classical hydrological tasks, such as hydrograph classification, clustering, change detection, and model evaluation, selection, and calibration. Despite the potential benefits of using signatures in groundwater studies, their application has not yet been thoroughly explored. To support research into the application of signatures in groundwater studies, we introduce the new groundwater signatures module from the Pastas software. The signatures module is written in Python, fully tested and documented, and available as open-source software under the MIT license. In this paper, it is shown how the signatures are tested and can be used in practical applications through two examples. In the first example, signatures are used to characterize and cluster monitoring wells in a nationwide monitoring network in Switzerland. In the second example, signatures are used to evaluate how well different groundwater model structures simulate the heads. Future research opportunities involving groundwater signatures are discussed.

Surface water (SW) and groundwater (GW) models, such as MODFLOW and HEC-RAS, have been explored to simulate the complexities of SW–GW interactions. However, individual models are not capable of capturing the full complexity of these interactions. To overcome individual models' shortcomings, researchers introduced the model coupling concept. This concept helps compensate for each individual model's shortcomings and incorporates the models' advantages. However, challenges arise from temporal scale disparities between SW and GW models. To tackle the temporal scale issue, this study introduces the novel explicit solver (EXP1) for MODFLOW 2005, enabling GW modeling using small time steps matching SW models (i.e., 15 min) by reducing runtime and computational burden. The EXP1 solver incorporates an integrated stability criterion to ensure the stability of explicit schemes, and it was systematically evaluated against the Preconditioned Conjugate Gradient (PCG) solver across various scenarios, including a 1-dimensional, 2-dimensional, and a vast 3-dimensional model. Results demonstrated the efficiency and accuracy of EXP1 in predicting groundwater heads and water budget, along with considerably reduced runtimes of up to 33% compared with the PCG solver, with less than 0.4% discrepancy in the water budget. These findings underscore the effectiveness of EXP1 in facilitating groundwater small time step simulations and bridging the temporal scale gap between SW and GW models.

Groundwater age distributions provide fundamental insights on coupled water and biogeochemical processes in mountain watersheds. Field-based studies have found mixtures of young and old-aged groundwater in mountain catchments underlain by bedrock; yet, the processes that dictate these groundwater age distributions are poorly understood. In this work, we use the coupled ParFlow-CLM integrated hydrologic and EcoSLIM particle tracking models to simulate groundwater age distributions on a lower montane hillslope in the East River Watershed, Colorado (USA). We develop a convolution-based approach to propagate fracture-matrix diffusion processes to the EcoSLIM advection-dominated age distributions. We compare observed ³H and ⁴He concentrations from two groundwater wells against model predictions that have varying advective transport times and matrix diffusion magnitudes. Based on a Monte Carlo analysis that considers uncertain matrix and fracture parameters, we find that matrix diffusion is needed to jointly predict ³H and ⁴He observations at both wells. The advection-dominated age distributions lack adequate mixing of young and old-aged water to capture the observed co-occurrence of ³H and ⁴He. The model scenario that best matches the ³H, ⁴He, and water level observations when considering both advective flowpath and matrix diffusion mixing processes has a dynamic bedrock groundwater reservoir that is susceptible to considerable storage losses during low-snow periods. This dynamic groundwater system amplifies the need to assimilate deeper bedrock groundwater into watershed hydro-biogeochemical predictions. This work further highlights the importance of considering matrix diffusion when interpreting environmental tracers in bedrock groundwater systems.

Aerobic bioremediation enhanced by tandem circulation well (TCW)-generated aeration in a groundwater circulation systems has emerged as a novel, environmentally friendly, and cost-effective remediation approach with growing recognition. For TCW, previous investigations have been limited to few laboratory experiments, simulation precision, acquisition of reaction kinetic parameters, and effective guidance for technology optimization. In this work, we employed regionalized sensitivity analysis with Latin Hypercube Sampling (LHS) to identify the most sensitive parameters in laboratory TCW experiments, reducing the number of parameters to estimate. The estimated parameters were utilized to construct a reactive transport model with periodic boundary conditions, enhancing its universality for in-situ trichloroethylene (TCE) bioremediation through electrolysis considering mutual interactions among well clusters. The results revealed the influence mechanisms of the operating parameters and well spacing on remediation performance. Besides, it was found that degradation efficiency was limited by DO oversaturation in the wellbore. However, it could be promoted by optimization of operation parameters, using an optimization index, the ratio of current to pumping rate ($��\alpha �$). Finally, simulation results implied two suggestions for well spacing: (1) Designing a remediation site with a higher aspect ratio will enhance the performance of this technology. (2) With a larger area, both current intensity and pumping rate need to be proportionally increased in alignment with the enlarged area to ensure optimal efficiency. This work improves the precision of characterizing the TCW system, guiding the determination of reaction kinetics parameters and optimization of critical design parameters, including operational parameters and well spacing, in remediation sites, thereby achieving superior remediation performance in field applications.

Recent modeling of groundwater at the scale of the Conterminous United States (CONUS) has often relied on relatively large square grid cells (Maxwell et al. 2015; Zell and Sanford 2020). Consequently, features such as streams can become generalized. An important issue, therefore, is the relationship between grid-cell size and representation of the stream network. This technical commentary addresses two questions related to this issue. First, what cell size is required to accurately represent all mapped streams in CONUS (McKay et al. 2012)? Second, given a 1-km cell, what order stream network can be accurately represented? This commentary focuses only on geometry and does not address other important aspects of modeling stream –aquifer interactions such as small-scale sinuosity (Cardenas 2009) or developing sound conceptual hydrogeologic models (Anderson et al. 2015).

