HR: 16:45h
AN: V14B-04    [Abstracts]
TI: How Local Stresses can Prevent Volcanic Eruptions and Slips on Ring-Faults: The Need for Stress Monitoring of Volcanoes
AU: * Gudmundsson, A
EM: Agust.Gudmundsson@gwdg.de
AF: Department of Structural Geology and Geodynamics, Geoscience Center, University of Gottingen, Goldschmidtstrasse 3, Gottingen, 37077 Germany
AB: Local stresses in a volcano result from various volcanotectonic processes, the material properties of the rock units, and the tectonic environment within which the volcano is located. Ring faults and dikes (including inclined sheets) are rock fractures: they do not initiate and propagate unless the local stresses are favorable to these specific types of fracture. More specifically, no dike-fed eruption can occur unless the local stress field within the volcano is favorable to feeder-dike formation. Since most dikes are extension fractures, this means that the local stress in each rock unit through which the dike must pass on its way to the surface must favor the development of a magma-driven extension fracture, or else the dike becomes arrested. Thus, if there is any rock unit where the stress field is unfavorable to dike propagation along its potential pathway to the surface, the dike becomes arrested and an eruption is prevented. Recent field data indicate that most dikes never reach the surface; most potential feeder-dikes become arrested in some rock units and thus do not feed eruptions. Similarly, no ring-fault formation can occur, nor can there be slip on an existing ring fault, unless the local stress field within each rock unit at the location of the potential ring fault favors the formation or reactivation of a dip-slip fault. Since ring faults are normally very extensive, ranging in lengths from about 5 km to more than 200 km and reaching depths of many kilometers, a large rock mass has to be stress-homogenized for a ring fault to form or an existing one to slip. Most volcanic eruptions are fed by dikes; most collapse calderas are associated with ring faults. Thus, to make reliable forecasting of an eruption or caldera slip one must understand the local stresses in volcanoes. A composite volcano is composed of numerous layers (strata) many of which have contrasting mechanical properties, particularly at shallow depths. Some lava flows, welded pyrolastic units, and intrusions, for example, may be very stiff (with a high Young's modulus), whereas young and non-welded pyroclastic and sedimentary units are often very soft (with a low Young's modulus). The results of new conceptual and numerical stress-field models show, first, how the variation in the stress field ahead of a dike tip is determined by the mechanical layering; second, why most dikes are non-feeders; and, third, what stress conditions must be satisfied for a ring fault to form or slip. In particular, the numerical models explain (1) why the ring faults of most collapse calderas have neither the dips of ordinary normal faults nor those of ordinary reverse faults, but are rather close to vertical, and (2) why slips on existing ring faults are much more common in basaltic edifices, such as in Hawaii and Galapagos, than in ordinary composite volcanoes where there is no slip during most unrest periods. To improve our understanding of unrest periods and our assessment of the probability of a caldera slip or a dike-fed eruption we need stress monitoring of volcanoes. Stress monitoring through volcano seismicity and deformation is important but gives only very generic information on the state of stress. Effective stress monitoring can only be made through in situ stress measurements (perhaps using hydraulic fracturing) in many of the layers that constitute the volcano and by comparing the results with field observations and model predictions.
DE: 8164 Stresses: crust and lithosphere
DE: 8178 Tectonics and magmatism
DE: 8419 Volcano monitoring (7280)
DE: 8439 Physics and chemistry of magma bodies
DE: 8440 Calderas
SC: Volcanology, Geochemistry, Petrology [V]
MN: Fall Meeting 2005