Abstract
The relationship between the days of earthquake in Japan and its vicinity and the Be-7 concentrations in the surface air at Takasaki were statistically analyzed. A day when one or more earthquakes of M 5 or greater occurred in Japan and its vicinity was defined as a “day of earthquake”. Earthquakes which were deemed aftershocks were not included in the analysis. In order to avoid the effect of seasonal variation, the residuals obtained by deducting 31-day moving averages from Be-7 concentrations and those obtained by deducting the ‘21-day’ moving averages of the periods from −15 to +0 days and from +11 to +15 days were analyzed. The ‘21-day’ moving averages were also calculated to avoid the effect of aftershocks. Rank-sum tests of the residuals showed that the Be-7 concentrations decreased from the 31-day and ‘21-day’ moving averages on days of earthquake. A binomial test showed that there were significantly more days of earthquake when the Be-7 concentrations were smaller than the medians of Be-7 concentrations for the 31-day periods (from 15 days before to 15 days after days of earthquake) than those when the Be-7 concentrations exceeded the medians. These suggest that the Be-7 concentrations decrease on days of earthquake. It is noted that the Be-7 concentrations on days of earthquake decreased on days of both precipitation and non-precipitation, respectively.
