Department of Geoscience

Workpackage 4

WP4. Coupled three-dimensional hydrological and hydrogeophysical modeling.

WP leader: Associate Professor Peter Bauer-Gottwein (DTU)

In WP4 we are developing automatic procedures linking geophysical information with lithological/hydrological information. The way of describing the link is the key difference between the three parallel developments in WP4:

• Coupled hydrogeophysical modeling. We are developing a method for small- and medium-scale (<10km2) coupled hydrogeophysical inversion of groundwater model parameters and geophysical model parameters using hydrogeological observations (potential heads, discharge, etc.) together with Electromagnetic and MRS data. The link is a hydrogeophysical relationship (in the simplest form Archie's law). The inversion minimizes an object function including the misfit between calculated and observed geophysical data, calculated and observed hydrological data, and regularization terms.

• Uncoupled hydrogeophysical modeling using probability functions. In this development we are building a lithological/hydrological model using a probability link between lithology and resistivity (Gunnink and Siemon, 2009). The link is based on borehole data and geophysical data established in WP3. Sequential indicator simulation (SIS) and/or transition probability indicator simulation (T-PROGS (Carle, 1999)) is used to produce a stochastic model of the subsurface based on drilling data; the three-dimensional probability link constrains the results to the local prior information from the geophysical models. Several realizations finally form the basis for average models as well as uncertainty fields for the different lithologies and/or hydrological parameters.

• Uncoupled hydrogeophysical modeling using an inversion approach. The inversion parameters describes a function linking the geophysical data with the lithological/hydrological parameters and the objective function minimizes the misfit between the geophysical models and borehole information. A simple version of this has been developed at Aarhus University and GEUS under the name SSV. In the SSV the link is a simple weight function between resistivity and clay content (low resistivity=clay; high resisivity=not clay). Preliminary investigations have shown that, by generalizing the link function, the methodology can be used in a more general context to build automatic lithological and hydrological models on survey scale.

All three parallel developments use the results from WP1-3 in their work.