基于EOF分解和KZ滤波的2019~2021年中国臭氧时空变化及驱动因素分析

Spatial-temporal Variation and Driving Factors of Ozone in China from 2019 to 2021 Based on EOF Technique and KZ Filter

基于我国337个地级行政区2019~2021年的3~8月O_(3)逐时浓度数据及同期气象数据,使用经验正交函数(EOF)分解,分析了我国O_(3)浓度的主要空间分布模态变化趋势及其主要气象驱动因素.选取31个省会城市,利用KZ滤波将O_(3)及气象因子的时间序列分解为短期分量、季节分量和长期分量,结合逐步回归模型,建立O_(3)与气象要素之间的关系,进而重构“气象调整”后O_(3)浓度的长期分量时间序列.结果表明,在我国O_(3)浓度变率的总体空间特征基本稳定的背景下,O_(3)浓度的第一模态整体呈趋同性变化,即O_(3)浓度变率的高值区域波动性减弱,低值区域波动性增强.气象调整前后不同城市O_(3)浓度的变化趋势存在一定差异,大部分城市O_(3)浓度经调整后的长期分量较调整前更“平缓”.其中,气象因素对石家庄、济南和广州等城市O_(3)浓度长期变化的影响较大,而前体物排放变化对福州、海口、长沙、太原、哈尔滨和乌鲁木齐等城市长期变化的贡献相对较大,北京、天津、长春和昆明O_(3)浓度长期变化趋势受排放和气象的影响均较大.

Based on the hourly O_(3) concentration data of 337 prefectural-level divisions and simultaneous surface meteorological data in China,we applied empirical orthogonal function(EOF)analysis to analyze the main spatial patterns,variation trends,and main meteorological driving factors of O_(3) concentration in China from March to August in 2019-2021.In this study,a KZ(Kolmogorov-Zurbenko)filter was used to decompose the time series of O_(3) concentration and simultaneous meteorological factors into corresponding short-term,seasonal,and long-term components in 31 provincial capitals.Then,the stepwise regression was used to establish the relationship between O_(3) and meteorological factors.Ultimately,the long-term component of O_(3) concentration after"meteorological adjustment"was reconstructed.The results indicated that the first spatial patterns of O_(3) concentration showed a convergent change,that is,the volatility of O_(3) concentration was weakened in the high-value region of variability and enhanced in the low-value region.Before and after the meteorological adjustment,the variation trend of O_(3) concentration in different cities was different to some extent.The adjusted curve was"flatter"in most cities.Among them,Fuzhou,Haikou,Changsha,Taiyuan,Harbin,and Urumqi were greatly affected by emissions.Shijiazhuang,Jinan,and Guangzhou were greatly affected by meteorological conditions.Beijing,Tianjin,Changchun,and Kunming were greatly affected by emissions and meteorological conditions.