This paper presents two methods for joint inversion of aquifer test data, magnetic resonance sounding (MRS) data, and transient electromagnetic data acquired from a multilayer hydrogeological system. The link between the MRS model and the groundwater model is created by tying hydraulic conductivities (k) derived from MRS parameters to those of the groundwater model. Method 1 applies k estimated from MRS directly in the groundwater model, during the inversion. Method 2 on the other hand uses the petrophysical relation as a regularization constraint that only enforces k estimated for the groundwater model to be equal to MRS derived k to the extent that data can be fitted. Both methodologies can jointly calibrate parameters pertaining to the individual models as well as a parameter pertaining to the petrophysical relation. This allows the petrophysical relation to adapt to the local conditions during the inversion. The methods are tested using a synthetic data set as well as a field data set. In combination, the two case studies show that the joint methods can constrain the inversion to achieve estimates of k, decay times, and water contents for a leaky confined aquifer system. We show that the geophysical data can assist in determining otherwise insensitive k, and vice versa. Based on our experiments and results, we mainly advocate the future application of method 2 since this seems to produce the most reliable results, has a faster inversion runtime, and is applicable also for linking k of 3-D groundwater flow models to multiple MRS soundings.
Hydrological models are often set up to provide specific forecasts of interest. Owing to the inherent uncertainty in data used to derive model structure and used to constrain parameter variations, the model forecasts will be uncertain. Additional data collection is often performed to minimize this forecast uncertainty. Given our common financial restrictions, it is critical that we identify data with maximal information content with respect to forecast of interest. In practice, this often devolves to qualitative decisions based on expert opinion. However, there is no assurance that this will lead to optimal design, especially for complex hydrogeological problems. Specifically, these complexities include considerations of multiple forecasts, shared information among potential observations, information content of existing data, and the assumptions and simplifications underlying model construction. In the present study, we extend previous data worth analyses to include: simultaneous selection of multiple new measurements and consideration of multiple forecasts of interest. We show how the suggested approach can be used to optimize data collection. This can be used in a manner that suggests specific measurement sets or that produces probability maps indicating areas likely to be informative for specific forecasts. Moreover, we provide examples documenting that sequential measurement election approaches often lead to suboptimal designs and that estimates of data covariance should be included when selecting future measurement sets.
Abstract. Nitrate contamination of subsurface aquifers is an ongoing environmental challenge due to nitrogen (N) leaching from intensive N fertilization and management on agricultural fields. The distribution and fate of nitrate in aquifers are primarily governed by geological, hydrological and geochemical conditions of the subsurface. Therefore, we propose a novel approach to modeling both geology and redox architectures simultaneously in high-resolution 3D (25m×25m×2m) using multiple-point geostatistical (MPS) simulation. Data consist of (1) mainly resistivities of the subsurface mapped with towed transient electromagnetic measurements (tTEM), (2) lithologies from borehole observations, (3) redox conditions from colors reported in borehole observations, and (4) chemistry analyses from water samples. Based on the collected data and supplementary surface geology maps and digital elevation models, the simulation domain was subdivided into geological elements with similar geological traits and depositional histories. The conceptual understandings of the geological and redox architectures of the study system were introduced to the simulation as training images for each geological element. On the basis of these training images and conditioning data, independent realizations were jointly simulated of geology and redox inside each geological element and stitched together into a larger model. The joint simulation of geological and redox architectures, which is one of the strengths of MPS compared to other geostatistical methods, ensures that the two architectures in general show coherent patterns. Despite the inherent subjectivity of interpretations of the training images and geological element boundaries, they enable an easy and intuitive incorporation of qualitative knowledge of geology and geochemistry in quantitative simulations of the subsurface architectures. Altogether, we conclude that our approach effectively simulates the consistent geological and redox architectures of the subsurface that can be used for hydrological modeling with nitrogen (N) transport, which may lead to a better understanding of N fate in the subsurface and to future more targeted regulation of agriculture.
Groundwater model predictions are often uncertain due to inherent uncertainties in model input data. Monitored field data are commonly used to assess the performance of a model and reduce its prediction uncertainty. Given the high cost of data collection, it is imperative to identify the minimum number of required observation wells and to define the optimal locations of sampling points in space and depth. This study proposes a design methodology to optimize the number and location of additional observation wells that will effectively measure multiple hydrogeological parameters at different depths. For this purpose, we incorporated Bayesian model averaging and genetic algorithms into a linear data‐worth analysis in order to conduct a three‐dimensional location search for new sampling locations. We evaluated the methodology by applying it along a heterogeneous coastal aquifer with limited hydrogeological data that is experiencing salt water intrusion (SWI). The aim of the model was to identify the best locations for sampling head and salinity data, while reducing uncertainty when predicting multiple variables of SWI. The resulting optimal locations for new observation wells varied with the defined design constraints. The optimal design (OD) depended on the ratio of the start‐up cost of the monitoring program and the installation cost of the first observation well. The proposed methodology can contribute toward reducing the uncertainties associated with predicting multiple variables in a groundwater system.
The nuclear magnetic resonance sounding (MRS) method is used increasingly as a tool for hydrological investigations. Compared to other geophysical methods, the advantage of MRS is that it is directly sensitive to the presence of water in the subsurface. Data interpretations can also be used to get information about the subsurface pore structures, which under special conditions can be related to hydraulic properties such as aquifer transmissivity. However, to broaden the usage of this information in hydrological modeling, the uncertainties related to these transmissivity estimates must be determined. Otherwise, properly balanced weights cannot be given to the prior information obtained from MRS transmissivity estimates as compared to the hydrological data sets when used for groundwater model calibration. We have developed a methodology to estimate the uncertainties of MRS-based transmissivity estimates. Compared to previous studies, the methodology is well defined, and it takes into account important factors such as the uncertainties of the hydraulically estimated transmissivities, the uncertainty of the correlation factor in the petrophysical relation, and the uncertainties and correlations of the geophysically estimated parameters. We have determined the correlations and uncertainties of the geophysical parameters using a linear and a nonlinear method, and we find that the results are comparable.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.