A study of the homogeneity of the NOAA earthquake data file in the Mid-America region by the magnitude signature technique

Abstract

The Earthquake Data File (EDF) compiled by the National Oceanic and Atmospheric Administration (NOAA) of the U.S. contains information that can be of great importance when studying seismicity of a certain region of the world. However, the homogeneity of this catalog, which comprises mostly data gathered by PDE (Preliminary Determination of Epicenters), depends on the characteristics of the stations and agencies which provided the data in the first place. In this study, the homogeneity of the NOAA catalog, with specific focus on the Mid-America region, is investigated by means of a t~chnique recently proposed which simulates changes in the events reported with respect to time. These changes may account for the observed variations in number and characteristics of the events listed in the catalog. The method is briefly described. Times of occurrence of such changes, as given by the magnitude signature method, are determined as well as their possible causes, and proposed corrections are given. The times for which changes were found to be statistically significant are: end of September 1965; mid-October 1967; early October 1969, end of May 1972 and end of July 1979. The corrections proposed for events in the Mid-America region listed in the NOAA catalog are: events from July 1964 to September 1965 should be shifted -0.1 magnitude (m_b) units; events from September 1965 to May 1972 should be changed by+ 0.1 magnitude units and events after May 1972 ought to be left without change. With this results the probable inclusion of artificial changes in seismicity that could otherwise be taken as real seismicity variations can hopefully be avoided. Thus, the study of seismicity in the Mid-America region for earthquake prediction purposes can be undertaken with more accurate results.