WP 6.1 Earthquake probability changes due to stress transfer

Start date or starting event:

M0

Lead contractor:

GFZ POTSDAM

Participants:                                                GFZ POTSDAM, DF.UNIBO, CNRS-UMR 5562

1

Objectives:

Improving probabilistic earthquake hazard assessment through stress field models for media with both elastic and inelastic layers.

Achieving a time-dependent hazard analysis by deducing changes in the probability of future earthquakes due to stress transfer.

2

Inputs:

·         The historical seismicity in the region are needed and exist,

·         Data on the aftershock activity of recent strong events are needed and available too,

        the same is true for the deformation fields obtained from Continuous GPS and INSAR.

 

Methodology / work description:

The models of stress changes due to strong earthquakes in the South Iceland seismic zone (SISZ) produced in the framework of the EU-funded projects PRENLAB-1 and PRENLAB-2  will be extended to include inelastic behaviour. The stresses will be calculated with the elasticity theory of dislocations for a layered elastic/inelastic medium using existing programs. The algorithms were fully redesigned  and then extended to inelastic medium properties for layers below the seismogenic layer. By this, now changes in the stress field due to plate motion, from seismic events, from stress relaxation and aseismic slip/creep can be taken into account.

 

Moreover, we will now consider - besides the variations in the shear stress distribution - also the changes in the Coulomb failure stress (CFS), as these have shown to correlate well with aftershock distributions and - in some cases - with stress transfer in series of strong events. Coulomb stresses will be determined expanding a fully elastic approach to inelastic constitutive laws using own software.

 

Both, the shear stress and the Coulomb stress variations, will be converted into an increase/decrease of changes in the occurrence probability of future earthquakes (a permanent offset plus a temporal increase for the aftershock period). Time-dependent earthquake probabilities on fault segments near those that ruptured recently (e.g. in June 2000) will be calculated by two methods: (i) the stationary Poisson model and (ii) the conditional probability model. Permanent and transient effects - determined via the rate and state dependent fault property model  - of stress changes are taken into consideration. The recurrence rate of earthquakes in the SISZ determined via the long-term moment release will be used to check whether the historical events show stress triggering. For the present situation (time period of the project), the probability of a strong event in parts of the SISZ that did not rupture for a long time is recalculated using the stress transfer method.

 

Results of the research activities in PRENLAB and PRENLAB-2 are now extended and used to provide an important contribution to an early warning and information system.

Close cooperation will be with WP 6.2 and with WP 2.3, 2.4, 2.5, 4.4 and 4.5.

 

3

Deliverables including cost of deliverable as percentage of total cost of the proposed project:

D96

Inelastic model for the earthquake series (M>=6) in the SISZ since 1706.  

M12  Re  PU  2,5%

D97

Article and report: Probability increase of each of these 13 events compared to the model. 

M24  Re  PU  2,5%

                                                                                                  

4

Milestones: Delivery of the above items at the date indicated.