Coda Envelope Moment Magnitudes for Improved Catalogs and Source Scaling in Central and Eastern North America

For hazard assessments and other research requiring accurate and consistent magnitudes direct determination of moment magnitude (MW) is preferable to relying on conversion relations. These may not have been validated across the complete range of magnitudes in an earthquake catalog and do not necessarily give the correct MW for any given earthquake. The coda envelope method (Mayeda et al., 2003) has emerged as a promising and stable technique for direct determination of MW for earthquakes too small for MW (~3.5-4.0) to be determined by waveform modeling methods. The scattered waves that comprise the coda are only minimally affected by source radiation pattern and directivity. Thus, the method provides stable results even with sparse or azimuthally unequal stations. Recent improvements to the Java-based software through collaboration between AFTAC and LLNL to facilitate the process make the coda envelope moment magnitude attractive for use as a standard magnitude. The method has been successfully applied to several regions of central and eastern North America among which are the southern Canadian shield and St. Lawrence corridor, Virginia and Kansas-Oklahoma, decreasing the MW threshold by a couple of orders of magnitude. The results highlight the need for region-specific calibrations due to attenuation differences. The results also showed some intriguing variations in apparent stresses between regions with unusually high stresses noted in the shield regions. In this study, we extend the application of the coda envelope method to areas of eastern North America not yet calibrated with priority given to the northern Appalachians, which link the previously studied regions. The results will be used to produce a uniform catalog for eastern North America for use in hazard and seismicity studies; current catalogs are maintained by a multitude of institutions using several magnitude scales and procedures for applying them. They will also lead to improved magnitude conversion relations and the development of source scaling relations across a broad region.