Insights from deformation during the Pu'u O'o eruption of Kilauea volcano
Abstract
The Pu'u O'o eruptive period has been one of relatively stability, characterized by quasi steady deformation punctuated by discrete events and transients. During this time the south flank of Kilauea displaced SSE at up to 8 cm/yr, while sites N of the East Rift Zone (ERZ) were stable. Velocities decreased to the east and west of the central south flank such that the distal ends of both rift zones were stable, while the ERZ extended significantly. This deformation is believed to be driven by slip on a decollement at the interface between the sea floor and the volcanic pile, combined with dilation of the ERZ at depths of 3 to 9 km. GPS data shows that the south flank of Mauna Loa also undergoes SSE directed motion. Since 1983 Kilauea's caldera has subsided over 2 m, due to a combination of deflation of a shallow magma reservoir and rifting. The shallow reservoir, however, supplied only a small fraction of the magma to the eruption. Tilt signals indicated that the dike that initiated the Pu'u O'o eruption propagated at rate of 0.5 km/hr, consistent with the seismic swarm. Smaller intrusions and eruptions have occurred since 1983. Given that Pu'u O'o presents an open conduit to the surface, it is unlikely that magma pressures could build up to the level necessary to initiate dike propagation. Furthermore, these intrusions were not preceded by summit inflation. We believe that these intrusions are "passive", occurring because the ERZ is extended by south flank slip, causing the ERZ to fail in extension. Deformation during the 1997 ERZ intrusion/eruption was particularly well recorded by continuous GPS. The data show that dike propagation began 8 hours prior to the eruption, coincident with an increase in tremor. The inferred flux of melt into the growing dike indicates that the pressure at the dike inlet decreased as source reservoirs were depleted and/or as fluid flow into the dike failed to keep up with dike growth. The depth to the bottom of the dike is well constrained at 2-3 km, consistent with the inferred transition to steady rift opening at depth. The stability of Kilauea's north flank has been explained by the "buttressing effect" of Mauna Loa. The finding that the south flank of Mauna Loa exhibits SSE motions suggests a more general explanation is required. Other unexplained observations include the fact that the flank fails in large (M7) earthquakes but also slips steadily. Aftershocks following the M 7.2 1975 earthquake were largely restricted to the narrow band between the rift and the Pali, yet we know from the tsunami that slip occurred offshore. We propose that the mechanical properties of the decollement change as they are buried beneath the volcano. At its seaward end, the fault zone contains fluid saturated abyssal and volcaniclastic sediments, and is unlikely to experience stick-slip. As these sediments are buried beneath the volcano they are subject to elevated pressures and temperatures eventually becoming lithified and exhibiting velocity weakening and stick-slip. At greater temperatures near the volcanic axis the interface becomes steady-state velocity strengthening. Finally, north of Mauna Loa the fault zone pore-pressures may be sufficiently reduced, by a combination of fluid diffusion and metamorphic reactions, that the near horizontal interface can no longer slide under the ambient stresses. While, at this point qualitative, this model explains the asymmetry in the observed deformation, the distribution of seismicity, and presence of both stable and unstable sliding on the decollement.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2002
- Bibcode:
- 2002AGUFM.V62C..02S
- Keywords:
-
- 1242 Seismic deformations (7205);
- 8419 Eruption monitoring (7280);
- 8434 Magma migration