The overall approach for assessing accuracy to answer these questions is presented in the methods section, including how the relationship between cell size and stream order is evaluated. The approach is applied to 18 representative surface water basins distributed across CONUS (Van Metre et al. 2020). The Delaware River Basin is used to illustrate the approach and for the purposes of discussion.

A uniform-square grid that accurately represents the geometry of a stream network is one in which contributing areas can be differentiated from streams, such that most of the cells do not contain a stream. A grid where every cell contains one or more streams would not accurately represent the geometry of the network. For the purposes of discussion, the criterion for accurate representation is that streams are in no more than 20% of all cells. A different criterion could be chosen depending on the purpose of a particular study. The graphs presented in this commentary allow for a different criterion.

The question of developing an accurate grid also depends on stream order. Representation of the geometry of a first-order stream network (headwaters) requires a smaller cell size than the representation of higher-order stream networks (large rivers). In this paper, we evaluate stream networks ranging from first through fifth stream orders across the CONUS.

The relationship between grid-cell size and stream network geometry is evaluated in 18 river basins distributed across CONUS (Figure 1; Table 1). These 18 river basins were identified as candidates for future intensive monitoring and assessment by the USGS (Van Metre et al. 2020) per a ranking scheme that represents the range of physiographic, climatic, and land use characteristics within CONUS. For each river basin, we used the Multi-Order Hydrologic Position (MOHP) dataset (Belitz et al. 2019; Moore et al. 2019) derived from the NHDPlusV2 stream network (McKay et al. 2012) as the basis for i

The separation of advection and dispersion from the breakthrough curve of a potentially reactive solute can help determine if reactive transport mechanisms occurred. This is typically done by solving the advection–dispersion equation and fitting the breakthrough curve of an applied non-reactive solute tracer by adjusting groundwater velocity and the dispersion coefficient; the values of velocity and dispersion are then applied to the breakthrough curve of the potentially reactive solute, and any residuals can be fitted with the appropriate reactive transport mechanisms. A simpler approach is to plot the dimensionless relative concentrations of the non-reactive and reactive solutes on the same breakthrough curves; thus, any differences between the two curves can be attributed to reactive transport. The method proposed here can allow for separating advection and dispersion from the breakthrough curve of a potentially reactive solute based on data only, as opposed to model-derived fitting of groundwater velocity and dispersion, all while preserving the true concentration, as opposed to the dimensionless relative concentration, of the potentially reactive solute. A new measure of overall solute reactivity is also introduced that summates relative temporal moments to quantify and rank the reactivity of a suite of solutes. The method is described and applied to numerical model simulations and field tracer data to demonstrate its utility for combined visual–quantitative breakthrough curve separation to better characterize reactive solute transport in applied tracer studies.

Engineering practice in monitoring design aims at the optimum number of observation wells needed to assess the growth of a contaminated volume groundwater, the plume. Available methodologies rely on a combination of a numerical groundwater transport model, GIS-techniques and an optimization technique and require a relative huge amount of data and computer resources. The method of advective transport phenomena enables to calculate the longitudinal and vertical growth of a contaminant plume along the flow path by simple analytic expressions using only three stochastic parameters, the log conductivity variance and the horizontal and vertical characteristic lengths, that together describe the heterogeneity of the aquifer. In previous work, the calculated plume growth has been verified in 12 large experiments all over the world. The method is used to investigate the relationship between uncertainty in the conductivity variation and the plume growth by calculation of the spreading of water particles in a vertical section along the traveled path. In a very heterogeneous aquifer, virtually all water particles spread forward about equally generating a limited forward growth compared to the traveled distance that is not sensitive to uncertainty in the conductivity. In a nearly homogenous aquifer, only a part of the water particles is spread forward, which is repeated at different depths along the traveled path causing significant uncertainty in the position and length of the plume growth. Therefore, an observation network should be designed more densely in a homogeneous aquifer than in a heterogeneous one. A calculation tool is provided.

Transient hydraulic tomography (THT) is proven to be effective in representing hydraulic and storage properties in diverse hydrogeologic settings. Sequential inversion of THT is computationally efficient, however, its accuracy is constrained by the number and sequence of pumping datasets used in the inversion. While signal-to-noise ratio (SNR) is commonly used to regulate the order of pumping datasets, it often disregards the information content. We propose an alternate strategy to rank the pumping ports based on the information contained in the data for use with inversion. A non-parametric Gringorten plotting position was used to generate cumulative distribution functions (CDFs) of the transient datasets, with the CDF corresponding to the maximum drawdown port set as a reference. The Kullback–Leibler divergence (KLD) is employed to quantify variations in time-drawdown datasets by statistically measuring the divergence from the reference distribution. Pumping ports are then ranked in the decreasing order of KLD and further used in the inversion. The proposed methodology is tested under a controlled environment using a laboratory sandbox model. Discrete wavelet transform (DWT) was applied to denoise the raw pumping datasets, and PEST coupled with MODFLOW was used to perform the inversion. The performance of KLD-assisted inversion (RMSE_KLD = 0.278 ± 0.177 cm) is found to be superior to SNR-assisted inversion (RMSE_SNR = 1.075 ± 0.990 cm). Further, a reduction in THT data (by 68%) by specifying a threshold on KLD (>10) has drastically reduced the computational time (by 64%) with commensurable accuracy (RMSE_KLDF = 0.265 ± 0.121 cm). Our findings lead to the conclusion that sequential inversion of THT with information-driven datasets outperforms quality-driven datasets, even with reduced pump-test data.