TRI-MODEL DIAGRAMS FOR DEPICTING RELATIONS OF MEASURABLE ATMOSPHERIC ION DENSITIES AND MAGNITUDES OF SUBSEQUENT EARTHQUAKE

K. WADATSUMI, R. Haraguchi, K. Okamoto (Faculty of Informatics, Okayama University of Science)

The Western Tottori Prefecture Earthquake in Japan (WTPE; Mj=7.3) on October 6, 2000 occurred at 80 km distance northward from Okayama University of Science(OUS). Since July of 1997, the atmospheric ion density has been continuously measured at our laboratory in OUS in order to evaluate the relation between the ion densities and magnitudes of subsequent earthquakes. Usually, we have had the ion densities ranging between 500 to 2,000 ion/cc as a background through a year. Prior to earthquakes less than Mj=5, the density peaks of positive charged larger ion are used to take the values of less than 5,000 ion/cc before several days of an earthquake event. In this case, it is enough to use a diagram with vertical axis scaled to 5,000 ion/cc and a lateral time axis scaled to a week. This diagram is used to upload daily on an "e-pisco" website (http://www.pisco.ous.ac.jp/). In the case of WTPE on October 6, 2000, it was difficult to manage this weekly diagram and the 5,000 ion/cc upper limit. Because of the reasons that the first appearance of abnormal density had been recorded 11,105 ion/cc on June 22 just 100 days prior to WTPE and the 13 abnormal density peaks (9,774 to 3,000) ion/cc only in July, comparing with 15 peaks in other months of a year. For large earthquake, it required to represent the information in a special diagram that consists of a vertical axis of ion density 1,000-16,000 ion/cc and 3.00-4.20 on the log scale, using a lateral time scale of 12 months in a year. Each of the peaks of WTPE fitting on a curve is represented nearly as a straight line on a log diagram and the lower end of the descending line seems to cross the time axis at the time that WTPE occurred. The results of the ion densities of earthquakes tested during past three years (including that of WTPE) show a clear correlation between the ion densities and earthquake magnitudes. The peaks of Mj4 earthquake takes a position of 3.30 to 3.50 on the log scale, Mj5 of 3.50 to 3.70, Mj6 of 3.70 to 3.90, Mj7 of 3.90 to 4.10, and Mj8 of 4.10 and greater. In order to apply to all earthquakes, the data must be depicted in on tri-models diagrams. The first model represented in the diagram on a lateral time scale of 12 months during the year for the biggest earthquake, 30 days in a month for big and moderate earthquakes, and 7 days in a week for moderate and smaller one. This method is acceptable within a 300 km radius of OUS monitoring location. The ion density measured is commonly referred to as the natural density. Using with an empirical equation to get an estimated density at the location far from an epicenter, the normalized standard density can be calculated at the point of 100 km. To identify the fore-shock in the case of WTPE, we recognized with the second peak of abnormal density 9,774 ion/cc in the afternoon on 11th of July, 2000. In this critical condition, the moderate size of earthquake (Mj4.3, 8:00 July 17) at the western Tottori prefecture had come, indicating that this should be a fore-shock with high probability such as a forthcoming big earthquake (Mj=6.5) under abnormal condition of the ion densities. This consideration was reported on our Web site at 19:00 on July 17. We had downloaded data of micro-earthquakes from the database of Japan Meteorological Agency and found the characteristic changes of time and space distribution of micro-earthquakes before WTPE, which were well correlated with a variety of the ion densities.