Analysis of tectonic and magmatic stresses controlling volcanic spreading over multiple time scales. A case study from Kilauea Volcano, Hawaii, USA
Geophysics
Final Report Abstract
Many volcanoes show evidence of flank motion, landsliding, or sector collapse. It is key to understand the causes and controls of this motion in order to estimate the hazard from these processes in volcanic regions. The process can be monitored by surface displacement data from space-geodetic measurements, in particular Global Positioning System (GPS) and InSAR (Interferometric Satellite Aperture Radar). The time-series analysis allows distinguishing between periods of steady motion and episodic events, such as flank motion facilitating earthquakes or the opening of flank-perpendicular dikes. There is no general implication, however, on how these motions affect the long-term stability of the volcanic flanks. For example, active dike intrusions occur due to magmatic overpressure and steepen the flanks and thus destabilize it, while passive intrusions allow the near-surface flank to “catch-up” with deep seated deformation and thus can stabilize the flank. In order to address the effect of steady flank motion and episodic deformation events the mechanical disequilibrium between the body forces from gravity and stress gradients, subject to the individual topography, structure and material properties was addressed using numerical modeling. Here, a case study on the Kilauea Volcano, Hawai’i was carried out. Kilauea Volcano shows a steady seaward flank motion rate of up to 6cm/yr at the surface, modified by episodic magmatic diking and fault slip events at the basal decollement of all kinds of slip modes (from aseismic creep over slow slip events to large magnitude earthquakes such as in 1972 and 1918). In a first attempt a numerical model was designed to understand the effect of gravity on high-density material (olivine crystals) in a deep-seated magma mush. It was found that the density contrast with respect to the host-rock is not sufficient to explain the observed flank motion. A lower viscosity of this body (10^19 Pa.s) with respect to the host rock (10^23 Pa.s), however, can facilitate accelerated continuum deformation at depth in accordance with geodetic rates of surface motion (6 cm/yr). The limitation of this model is, that the flank motion is only moving in a steadystate process if the volcanos’ decollement is allowed to slip and the dikes are allowed to open at any time. Both of these deformation processes, however, are known to occur mostly as episodic events (at most, only part of the decollement may slip continuously in an aseismic matter). Therefore, the model was modified to include fault friction on structural discontinuities and to allow for episodic dike opening. Active dike opening in response to magmatic overpressure, however, could only be included as a boundary condition. With gravitational forces being the only driver of the model, the deformation is a response to the model assumptions made on the internal structure of the volcano and friction laws. A major task was to examine the internal structure and build a three-dimensional model geometry using seismicity data to define the location of deformation zones. On the same time, the influence of the assumptions made on the geometry internal structure, such as the location, length, and dip of faults were tested using 2D models. The major finding of this work is that active dike intrusions in the shallow rift zone of Kilauea with typical opening rates as obtained from geodetic data analysis induce significant relative displacement across the decollement. Further analysis on the realistic assumptions of the fault friction and the consideration of deformation gradients in the third dimension are necessary to provide a better constraint on these results.
Publications
-
The role of viscous magma mush spreading in volcanic flank motion at Kīlauea Volcano, Hawai‘i. Journal of Geophysical Research: Solid Earth, 118(5), 2474-2487.
Plattner, C.; Amelung, F.; Baker, S.; Govers, R. & Poland, M.
