Figure 4
Swapping of the location of fault activity/strain accumulation across the strike of the normal fault system. We concentrate on the time since ~ 15 ka because slip is recorded at all 3 sites during this time period. (a) Slip-histories inferred from 36Cl data interpreted to show the approximate durations of the principal clusters. (b) Locations of the faults relative to the 025A-007A GPS transect. (c) Cross-section showing the elevations and preservation ages of each site and geomorphic and geological offsets. (d) Strain rates implied by high slip-rate periods; strain-rates in each cluster are calculated over the interpreted duration of the cluster, as shown in (a); uncertainty for each strain-rate value is calculated assuming a cluster duration of 2000 kyears and 4000 kyears. We reject the hypothesis that each of the seven principal faults slip at a constant rate taking up 1/7 (~ 14%) of the regional strain-rate because the required clustered durations are not consistent with the inferred least squares slip histories. We interpret the existence of earthquake clusters that are out-of-phase on faults located across from each other suggesting that only a few faults are active at any one time on a millennial timescale and fault activity swaps across strike to maintain the regional strain-rate.
