The reliability of pollution prediction with regression analysis and the possibility of dispersion and receptor models

This paper presents the results of regression analysis to predict the pollution degree of outdoor insulator with the meteorological conditions, and the proposal to improve the reliability of the pollution prediction technique, applying the atmospheric dispersion and receptor model. Through the preceding study, we had acquired close relations, r over 0.9, between the pollution degree and the conditions. But the old method, multiple regression analysis, has many problems, such as needs measuring data for each sites, don´t consider the geographical condition, etc. So, in this study, we integrate the old research results of us to confirm the reliability of the old method, and through the consideration on the integrated results, proposed and demonstrated the atmospheric dispersion and receptor model to overcome the shortcoming of the old method.