Global sampling to assess the value of diverse observations in conditioning a real‐world groundwater flow and transport model

Abstract

The use of additional types of observational data has often been suggested to alleviate the ill-posedness inherent to parameter estimation of groundwater models and constrain model uncertainty. Disinformation in observational data caused by errors in either the observations or the chosen model structure may, however, confound the value of adding observational data in model conditioning. This paper uses the global generalized likelihood uncertainty estimation methodology to investigate the value of different observational data types (heads, fluxes, salinity, and temperature) in conditioning a groundwater flow and transport model of an extensively monitored field site in the Netherlands. We compared model conditioning using the real observations to a synthetic model experiment, to demonstrate the possible influence of disinformation in observational data in model conditioning. Results showed that the value of different conditioning targets was less evident when conditioning to real measurements than in a measurement error-only synthetic model experiment. While in the synthetic experiment, all conditioning targets clearly improved model outcomes, minor improvements or even worsening of model outcomes was observed for the real measurements. This result was caused by errors in both the model structure and the observations, resulting in disinformation in the observational data. The observed impact of disinformation in the observational data reiterates the necessity of thorough data validation and the need for accounting for both model structural and observational errors in model conditioning. It further suggests caution when translating results of synthetic modeling examples to real-world applications. Still, applying diverse conditioning data types was found to be essential to constrain uncertainty, especially in the transport of solutes in the model

Beyond Classical Observations in Hydrogeology: The Advantages of Including Exchange Flux, Temperature, Tracer Concentration, Residence Time, and Soil Moisture Observations in Groundwater Model Calibration

“…However, as soon as observations of Q GW in the form of spring discharge were included alongside observations of H, Q SW , and T, the information from T for hydraulic parameters became negligible—observations of T then only helped informing thermal parameters. In another equally rigorous study that systematically analyzed the data worth of observations of T alongside other observation types, Delsmann et al () quantified the value of observations of H, T, Q GW , and C for the calibration of an IFM. The worth of different observation types in calibrating an IFM was estimated using a uniform random sampling approach (generalized likelihood uncertainty estimation; Beven & Binley, ; Stedinger et al, ).…”

“…The worth of different observation types in calibrating an IFM was estimated using a uniform random sampling approach (generalized likelihood uncertainty estimation; Beven & Binley, ; Stedinger et al, ). Like Bravo et al () and Kurtz et al (), Delsmann et al () carried out both synthetic and real‐world modeling experiments to investigate the data worth of different observation types for flow model calibration. Delsmann et al () showed that, overall, the inclusion of unconventional observations improved model parameters and predictions if the observations were of a sufficient quality, that is, if the observations were associated with a sufficiently small measurement error.…”

“…Like Bravo et al () and Kurtz et al (), Delsmann et al () carried out both synthetic and real‐world modeling experiments to investigate the data worth of different observation types for flow model calibration. Delsmann et al () showed that, overall, the inclusion of unconventional observations improved model parameters and predictions if the observations were of a sufficient quality, that is, if the observations were associated with a sufficiently small measurement error. They found, however, that with observations of T only two parameters could be calibrated well: f and the thermal conductivity of the porous medium ( κ bulk ).…”

“…They found, however, that with observations of T only two parameters could be calibrated well: f and the thermal conductivity of the porous medium ( κ bulk ). Delsmann et al () also found that (1) all other observation types employed (i.e., H, Q GW , and C) constrained the hydraulic gradients and flow field better than observations of T and that (2) by including observations of T, all model outcomes other than the simulation of heat transport worsened. Only the inclusion of observations of C (in the form of total salinity load) allowed simultaneously improving all model outcomes.…”

“…The review showed that more often than not, the validity of flux or mass transport simulations calibrated against a combination of classical observations and T was not assessed by directly comparing the simulations to flux or mass transport observations. This is not good practice, especially in the light of the findings from the thorough analyses carried out by Bravo et al (), Kurtz et al (), Masbruch et al (), and Delsmann et al (), who demonstrated that as soon as there are multiple unknowns among the hydraulic and thermal properties and relevant exchange fluxes, and as soon as measurement errors are taken into account, calibration of an IFM against H and T is likely to result in correct temperature reproduction but incorrect representation of fluxes. If observations of T were complemented by a combination of classical and additional unconventional observations (e.g., by C or Q GW ), an acceptable calibration of flow, heat, and mass transport was much more likely.…”