Earthquake recurrence intervals on geological timescales
Andy Nicol1,2, & John Walsh1 & Tom Manzocchi1
Fault Analysis Group, Dept Geology, University College Dublin, Dublin 4, Ireland
Institute of Geological and Nuclear Sciences, PO Box 30368, Lower Hutt, New Zealand
a.nicol@gns.cri.nz
Faults within the earth's upper crust generally grow and accumulate slip during earthquakes. In circumstances where faults remain active for millions of years they can accommodate hundreds of earthquakes. The rate of fault growth is dependent upon the amount of coseismic slip and the period of time between earthquakes, i.e. the recurrence interval. Fault slip rate, or displacement rate, is a measure of growth rate and is a function of fault size, with larger faults growing faster than small (Nicol et al. 1997; Wesnousky, 1999). Correlations between fault size and slip rate could arise from This relationship is in part produced by the proportional scaling of earthquake slip and rupture length, with longer faults accommodating greater coseismic slip (e.g., Wells and Coppersmith, 1994), or may demand a decrease in recurrence interval with fault size. The basic question arising from these observations is whether slower moving faults also accommodate fewer earthquakes and have longer recurrence intervals than faults which slip more rapidly?
To assess the impact of fault size and regional strain rates on the long-term stability of earthquake recurrence intervals we analyse published data for a range of fault sizes from four regions of crustal extension (Aegean, North Sea, Timor Sea, and Taupo Volcanic Zone). Slip rates, lengths and average recurrence intervals of 283 normal faults have been estimated for periods of time ranging from c. 20 kyr to 7 Myr. Although faults do not always rupture in their entirety and multiple faults can slip in a single earthquake, earthquake-scaling properties, such as the Gutenberg-Richter relationship, demonstrate that large earthquakes contribute disproportionately more to the strain accommodated within a fault system or along a fault. General conclusions can therefore be drawn on the characteristics of recurrence intervals and fault behaviour on the basis that each large earthquake on a fault Our analysis is based on the notion that each earthquake ruptures the entire fault length; apparently significant departures from this behaviour do not however impact the main conclusions of our analysis.
. Although faults do not always rupture in their entirety and multiple faults can slip in a single earthquake, earthquake-scaling properties, such as the Gutenberg-Richter relationship, demonstrate that large earthquakes contribute disproportionately more to the strain accommodated within a fault system or along a fault. Therefore, to a first-order this assumption is valid and permits general conclusions to be draw on the characteristics of recurrence intervals and fault behaviour.
We find that increases of fault displacement rates with fault size are accompanied by recurrence intervals which are, to a first approximation, constant. This suggests that higher displacement rates on larger faults are mainly achieved by greater slip per earthquake and not by more frequent earthquakes. Further analysis indicates that decreases in recurrence intervals between fault systems arise principally due to increases in regional strain rates.
References
Nicol, A., Walsh, J.J., Watterson, J., Underhill, J.R., 1997. Nature 390, 157-159.
Wells, D.L., Coppersmith, K.J., 1994. Bull. Seis. Soc. Am. 84, 974-1002.
Wesnousky, S.G., 1999. Bull. Seis. Soc. Am. 89, 1131-1137.