Minimum Magnitude of Completeness and Rates of Seismicity in the New Madrid Region

Abstract

Abstract We apply the Schuster test to an augmented Central Mississippi Valley (CMV) seismic catalog to determine the minimum magnitude of completeness (MMC). We test the null hypothesis that the events in the catalog are distributed as a Poissonian process at a period of 24 hours. The test is based, in part, on the assumption that event detection is modulated by differences between day and night noise. We tested different temporal portions of the catalog corresponding to the whole catalog, 1974 – 1990; the period of network improvement in 1980 – 1987; and post-1987. The minimum magnitude of completeness between 1980 and 1987 is 1.8 within the region reported in the CMV Earthquake Bulletin. The catalog shows anomalous distribution after 1987, which may be related to a combination of mistaken identification of blasts as earthquakes and changes in network personnel. A significant (better than 95% confidence) increase in the rate of seismicity above magnitude 2.0 occurred in the central New Madrid seismic zone in 1981, followed by the Arkansas swarm less than a year later and by a significant rate increase in the enveloping region less than two years later. A significant rate decrease occurred in the whole region in June 1987, although the anomalous nature of the post-1987 catalog makes this change suspect. We suggest that the early rate increase is caused by a change in local (New Madrid) fault conditions, rather than a change in far-field conditions, because the rate increase appears first in the central region.