Mw vs. Ml Evaluation for Small Earthquakes in Oklahoma and Kansas

Session: Earthquake Source Parameters: Theory, Observations and Interpretations [Poster]
Type: Poster
Date: 4/30/2020
Time: 08:00 AM
Room: Ballroom
Description: 

Probabilistic seismic hazard analysis requires recurrence rates with respect to a consistent magnitude scale, preferably Mw. Low signal levels prevent the use of regional moment tensor inversion techniques to estimate Mw below about 3 to 3.5, thus affecting catalog completeness. However, with an improved velocity model, waveform inversion is possible using observations within 50 km at higher frequencies.

We focused on five source zones in the study area where broadband instruments were deployed following a mainshock. Given the lack of open information on shallow velocity structure, we carefully defined each local upper-crust velocity model to satisfy teleseismic P-wave receiver functions, regional surface-wave tomography, local phase velocities obtained using p-omega stacking and regional earthquake P-wave receiver functions. These models were tested by comparing the model predictions to the observed pulse-like reverberations in SH signals. Finally, moment tensor inversion was performed to fit the P and SH pulses observed at short distances. The result was to create a data set with Mw as low as 2.0.

Our determination of ML corrects for differences in regional attenuation with the assumption that similar ground motions at short distances in Southern California and in the study regions must have the same ML, hence the definition ML = log A(nm) -1.11 log R – K R -2.09, where the K is locally determined. This approach does not require the arbitrary assumption that ML = Mw.

A preliminary relationship of Mw ~ 2/3ML+1.0 between the local Mw and ML was determined with the ~2/3 slope matching expectation from source scaling. The short distance Mw vs ML population overlaps the regional determinations near ML= 3.3.

Presenting Author: Quentin Seydoux

Authors