Abstract
According to the direction and distance from Mt. Sakurajima, thirty-nine local communities in the surrounding area of this volcano were classified into four regional groups, in which the week-to-week variations in mortality covering a 3-year period 1982 to 1984 were investigated in relation to volcanic air pollution. Mumber of cruption times of the volcano and atmospheric concentration of sulfur dioxide were used as the indices for correlation with volcanic air pollution.
In every regional group, the variations of weekly deaths showed a seasonality with the highest in winter and the lowest in summer seasons though there were some additional irregular short-term fluctuations. When the week-to-week variation in mortality was reviewed for the increase or decrease of deaths on the basis of consecutive two weeks, the number of deaths for the week following an abrupt rise in the atmospheric concentrations of sulfur dioxide was found to increase as compared with those for the preceding week without exception in the two regional groups within 20km north-west and within 30km south-east from the volcano's summit. Such increase was not evident for the other two regional groups more than 20 to 30km distant from the volcano. The weekly maximum of hourly sulfur dioxide concentrations at and above which significant association occurred with an increase in mortality was 0.2ppm. Since this measurement was obtained at a single point in the study area, however, it should be presumed that a relatively wide range of values around this level actually contributed to the increased deaths. For every regional group, there were no examples given where number of volcanic eruptions appeared to be related to increased mortality.
From the above, it is suggested that excess mortality as an “iceberg” of acute health effects can be detectable in large population, if the atmospheric concentrations of sulfur dioxide rise abruptly to the level at or above 0.2ppm in the presence of volcanic ash.
