Abstract
In this paper, the factor analysis by the linear multiple regression model was carried out using the home CO2 statistical survey and the heating degreeday. Factor analysis was conducted in two models, nationwide model and regional model. Furthermore, by using parameters obtained by factor analysis and municipal statistics, CO2 emissions per household in the household sector were estimated by municipalities across the country.For the nationwide model, the regional model has a multiple correlation coefficient improving accuracy of about 1 to 3 points, and in the four regional categories of Hokkaido, Tohoku, Kanto Koshin, and Kyushu, the heating degrea in the district It was revealed that there was a big difference and a significant difference in CO2 emissions. From the estimation result of CO2 emissions per household by municipality using regional model, it was clarified that the quartile deviation in the four regional categories of Tohoku, Kanto, Hokuriku, and Kinki is 0.4 or more. The difference in CO2 emissions per household among municipalities among the four regions was shown to be particularly large.
