Simulation of Seismicity in the Japanese Islands with an Interacting Fault System Model Derived from Geodetic Data

Abundant geodetic data enabled us to estimate fault deficit rates in and around the Japanese Islands. Hashimoto and Jackson (1993) first estimated fault slip deficit rates using triangulation and trilateration data with a block and fault model. Here we present results of a simulation of seismicity using this fault model, in order to investigate correlations between interplate and intraplate events, long range correlations of seismicity, and discrepancies in strain rates derived from geodesy, geology and historical seismicity. Hashimoto and Jackson's original model has 104 faults. However, the original fault sizes are too large to accommodate small earthquakes, so we divide each modeled fault into 5 x 5 elements in order to be able to generate earthquakes as small as M5. In total there are 2600 elements. We assume stress accumulates according to the estimated slip deficit rates. In our model we deal with only Δ CFF, among the stress components that promote rupture consistent with the estimated slip deficit rates. Interaction between fault elements is represented by Δ CFF. When stress reaches a threshold level, accumulated stress is released by forward slip and redistributed to surrounding faults according to their geometrical relationship. The thresholds for rupture are assumed to be 0.5 MPa for interplate faults and 2.0 MPa for inland faults, respectively, and are uniform on each fault. These values are consistent with the typically observed slips of earthquakes for the model faults with their original size. If redistribution of stress causes the next rupture, the same process repeats. In this simulation, the time step is 1 year. Due to the computer limitations, we could calculate only for 5,000 years. The rigidity is 40GPa and friction coefficient varied from 0 to 0.7. Results of the simulation show that large events in which entire faults rupture rarely occur. This suggests that strain rates from geology or historical seismicity might be underestimated. However, Gutenberg-Richter law does not hold in this model, which suggests that we need heterogeneity in the rupture threshold or the size of elements on a fault plane. Seismicity of the model is much higher than the rate of earthquakes in Japan. This may be attributed to the use of slip deficit rates derived from geodetic data and suggests that geodetic strain is not entirely released by earthquakes. Correlation between interplate and inland events is not clear, however migration of events such as those observed along the Nankai trough can be seen in this simulation.