← Back to Sessions

Random and Systematic Uncertainties in EGF Spectral Ratio Analysis and Their Implications for Source Scaling

Description: 

Most estimates of earthquake stress drop, and any resulting magnitude dependence are made by modeling amplitude spectra. Trade-off between high-frequency attenuation (path and site) and source shape leads to systematic differences in estimates of spectral stress drop for small earthquakes, leading to different scaling relationships between with magnitude or depth. Spectral ratio, empirical Green’s function (EGF), techniques in which the amplitude spectrum of a larger earthquake is divided by that of a closely located smaller earthquake to obtain estimates of their corner frequencies, moment ratio, high-frequency asymptote, and hence spectral stress drop, are widely used. In theory, these techniques remove propagation effects without requiring a reference site effect or source spectral shape inherent in generalized inversion and spectral decomposition techniques. In practice, source model assumption, and data quality and availability have all limited the resolution and reliability of results. Some studies treat both corner frequencies and spectral stress drop as equally well resolved, and others do not, leading to different conclusions concerning earthquake source scaling. The well-studied data set of the Ridgecrest Community Stress Drop Validation Study provides an opportunity to identify whether there is systematic bias or large random uncertainties in the small event corner frequency. We select closely-located clusters of earthquakes that have a wide enough magnitude range for some events to form both the numerator and denominator in different spectral ratios. Initial results show that the corner frequency obtained for an individual event is systematically lower when it is the denominator, compared to the numerator, in a spectral ratio. We compare three independent approaches to spectral ratio analysis, varying time window selection, wave-type and assumed source model to determine under what conditions the estimates of spectral stress drop for both earthquakes in a ratio are reliable and comparable, and what causes systematic differences observed in preliminary tests.

Session: Advances in Reliable Earthquake Source Parameter Estimation - III

Type: Oral

Date: 4/16/2025

Presentation Time: 03:00 PM (local time)

Presenting Author: Rachel

Student Presenter: No

Invited Presentation: Yes

Poster Number:

Authors