Are the Long Recurrence Intervals of Small Repeating Earthquakes due to the Slow Slip Rates of Small Fault Strands?

Repeating earthquakes are earthquakes that repeatedly rupture a patch of a fault. One might expect the cumulative slip of these repeaters to be equal to the long term slip. However, observations on the San Andreas Fault imply that the cumulative repeating earthquake slip is smaller than the geologically and geodetically estimated slip. This difference in slip means earthquakes occur significantly less often than one would expect. The difference is especially large for smaller repeating earthquakes. Earthquake recurrence intervals are usually expected to scale as M₀^1/3 given slips estimated with standard rupture models, but observed recurrence intervals scale as M₀^1/6.

Here we assess whether the long recurrence intervals could arise because the total slip on the San Andreas accumulates on an array of large and small fault strands. Smaller earthquakes are more likely to occur on smaller fault strands, and those strands are likely to slip more slowly, potentially allowing for longer recurrence intervals. So here we seek to isolate and examine groups of repeating sequences that occur on the same fault strands, as we would expect the recurrence intervals of sequences on the same fault to scale "normally" with moment, as M₀^1/3, even though recurrence interval scales as M₀^1/6 for the general population of repeating sequences.

We use the Parkfield-based repeater catalogue of Lengliné et al. (2009) to identify closely spaced repeating sequences which could occur on the same fault strand, and analyse how their recurrence intervals vary with moment. In addition, we analyse the locations and timing of earthquakes in the double-difference NCSN catalogue (Waldhauser, 2009). We show that we can often estimate the recurrence intervals of repeating earthquake pairs by extracting the time from an earthquake to the next similarly sized nearby earthquake. Then we examine how recurrence interval varies with moment for closely and widely spaced repeating earthquake pairs.