Abstract
We have compiled source-time functions (STFs) of 54 Mexican earthquakes (3.0 ≤ M ≤ 8). The depth, H, of 48 of these earthquakes lies between 10 and 50 km. The STFs of 43 earthquakes were determined from local and regional strong-motion and/or broadband recordings. For the rest of the events, the STFs were taken from published analysis of teleseismic body waves, or from the University of Michigan Source-Time Function catalog. If available, the seismic moments were taken from the analysis of teleseismic body waves; if not, as was often the case for earthquakes with M ≤ 5.5, they were estimated from the local recordings themselves. From the STF of each earthquake we have measured pulse duration of each subevent (
p), total rupture duration (t), and rise time (r), along with their respective seismic moments. The data fit the relation M0/3p ~ 6.7 x 1023 dyne cm/s3. This relation is similar to that reported for deeper earthquakes (mostly in the 50 to 125 km depth range) in the Kanto district of Japan. The pulse duration of Mexican and Kanto earthquakes, scaled to a seismic moment of 1026 dyne cm, ps, is about 5 s, nearly independent of depth for 10 ≤ H ≤ 125 km. We compliment these two data sets with those for California (H ≤ 20 km) and for deep earthquakes (H ≥ 100 km). When viewed over the entire depth range, the total rupture duration (t) is related to M0 by log t = (0.363 +/- 0.008) log M0 - 8.58 (+/-0.190), and the duration, within the scatter of the data, appears to be independent of depth. When the data are examined in smaller subsets, as has been done by several authors, certain trends manifest themselves. For example, the pulse and the total duration of deep earthquakes (H ≥ 100 km) scales as M0.160 and M0.250, respectively, and the moment-scaled pulse and total duration decreases by about 20% between 100 and 600 km depth. This decrease is much less than that reported previously. Furthermore, an earlier finding of fast rise times for very deep earthquakes is not substantiated by the recent data. Thus, the anomaly in the duration of the deep earthquakes, as compared to the shallow ones, appears to be smaller than previously suspected.