4.5 3-Week Postseismic + Coseismic Fit at the GPS Campaign Sites.
We also consider a second data set to further validate our conclusion that the coseismic slip distribution must be compact in the downdip direction, so that afterslip remains sufficiently far offshore. It is possible that the stress-driven afterslip could occur within the coseismic rupture area due to complex frictional properties, or a complex coseismic slip distribution Johnson et al., 2012, Avouac 201. Thus, we also use displacements from campaign GPS sites measured three weeks after the mainshock, which include coseismic slip plus three weeks of postseismic deformation. We consider the same range of coseismic slip models as in the previous sections.
For each coseismic slip model, we compute stress-driven afterslip and use the measured 3-week displacements from the continuous GPS sites to search for the best afterslip frictional parameterV0 , using the same method as for the 3-month case. Then we compare the coseismic displacements plus the predicted 3-week postseismic displacements to the observed coseismic + 3-week postseismic displacements at the GPS campaign sites. Viscoelastic relaxation is negligible over the first three weeks, based on the models already discussed above. Again, we find that a narrower fault of 110km~140km is preferred (light orange bars in Figure 8). This model also reduces the angular misfit at two island campaign sites with large displacements (YUK and SEMI) (Figure 12). This provides further support for our conclusion that the postseismic observations require a relatively compact rupture in the downdip direction.
(a) (b) (c)