Moment rate scaling for earthquakes 3.3≤M≤5.3 with implications for stress drop
Ralph J. Archuleta, & Chen JiPublished December 1, 2016, SCEC Contribution #7201
We have determined a scalable apparent moment-rate function (aMRF) that correctly predicts the peak acceleration (PGA), peak velocity (PGV), local magnitude and the ratio of PGA/PGV for earthquakes . Using the NGA-West2 database for 3.3≤M≤7.7, we find a break in scaling of LogPGA and LogPGV vs M around M~5.3 with nearly linear scaling for LogPGA and LogPGV for 3.3≤M≤5. Temporal parameters td and tp–related to rise time and total duration–control the aMRF. Both scale with seismic moment. The Fourier amplitude spectrum of the aMRF has two corners between which the spectrum decays f^-1. Significant attenuation along the raypath results in a Brune-like spectrum with one corner fc. Assuming fc≈1/td, the aMRF predicts non-self-similar scaling Mo~fc^3.3 and weak stress drop scaling ≈Mo^0.091. This aMRF can explain why stress drop is different from the stress parameter used to predict high-frequency ground motion.
Citation
Archuleta, R. J., & Ji, C. (2016). Moment rate scaling for earthquakes 3.3≤M≤5.3 with implications for stress drop. Geophysical Research Letters, 43(23), 12,004-12,011. doi: 10.1002/2016GL071433.
Related Projects & Working Groups
High-frequency path and source parameters determined from recorded ground motion in central California, Ground Motion Prediction
