Estimation of effective porosity in large-scale groundwater models by combining particle tracking, auto-calibration and <sup>14</sup>C dating

Abstract

Abstract. Effective porosity plays an important role in contaminant management. However, the effective porosity is often assumed constant in space and hence either neglected or simplified in transport model calibration. Based on a calibrated highly parametrized flow model, a three-dimensional advective transport model (MODPATH) of a 1300 km is improved significantly by the calibration of seven porosity units by a reduction of ME of 99% and RMS of 82% compared to a uniform porosity approach. Groundwater ages range from a few hundred years in the Pleistocene to several thousand years in Miocene aquifers. The advective age distributions derived from particle tracking at each sampling well show unimodal (for 10 younger ages) to multimodal (for older ages) shapes and thus reflect the heterogeneity that particles encounter along their travel path. The estimated effective porosity field, with values ranging between 4.3% in clay and 45% in sand formations, is used in a direct simulation of distributed mean groundwater ages. Although the absolute ages are affected by various uncertainties, a unique insight into the complex three-dimensional age distribution pattern and potential advance of young contaminated groundwater in the investigated regional aquifer system is provided, highlighting the importance of estimating effective porosity 15 in groundwater transport modelling and the implications for groundwater quantity and quality assessment and management.

Origin and Dynamics of Saltwater Intrusion in a Regional Aquifer: Combining 3‐D Saltwater Modeling With Geophysical and Geochemical Data

“…Distributed net recharge was available from the Danish national water resource model (Henriksen et al, ). In contrast to the previous studies (Meyer et al, , ) no groundwater abstraction was implemented since abstraction has not taken place in most of the historical periods considered here. Rivers and drains were implemented as head‐dependent flux boundaries with estimated conductances of 21.3 (m 2 /d)/m and 0.45 (m 2 /d)/m 2 , respectively.…”

“…The samples were analyzed for major ions in the analytical laboratory at GEUS in Copenhagen, Denmark. Of these, 18 were also analyzed for δ 13 C ‰ VPBD and 14 C at the AGH University of Science and Technology in Kraków and the Poznań Radiocarbon Laboratory, Poland (details on radiocarbon dating and the correction terms that had been applied to derive groundwater ages can be found in Meyer et al, ).…”

“…The model setup and parameter estimates of a calibrated flow and advective transport model using MODFLOW and MODPATH (Meyer et al, , ) formed the base for simulating saltwater intrusion with SEAWAT. A short summary of the relevant information is provided in the following.…”

Pathline Density Distributions in a Null‐Space Monte Carlo Approach to Assess Groundwater Pathways

“…Overall, the reduction in uncertainty is relatively limited, indicating that the current dataset does not provide sufficient information for the inverse parameter estimation. Adding more and different observations would likely help to better constrain the parameter estimation (Bravo et al ; Anderson ; Anderson et al ; Colombani et al ; Meyer et al ).…”

“…In addition to the aforementioned structural error (Doherty and Welter ; Ferré ), measurement type and noise can also influence the calibrated heterogeneity in the K field (Poeter and Hill ; Doherty ; among others) and consequently different pathline distributions can be obtained. Apart from measured hydraulic heads, other observations such as temperature, solute concentrations, and flux observations could also be used as a means to constrain the estimates of groundwater model parameters (Bravo et al ; Anderson ; Anderson et al ; Colombani et al ; Meyer et al ). It can be speculated that adding more and different observations will help to better constrain the parameter estimation.…”