Study On The Four Seasons Forecast Of PM₁₀ Pollution In Baoji City Based On Factorial Analysis

Taking the daily PM₁₀ concentration observed values from 4 sampling spots in Baoji City from Mar. 2003 to Feb. 2006 and the meteorological element observed values during the corresponding three years as study object, the paper studies the PM₁₀ concentration pollution forecast in Baoji City.Using air pollution index grading and its concentration limit as well as probability statistics, the paper analyzes the current PM₁₀ concentration pollution situation in Baoji City and the PM₁₀ concentration pollution characteristics. Analysis of the space-time characteristics of the PM₁₀ concentration in Baoji City shows that the PM₁₀ yearly average concentration in Baoji City has been declining from Mar. 2003 to Feb. 2006, with seasonal concentration were Winter>Spring>Summer>Autumn; the PM₁₀ concentration in heating season doubling the PM₁₀ concentration in non-heating season; the monthly changing trend in 4 sampling stations every year basically resembling the monthly changing trend in whole Baoji City every year, and the 4 sampling spots in Baoji City taking this order in changing: the sampling station>the 3^rd Army Hospital>Baoji Technical School>Zhuyuan Gou.Based on the meteorological element data of Baoji City from Mar. 2003 to Feb. 2006, the paper also analyzes 22 conventional meteorological element characteristics in order to find out the meteorological element changing characteristics in Baoji City and the law of it, so as to discuss their effects on PM₁₀ concentration, and then prepare for establishing the PM₁₀ concentration pollution forecast model.Applying factorial analysis, the paper respectively establishes forecast models for the seasons from Mar. 2003 to Feb. 2006. Extracted the factors from the four seasons: seven, eight, six and seven, and they explained the total variance: 85.779%, 86.986%, 88.693%, 85.389%. In general, the using of factorial analysis removes the collinearity so as to ensure the notability of forecast factors in forecast model; factorial analysis overcomes the analyzing difficulty arising from regression as a result of the highly pertinence among variables. The accuracy rates of forecast formula in forecast models established using factorial analysis are 81.70%, 81.48%, 78.19%, 75.67%, and the grading probabilty of hit is 100%.The paper still proposes advice concerning the current PM₁₀ concentration pollution situation in Baoji City, government policy making, engineering and technology, environmental quality management, economic policy making, afforestation of city and so on, emphasizing "preventive defense focused" regional integrated control strategy, and provides certain references and proof for the PM₁₀ concentration pollution integrated control in Baoji City.