Impact of geostatistical reconstruction approaches on model calibration for flow in highly heterogeneous aquifers

Abstract

Our study is aimed at assessing the extent at which relying on differing geostatistical approaches may affect characterization of the connectivity of geomaterials (or facies) and, in turn, model calibration outputs in highly heterogeneous aquifers. We set our study within a probabilistic framework, by relying on a numerical Monte Carlo (MC) approach. The reconstruction of the spatial distribution of geomaterials and flow simulations are patterned after a field scenario corresponding to the aquifer system serving the city of Bologna (Northern Italy). Two collections of MC realizations of facies distributions, conditional on available lithological data, are generated through two alternative geostatistically-based techniques, i.e., Sequential Indicator and Transition-Probability simulation. Hydraulic conductivity values of the least- and most-conductive facies are estimated within each MC simulation in the context of a Maximum Likelihood (ML) approach by considering available piezometric data. We provide evidence that the choice of the facies reconstruction technique (1) impacts the degree of connectivity of facies whose proportions are close to the percolation threshold while (2) is not sensibly affecting the connectivity associated with facies whose proportions are much larger than the percolation threshold. By relying on the unique (lithological and hydrological) data-set at our disposal, we also explore the performance of ML-based model identification criteria to (1) discriminate amongst competitive facies reconstruction geostatistical models and (2) quantify the (posterior probabilistic) weight associated with each model. We then show that ML-based model averaging provides estimates of hydraulic heads which are slightly more in agreement with available data when compared to the best-performing realization in the T-PROGS set than considering its counterpart associated with the SISIM-based collection.

Using sequential conditioning to explore uncertainties in geostatistical characterization and in groundwater transport predictions

“…Because of this, we use stochastic frameworks to characterize heterogeneity structure, based on geostatistical or physical process-based modelling methods (e.g. Riva et al, 2006;Riva et al, 2008;Scheibe et al, 2015;Siena & Riva, 2020).…”

“…The choice of the most appropriate structural model of heterogeneity largely depends on the features that control the predictive response of concern (e.g., Jafarpour & Tarrahi, 2011;Ciriello et al, 2013;Riva et al, 2015). We opted to use TP simulation which has been shown to be superior for representing the connectivity of high permeability pathways (Siena & Riva, 2020). TP simulation is also well-established and used in numerous modelling studies, including groundwater modelling studies, that rely on a geostatistical description of the spatial dependencies of selected categories (e.g.…”

“…The non-uniqueness of the TP model hyperparameters, as depicted in Figure 4, reflects the lack of information regarding the OFG category mean lengths in the lateral direction (L x ). Nonuniqueness of geostatistical models is acknowledged and discussed in several studies (Harp & Vessilinov, 2012;Koch, 2013;He et al, 2014;Siena & Riva, 2020). Despite this, the impact of geostatistical model equifinality on the quantification of the uncertainty of groundwater model predictions of interest is seldom considered by practitioners in decision support models (Sanchez-Villa & Fernandez-Garcia, 2016).…”

Impacts of High-Concentration Turbid Water on the Groundwater Environment of the Tedori River Alluvial Fan in Japan

“…To analyze these complex factors, it is considered effective to conduct a scenario analysis using a hydrological model that represents the water circulation within the alluvial fan. It is generally believed that the spatial distribution of geology and soil properties has a significant influence on hydrological characteristics [27], and 3D groundwater models are often employed for impact assessment (e.g., Thakur [28]; Siena and Riva [29]). However, the Tedori River alluvial fan is composed of relatively uniform sandy gravel deposits.…”

A comprehensive framework for stochastic calibration and sensitivity analysis of large-scale groundwater models

“…The latter stems from the (generally unknown) spatial distribution of medium properties, boundary conditions and/or forcing terms, and limited data availability. This issue could be addressed upon relying on a stochastic framework for model calibration (e.g., Neuman, 2003;Riva et al, 2009;Ye et al, 2010;Panzeri et al, 2015;Siena and Riva, 2020). However, stochastic model calibration presents significant challenges, particularly in terms of computational cost when dealing with multiple sources of uncertainty (e.g., Vrugt et al, 2008;Hendricks Franssen et al, 2009;Zhou et al, 2014).…”