EARTHQUAKE SCALING RELATIONS FOR MID-OCEAN RIDGE TRANSFORM FAULTS
BOETTCHER, M.S., MIT/WHOI Joint Program, Woods Hole Oceanographic Institution, Woods Hole, MA, 02543, email@example.com; and JORDAN, T.H., Dept. of Earth Sciences, University of Southern California, Los Angeles, 90089-0740, firstname.lastname@example.org.
The seismicity of a mid-ocean ridge transform fault (RTF) of length L, slip rate V, and moment release rate dM/dt can be characterized by a seismic coupling coefficient chi = AE/AT, where AE~ (dM/dt)/V is an effective seismic area and AT ~ L3/2V-1/2 is its thermal area of contact above some reference isotherm Tref. We show that the seismicity data from a global set of 65 RTFs with a combined length of 16,410 km can be well described by a linear thermal scaling relation (1) AE ~ AT. Assuming Tref = 600oC, we obtain chi = 0.26 plus or minus 0.05. We conclude that nearly 3/4 of the slip above the 600oC isotherm must be accommodated by subseismic mechanisms, i.e. steady aseismic creep, silent earthquakes, and infraseismic (quiet) events, and that this slip partitioning does not depend systematically on either V or the maximum age of the lithosphere in contact across the fault (1 My to 45 My in our data set). Above the threshold moment for catalog completeness, M0, the observed frequency-moment distributions of RTF seismicity can be fit by a truncated Gutenberg-Richter model with a slope beta = 2/3 in which the cumulative number of events N0 and the upper-cutoff moment MC vary systematically with thermal area. Data for the largest events are consistent with a self-similar slip scaling, DC ~AC1/2, and an areal scaling relation of the form (2) AC ~ AT1/2. If (1) and (2) apply, then moment balance requires that the dimensionless seismic productivity, nu0 ~ N0/AT V, should scale as (3) nu0 ~ AT-1/4, which we confirm with the data for small events. Thus, larger transform faults on average have bigger earthquakes but smaller seismic productivities. We conclude that the frequencies of both small and large earthquakes adjust with the thermal area in a way that maintains a constant coupling coefficient chi. The RTF scaling relations (1) & (2) appear to be inconsistent with the bimodal hypothesis, which states that a fault patch is either fully seismic or fully aseismic and thus implies AC less than or equal to AE. Non-bimodal behavior is consistent with the low aftershock productivity of RTF earthquakes, which exhibit very small ETAS branching ratios (10-2). We suggest that the heterogeneities in the stress distribution and fault structure responsible for the peculiar square-root scaling (2) arise from a thermally regulated, dynamic balance between the growth and coalescence of fault segments within a rapidly evolving, three-dimensional fault zone dominated by serpentinite and other hydrated mineralogies.