Landslides in the Sydney Basin - Geoscience Australia

Seismic Hazard in SydneyProceedings of the one day workshopFirstly, we use the expression relating seismic and geodetic moments:M o = µ A D (14)(Aki & Richards, 1980) where µ is the crustal rigidity (3.0×10 11 dyne/cm 2 ), A is the rupture area (inour case the fault length, L, times the width, W), and D is the average subsurface slip (in our case therecorded fault throw, although this is a minimum — see below). The results are shown in Table 4(L . W . D).Table 4 Proportion of moment release rate on dip-slip structuresDIP-SLIP FAULTS AS INDIVIDUAL FAULTS DIP-SLIP FAULTS IN ZONESL W D 12% 18%L 2 D 7% 13%L 3 16% 12%In the above (L . W . D) treatment we have assumed a fault width of 18 km, i.e. steeply dipping faultsgoing through the full seismogenic crust. An equally valid argument is to say that the mapped faultswill have an aspect ratio of 1:1 (until L>18 km), and so we calculate the proportion of momentrelease accordingly. This is our second way of apportioning moment using the mapped structuresand the results are also shown in Table 4 (L 2 D). The values obtained are somewhat less than for thefirst (L . W . D) treatment.A shortcoming of both the methods presented thus far is that the throws recorded for strike-slipfaults in the coalfield data are just the vertical throw and do not represent the total offset. This mayintroduce a bias towards more moment release on the dip-slip structures. Our third approach is toassume self-similar scaling and to calculate the moment release rate according to L 3 , thus avoidingthe use of measurements of fault throw. The results (bottom row on Table 4) are comparable withthe other approaches. All of the values in Table 4 have been used in probabilistic calculations.Given that all the approaches we have used so far have some limitations, we also use more genericearthquake scaling properties to estimate moment release. In general, moment release in a region isdominated by the largest events. The moment release associated with the maximum magnitudeearthquakes for the strike-slip and dip-slip structures can thus be used to estimate the proportion ofmoment release rate on dip-slip structures alone. We have already mentioned that the northwestsoutheastoriented strike-slip faults are the longest structures in the region. We thus assume that theM max already estimated for the region (M w 7.5) will occur on such a structure.To estimate the M max for the dip-slip structures we use the logic already developed on the otherbranch of the logic tree for the maximum likely rupture lengths (10 and 18 km), and use that toestimate M max . The result is M w 6.81±0.12, giving proportions of moment release of 8% and 6%based on the ratio of mean and mean-plus-2-sigma M max values respectively. We have used both inthe probabilistic calculations.Thus far we have obtained a range of 6–18% for the proportion of total moment released on the dipslipstructures, and conversely 82-94% on the strike-slip faults. Other information of relevanceincludes earthquake focal mechanisms. Studies by Denham (1980) and Denham et al. (1981) showa mixture of strike-slip and reverse focal mechanisms, in good agreement with our assumed modelof a mixture of active strike-slip and dip-slip structures. Most of those focal mechanisms are,however, closer to the continental interior than our study area. While these data reinforce thesuitability of our model, they are not sufficient to derive any accurate partitioning of moment releaserate between strike-slip and dip-slip fault motion. This is because there are (i) too few events in an