Imaging the magmatic system beneath the Krafla geothermal field, Iceland: A new 3-D electrical resistivity model from inversion of magnetotelluric data

Abstract

SUMMARY Krafla is an active volcanic field and a high-temperature geothermal system in northeast Iceland. As part of a program to produce more energy from higher temperature wells, the IDDP-1 well was drilled in 2009 to reach supercritical fluid conditions below the Krafla geothermal field. However, drilling ended prematurely when the well unexpectedly encountered rhyolite magma at a depth of 2.1 km. In this paper we re-examine the magnetotelluric (MT) data that were used to model the electrical resistivity structure at Krafla. We present a new 3-D resistivity model that differs from previous inversions due to (1) using the full impedance tensor data and (2) a finely discretized mesh with horizontal cell dimensions of 100 m by 100 m. We obtained similar resistivity models from using two different prior models: a uniform half-space, and a previously published 1-D resistivity model. Our model contains a near-surface resistive layer of unaltered basalt and a low resistivity layer of hydrothermal alteration (C1). A resistive region (R1) at 1 to 2 km depth corresponds to chlorite-epidote alteration minerals that are stable at temperatures of about 220 to 500 °C. A low resistivity feature (C2) coincides with the Hveragil fault system, a zone of increased permeability allowing interaction of aquifer fluids with magmatic fluids and gases. Our model contains a large, low resistivity zone (C3) below the northern half of the Krafla volcanic field that domes upward to a depth of about 1.6 km b.s.l. C3 is partially coincident with reported low S-wave velocity zones which could be due to partial melt or aqueous fluids. The low resistivity could also be attributed to dehydration and decomposition of chlorite and epidote that occurs above 500 °C. As opposed to previously published resistivity models, our resistivity model shows that IDDP-1 encountered rhyolite magma near the upper edge of C3, where it intersects C2. In order to assess the sensitivity of the MT data to melt at the bottom of IDDP-1, we added hypothetical magma bodies with resistivities of 0.1 to 30 Ωm to our resistivity model and compared the synthetic MT data to the original inversion response. We used two methods to compare the MT data fit: (1) the change in r.m.s. misfit and (2) an asymptotic p-value obtained from the Kolmogorov–Smirnov (K–S) statistical test on the two sets of data residuals. We determined that the MT data can only detect sills that are unrealistically large (2.25 km3) with very low resistivities (0.1 or 0.3 Ωm). Smaller magma bodies (0.125 and 1 km3) were not detected; thus the MT data are not sensitive to small rhyolite magma bodies near the bottom of IDDP-1. Our tests gave similar results when evaluating the changes in r.m.s. misfit and the K–S test p-values, but the K–S test is a more objective method than appraising a relative change in r.m.s. misfit. Our resistivity model and resolution tests are consistent with the idea of rhyolite melt forming by re-melting of hydrothermally altered basalt on the edges of a deeper magma body.

Probing the Southern African Lithosphere With Magnetotellurics—Part I: Model Construction

“…This approach was designed to test the impacts of heterogeneous station coverage and of noisy and distorted data on the results of each inversion algorithm. A similar strategy was pursued by Lee et al (2020), e.g., to homogenize data coverage.…”

Estimation of Spatial Distribution and Fluid Fraction of a Potential Supercritical Geothermal Reservoir by Magnetotelluric Data: A Case Study From Yuzawa Geothermal Field, NE Japan

“…Similar conductors at depths of 2–15 km in other volcanoes have been interpreted as magmatic fluids and melt (e.g., Mount St Helens: Bedrosian et al., 2018, Geysers: Peacock et al., 2020, Laguna del Maule: Cordell et al., 2020, Kirishima: Aizawa et al., 2014; Taupo: Bertrand et al., 2012, Krafla: Lee et al., 2020). A minimum solidus temperature of 725°C at a pressure of 100 MPa (Bowles‐Martinez & Schultz, 2020; Tuttle & Bowen, 1958), suggests that the part of C 1 at a temperature lower than 725°C is mostly crystallized and high conductivity values of the part of C 1 <725°C is derived from the existence of magmatic fluids, rather than melt.…”

Electrical Conductive Fluidized Zones and Their Influence on the Earthquake Nucleation, Growth, and Arrest Processes of the 2016 Kumamoto Earthquake Sequence, Kyushu Island, Japan

“…Therefore, we argue that the pore-uid pressure-which is considered to be high inside and near lowresistivity zones, such as deep magmatic uid zones (Fournier, 1999;Lee et al, 2020) or a fracture zone that transports magmatic volatiles (Aizawa et al, 2016; Lee et al, 2020)-plays an important role in the evolution of crustal earthquake rupture. We hypothesize that the pre-failure pressure/temperature (PT) gradient (spatial difference) of the pore uids contributes to the propagation and arrest of earthquake rupture.…”