The stratification of sedimentary aquifers introduces spatial variability in hydraulic conductivity, primarily between individual horizontal layers. On larger scales, the vertical heterogeneity enhances hydraulic anisotropy, with the horizontal conductivity typically exceeding the vertical one. In this study, the hydraulic anisotropy of a stratified aquifer is estimated from data of hydraulic tests in which water is sequentially extracted from well sections screened at different depths, and the hydraulic response is measured at various multilevel observation wells. The applicability of the method is demonstrated by field tests in a fluvial gravel aquifer in the Upper Rhine Valley, Germany. A homogeneous anisotropic model, and models with three and five anisotropic layers, are fitted to the measured drawdowns in the steady-shape regime, in which differences in hydraulic head between observation locations do not change over time even though the head values themselves change. The position of the five horizontal layers is based on the lithology of the drilling profile at the pumping-well location. The three-layer model is achieved by merging insensitive or similar layers with sensitive layers. The fits result in estimates of the radial and vertical hydraulic conductivities for all layers of the respective models, which are upscaled to effective parameters over the entire depth in the case of the multilayer models. The homogeneous model shows significantly higher errors than those of the heterogeneous models. The heterogeneous locally anisotropic models not only reveal vertical variability of hydraulic conductivity, but also lead to a three-times larger anisotropy ratio upon upscaling.

When applying environmental models, the choice of model complexity and the design of field campaigns depend on each other and on the modeling/prediction goal. We propose jointly optimizing model complexity and data collection (design) by maximizing the expected performance for the modeling goal. We use ensembles of highly resolved virtual realities and of less complex modeling variants that differ in their degrees of upscaling and simplified parameterization. For each design under consideration, we simulate hypothetical measurement data (subject to noise) with all realizations of all models. To mimic model calibration with hypothetical data, we identify pairs of best fitting realizations between virtual reality and each model variant for each design. Then, we emulate model choice by selecting (across the model variants, for each design and for each virtual reality) the pair that shows the best predictive match in the modeling goal. Finally, we identify the model/design combination that offers, on average over all virtual realities, the best predictive match. As a test application, we consider a heterogeneous, stratified aquifer, in which the stratification enhances hydraulic anisotropy on the macroscale. We define two different modeling goals: (a) estimating the hydraulic conductivity tensor upscaled to the full aquifer thickness and (b) predicting the pumping rate needed to dewater a construction pit. Our results indicate that jointly optimizing observation networks and model selection can reduce the prediction uncertainty of parameters at lower experimental costs. We also show that the involved trade‐offs between model complexity and required design depend on the target quantity.

ZusammenfassungDie horizontale Schichtung von Sedimentkörpern bewirkt, dass die effektive hydraulische Durchlässigkeit eines sedimentären Grundwasserleiters in horizontaler Richtung größer ist als in der vertikalen, was sich unter anderem auf die Ausbreitung des Absenktrichters bei Grundwasserentnahmen auswirkt. Um die großräumige hydraulische Anisotropie zu bestimmen, führten Maier et al. (2022a) an einem Testfeld am Oberrhein tiefenorientierte Pumpversuche mit einem Förderbrunnen mit drei Filterstrecken durch. Die Grundwasserentnahmen erfolgten nacheinander jeweils aus einer Tiefe. Die Absenkungen aller Versuche wurden in verschiedenen Abständen zum Förderbrunnen und verschiedenen Tiefen gemessen und gemeinsam mit einem Computermodell ausgewertet. In der vorliegenden Arbeit zeigen wir, dass die Interpretation von Absenkkurven bei Entnahme aus nur einer Tiefe zu Durchlässigkeitsbeiwerten führt, die vornehmlich die Bedingungen in der Entnahmetiefe widerspiegeln. Für die Bestimmung vertikaler Unterschiede der hydraulischen Durchlässigkeit und der hydraulischen Anisotropie ist es notwendig, Versuche mit mehreren Entnahmetiefen zu kombinieren. Der Arbeitsaufwand im Feld kann jedoch durch eine optimierte Beobachtungsstrategie reduziert werden.