A physical interpretation of field observations that precede large earthquakes
Abstract
A cellular automaton model of earthquake faulting adopting Coulomb's failure criterion developed by Sacks and Rydelek (1995) successfully generates catalogs that satisfy Gutenberg-Richter's Law, the observed decreases in b-value before large events, as well as the propagation of the rupture front. Model runs indicate that redistributed stresses remain on the ruptured area and that some slips recur on the same cells forming dynamic asperities of high slips. We found that the observed magnitude-dependent seismicity quiescence can be explained by the introduction of dilatancy hardening into the model. Only a few % of the total number of model cells need be strengthened by a small amount. This indicates the difficulty of detecting their presence using seismic imaging. However, the observed long term ( years) temporal changes in seismicity, gravity, and electrical resistivity may be causally linked to the volume change from microfractures and the effect of pore pressure changes on fault strength. Our model predicts the process occurs at points sparcely distributed. Water migrations into unfilled microfractures act to lower the strength, thus promoting the occurrence of seismic slips. These slips may expel water that will influence aquifer levels, which may be observed at regional water wells. Drilling in seismic fault zones, such as at the 1995 Kobe earthquake fault, has revealed that the permeability on the main fault plane was many orders of magnitude higher than the surrounding rocks. We suggest the same water migration process at highly permeable zone can occur at short time scale to grow into a large magnitude slip or may manifest as a slow slip. The aftershock sequence of the 1978 Izu-Oshima earthquake shows that it overlaps the inferred slow slip on the fault following the main shock, thus suggesting that a fault can slip in various ways in the same time interval. We propose new observations that are sensitive to crustal water migration such as vertical-component strainmeters to augment existing monitoring networks where large earthquakes occurred in the past.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2016
- Bibcode:
- 2016AGUFMNH51C1964S
- Keywords:
-
- 0619 Electromagnetic theory;
- ELECTROMAGNETICSDE: 2427 Ionosphere/atmosphere interactions;
- IONOSPHEREDE: 4333 Disaster risk analysis and assessment;
- NATURAL HAZARDSDE: 7223 Earthquake interaction;
- forecasting;
- and prediction;
- SEISMOLOGY