Influencing Factors Determination of MSW clearance volume Based on Spatial Dependency Consideration

摘要

Previous studies usually used regression analysis on OLS model to determine the influencing factors of MSW clearance volume, but OLS model did not take into account the spatial dependency of the dependent variable. In this study, firstly, as the dependent variable, the spatial autocorrelation of MSW clearance volume is tested to explore its spatial effect, and spatial regression model is introduced to establish a MSW-SEM (spatial error model) model. Secondly, variance inflation factor and partial correlation coefficient are used to remove the multicollinearity of 33 potential factors. Then using MSW-OLS model and MSW-SEM model respectively analyzes the influencing factors of MSW clearance volume in China mainland. Finally, comparative analysis of above two models' results is taken by Akaike Information Criterion, determination coefficient, significance of parameter estimation, and spatial dependence of residuals. Result shows that community health center visits, accommodation enterprises main business profit, passenger capacity, investment enterprises profit rate of Hong Kong/Macao/Taiwan/overseas and urban road lighting were dominant influencing factors of MSW clearance volume. Although MSW-OLS and MSW-SEM models have similar regression coefficients, MSW-SEM model shows better model fitting because of its lower AIC, higher R~2, and smaller global Moran's I value. In summary, compared with MSW-OLS model, MSW-SEM model is more successful in identifying the dominant factors of MSW clearance volume due to its consideration of spatial dependency.