In order to study the behavior of atmospheric ozone in Antarctica, the correlation coefficients between total amounts of ozone and altitudes and between those and upper-air temperature at Syowa station for each month of spring and summer have been computed during the period 1982 to 1987. It is statistically shown that the characteristics of those monthly correlation coefficients are found for the months of October and December, and they are divided into three seasonal patterns. Two of the three patterns have characteristics at Syowa, while one is also seen in Japan.
Characteristics of foreshock activities before large shallow (M≥5, depth≤30km) earthquakes in and near the Japanese islands have been investigated. Of 110 main shocks which occurred during the period 1961 through 1988, 41, or 37 per cent, were preceded by foreshock sequences that included events within 30 days of the main shocks. The distance criterion for foreshocks is simply that it be within the 20′×20′ square centered on each main shock. The percentage of earthquakes with foreshocks scarcely changes if we take for criteria 40 days or 30′×30′ square. Last foreshocks most likely occur within a day before the main shocks in close proximity to their epicenters. In the period after 1981, foreshocks were observed for 55 per cent of large earthquakes, while the ratio was 27 per cent for the period 1961-1970, and 37 per cent for the period 1971-1980, which fact shows that hypocenter determination capability of the Japan Meteorological Agency has remarkably improved in recent years. Considerable regional characteristics exist in foreshock activity. One of them is that earthquakes in the Izu region and in Kyushu island are likely to be preceded by foreshocks. In particular, 67 per cent of earthquakes in and near the Izu Peninsula have foreshocks, as compared with only 26 per cent in other regions. Moreover, foreshock sequences in the Izu region and Kyushu island are often swarm-like in contrast to those in other regions which usually contain not more than a few shocks. In most cases the swarm-like foreshock activity in the Izu region becomes quiet two or three hours before the occurrence of the main shock. The phenomenon is considered to imply that stress relaxation occurs in the focal region due to nucleation of fracture just before the occurrence of a large earthquake. Regional characteristics of foreshock activities presented in this study are consistent with the results by Mogi (1963), in which foreshocks for moderate and large earthquakes in and around the Japanese islands during the period 1926 through 1961 were investigated, which fact indicates that the characteristics are intrinsic properties and stem from regional differences of crustal structure and stress accumulation system as pointed out early by Mogi (1963). Time sequences of foreshock activities are classified into 4 Types. Type 1 includes such cases that only one or a few earthquakes occur several to tens days before the main shock. Some events of this Type may not be directly connected to the main shock occurrence. In Type 2 cases one foreshock occurs as in Type 1, but immediately (usually within several minutes) before the main shock occurrence. Foreshocks of this Type may indicate the commencement of fracture which is succeeded by the occurrence of a main shock. Type 3 is represented typically by cases where a moderate earthquake with a magnitude of 4 or so occurs several hours to one day before the main shock, and its aftershocks, sometimes its foreshocks as well, are observed. This seismic activity associated with the moderate earthquake usually disappears a few hours before the main shock. Type 4 corresponds to swarm-like activity. Occurrence of Type 3 and Type 4 foreshock sequences is almost confined to the Izu region, the Fossa Magna area, and the central part of Kyushu island. The appearance of a quiescent period a few hours just before the occurrence of a main shock in Type 3 and in a part of Type 4 sequences, especially in the Izu region, is considered to be a prospective precursory phenomenon for the imminent prediction of large earthquakes.