FACTOR ANALYSIS ON SO2 CONCENTRATION IN RELATION TO CLIMATIC CONDITION

Abstract

The objective of this study is to clarify some of principal factors which influence the variations and minute regional differences in the concentrations of air pollutants. The “Factor Analysis” method, which has been developed in psychology, was applied to the data on sulphur dioxide concentrations at nine stations in and around Sendai (Fig. 1) during the period from December, 1972 to February, 1973.First, mean and maximum daily concentrations were analyzed, and the result was con-firmed with hourly measurements at three hour intervals. Two main factors were extracted by the computation (Tables 1 and 3). Then, the author attempted to examine the meanings of the extracted factors by means of the distribution of factor loadings (Figs. 4, 5, and 7), the time variation of factor scores (Figs. S and 9), and the correlations between factor scores and some of the meteorological elements (Tables 2 and 4). The meteorological data used for this purpose were as wind direction, wind speed, vertical gradient of temperature, solar radiation, and precipitation observed at the Sendai District Meteorological Obse-rvatory and the Yagiyama station. The results obtained are as follows; 1) The first factor accounts for a large part of the variations in concentrations. Its loadings are plus everywhere, and the scores for them have negative correlation coefficients to wind speed and vertical gradient of temperature. These negative correlations indicate that the scores are higher under a calm and stable atmospheric condition, that is, the concentrations can be expected to be higher all over. Sendai under such a condition. Therefore, the first factor, the air pollution potential, is regarded as a condition which stagnates the pollutants over the large area. 2) The second factor mainly accounts for the local variations in the built-up area where medium and small emission sources are concentrated. The factor scores have positive correlation coefficients to wind speed and lapse rate. But, their values are generally lower on Sundays and higher on weekdays when a windy and unstable atmospheric con-dition; therefore, the second factor is considered to be conditioning wind-blown pollution. 3) Loadings of the second factor in the built-up area exceed those of the first factor as far as daily maximum value is concerned ; that is, higher concentration in the built-up area is closely related to the second factor. This indicates that, at any place in the vicnity of emission sources, higher concentrations tend to be accompanied more frequently by high wind speed and unstable lapse rates rather than by stable and calm atmospheric conditon.