Falsification and corroboration of conceptual hydrological models using geophysical data

Abstract

Geophysical data may provide crucial information about hydrological properties, states, and processes that are difficult to obtain by other means. Large data sets can be acquired over widely different scales in a minimally invasive manner and at comparatively low costs, but their effective use in hydrology makes it necessary to understand the fidelity of geophysical models, the assumptions made in their construction, and the links between geophysical and hydrological properties. Geophysics has been applied for groundwater prospecting for almost a century, but it is only in the last 20 years that it is regularly used together with classical hydrological data to build predictive hydrological models. A largely unexplored venue for future work is to use geophysical data to falsify or rank competing conceptual hydrological models. A promising cornerstone for such a model selection strategy is the Bayes factor, but it can only be calculated reliably when considering the main sources of uncertainty throughout the hydrogeophysical parameter estimation process. Most classical geophysical imaging tools tend to favor models with smoothly varying property fields that are at odds with most conceptual hydrological models of interest. It is thus necessary to account for this bias or use alternative approaches in which proposed conceptual models are honored at all steps in the model building process. This article is categorized under: Science of Water > Hydrological Processes

Bayesian model selection in hydrogeophysics: Application to conceptual subsurface models of the South Oyster Bacterial Transport Site, Virginia, USA

“…Schöniger et al (2015a) ;(2015b) ;(2014) found reliable evidence estimates with the BFMC method for different case-studies in hydrology. For our set-up with small errors and high data and model dimensions, we found that reliable evidence estimation with the BFMC method would need prohibitive computation times.…”

“…Geophysical data, for instance, warrant a detailed characterization of the hydrologic properties of the vadose zone and aquifers ( Binley et al, 2010;Hubbard and Linde, 2011;Hubbard and Rubin, 2005 ). Most published studies in the hydrogeophysical literature rely on a single conceptual representation of the subsurface, without recourse to explicit treatment of the actual uncertainty associated with the choice of a single conceptual model ( Linde, 2014;Linde et al, 2015 ). Geophysics-based model selection has received relatively limited attention, which is somewhat surprising, as geophysical data contain a wealth of information about the structure of the subsurface.…”

“…In contrast to current practice, we should not rely only on a single conceptualization and parameterization of the subsur-face, but instead determine as well the proper spatial arrangement of variables of interest such as porosity and moisture content. One approach of doing this is to implement model selection, and use the geophysical data to provide guidance about which representation of the subsurface is most supported by the available data among a set of competing conceptual models ( Linde, 2014 ). Such an approach will not only enhance the fidelity of our subsurface investigations, but will also further promulgate and disseminate the importance of geophysical data in hydrologic and environmental studies.…”

Probabilistic electrical resistivity tomography of a CO2 sequestration analog

“…Reduction of the model dimensions is particularly important for Bayesian inversion, since with growing number of free parameters the solution space becomes increasingly void and MCMC sampling very inefficient (Curtis and Lomax, 2001). Models can be represented by an expansion of base functions such that (e.g., Linde, 2014;Sambridge et al, 2013) …”

On the use of spatially distributed, time-lapse microgravity surveys to inform hydrological modeling

“…In particular, Efstratiadis and Koutsoyiannis [] and Beven and Westerberg [] indicated that epistemic uncertainty may be related to the noninformativeness (i.e., limited effectiveness) of observational data in conditioning model parameters and structure. Furthermore, Efstratiadis and Koutsoyiannis [] and Linde [] evidenced that, in some cases, the additional information provided by the adoption of a multiobjective framework may lead to rejection of an inappropriate conceptual model, which would appear as proper against a single criterion (i.e., the classical comparison of observed and simulated water discharges).…”