城市间大气污染相互影响的VAR模型分析

基于以往研究多采用大气数值模式分析污染物在不同城市之间的输送规律及内外源的贡献率,现从时间序列角度引入一个新方法,应用向量自回归模型的格兰杰因果关系检验、脉冲响应函数及方差分解技术,对广州、佛山、肇庆2003-2012年逐日空气污染指数进行模拟演算,得到城市间大气污染相互影响效应及其贡献。结果表明,广佛地区污染源对肇庆空气质量影响突出,肇庆并未成为广州和佛山的主要污染源。城市间大气污染相互影响存在明显的滞后效应,前7期累积作用较明显,污染物的累积效应容易导致区域性灰霾天气出现。佛山对广州的污染贡献达到了10.9%,广州对佛山的污染贡献相对偏小,佛山对肇庆的污染贡献接近30%,广佛肇经济圈应形成区域性大气污染联防联治机制。

Vector Autoregression Model for Analysis of Air Pollution Interactions between Cities

Based on the use of atmospheric models for studying transportation of pollutants between cities as well as the contributions of inner and outer sources,from perspective of time series,Granger causality test,impulse response function and variance decomposition of vector autoregression model was introduced for studying daily air pollution index in Guangzhou,Foshan and Zhaoqing from 2003 to 2012.The air pollution interactions between cities were proved.It was found that air quality of Zhaoqing was heavily affected by pollutants from Foshan and Guangzhou while Zhaoqing was not a major pollution source of Guangzhou and Foshan.Regional haze was caused by the lagged and cumulative effects of pollutants,obvious impact was appeared before lagged 7.The contribution from Foshan to Guangzhou was 10.9% while it was lower from Guangzhou to Foshan.The contribution from Foshan to Zhaoqing was nearly up to 30%.It was necessary to establish regional air pollution prevention mechanism in the economic circle of Guangzhou,Foshan and Zhaoqing.