2021 | Constraining MT models via laboratory measurements of volcanic rocks

Aims: It is possible to computationally extrapolate electrical measurements of rocks made at ambient conditions to higher P and T, comparable to conditions at depths sampled by magnetotelluric (MT) inversions. This employs a regression model (e.g. Eq. (4) from Ucok; 1979) to correct for temperature, salinity and saline species differences between laboratory and appropriate in situ conditions (Kluge, 2017). However, it is important to constrain these predictions with a few real measurements. We will thus measure resistivity and resistivity anisotropy of volcanic rocks saturated in varying salinity fluids, and at ambient and elevated P and T (Pconf < 20kbar; T < 800OC). The results will provide better constraints on inversions of magnetotellutic (MT) data from the Cebourco Volcano, Mexico, which is subject of current DFG-funded research (DEMITZ project; https://www.geophysik.uni-frankfurt.de/96829516/Projects) by Junge’s Applied Geophysics Group, building on a past PhD project (Hering, 2019). The results will be upscaled via numerical studies with COMSOL.

Cooperation: Experiments will be carried out both at Mainz (Toy’s Lab) and Frankfurt (Bagdassarov/Klimm lab). Castro will be responsible for appropriate sample selection. Junge will advise on forward modelling of electrical responses from laboratory data, and will employ these data as constraints on future MT inversions of Cebourco and the surrounding region.

Work plan: We will attempt to acquire samples of representative rocks from Ceboruco Volcano, by request from the main researcher in that region (https://ucol.academia.edu/NickVarley), or they will be collected by Prof. Junge’s team during their next MT fieldwork. Alternatively, we will analyze comparable rock types provided by collaborators from e.g. New Zealand (Taupo Volcanic Zone), Campi Flegrei Volcano (Italy), or other comparable samples already in the possession of Jon Castro. We will establish a benchtop system to measure resistivities at Universität Mainz. The proposed HiWi (Kluge) already used such equipment in Dr. David Lockner’s lab at USGS-Menlo Park, CA, during her MSc (Kluge, 2017; Kluge et al., in prep.). Kirilova has also used similar equipment guided by Prof. Andre Revil at Laboratoiré EDYTEM, Chambéry (cf. Abdulsamad et al., 2019). Both past collaborators who are very experienced in this topic (e.g. Lockner & Byerlee, 1985; Revil & Glover, 1997) have offered advice on setting up our equipment. Saline fluids and surface charges - especially on clay minerals such as smectites commonly encountered near surface in volcanic systems (Revil et al., 2019) – are the primary carrier of electrical charge in most rocks, and their connectivity varies greatly due to porosity reduction at elevated Pconf (cf. Toy et al., 2019). We will resolve the contribution of rock vs. fluid to the bulk electrical response of any sample (ie. fit Archie’s Law; Archie, 1945), by making measurements in variable salinity fluids at benchtop conditions in this new equipment at Universität Mainz. We will employ the existing piston-cylinder apparatus equipped with electrical meters in the highpressure lab at Frankfurt formerly belonging to Prof. Bagdassarov, now managed by Dr. Klimm) to make measurements of electrical properties of these samples at elevated pressure and temperature (Bagdassarov et al., 2001). It would be very valuable to introduce saline fluids into the samples in these experiments, but that is unlikely to be feasible in this small project. Nevertheless, we expect a combination of benchtop data that constrain the Archie’s Law fit, and dry high pressure experiments will permit adequate regression modelling of electrical properties at other P and T. A Hiwi based at Frankfurt, guided by Junge, will employ Comsol Multiphysics® FE software to compute bulk electrical properties of the volcano by simulating randomly distributed multiphase media comprising components with different conductivities indicated by the experimental measurements (cf. Rödder & Junge, 2016; Löwer & Junge, 2017).

Expected outcome: Potentially two joint publications: (i) Toy-Klimm-Bagdassarov describing the results of experiments, and (ii) Toy-Junge comparing lab results to COMSOL models.