ABSTRACT

A statistical method, named the Geochemical Filtering Method (GCFM) of earthquakes, has been proposed to identify seismogeochemical anomalies in the record of subsurface gas CH₄/Ar ratios at two deep borehole wells in Matsuyama, Japan. This is based on the statistical reasoning that the absence of seismogeochemical anomaly can confidently be shown only by the observational data. The earthquakes occurring in a period when geochemical monitoring does not indicate any significant anomalous changes, are very important to infer what kind of earthquakes does not accompany any seismogeochemical anomalies. This provides us a non-subjective criterion by which the earthquakes unlikely to accompany geochemical anomalies can be selected. The criterion expressed in terms of magnitude and epicentral distance for earthquakes makes it much easier to demonstrate the physical correlation between particular earthquakes and unusual changes in geochemical records. The criterion is thought to represent a threshold relation concerning the appearance of seismogeochemical anomalies. The threshold relation for CH₄/Ar anomalies in Matsuyama is found to be similar to those for Rn and geoelectric resistivity anomalies reported in the literature. The GCFM is applicable to other seismogeochemical monitoring data at any stations, and very useful in evaluating the reliability of the geochemical monitoring data.