• 关键词：判别方法  霾日  时空分布  小波变化

• 基金项目：2015年中国气象局气候变化专项（No.CCSF201532）；国家科技支撑计划项目（No.2014BAC16B03）；2015年四川省气象探测数据中心课题

• 孙彧
• 1. 四川省气象探测数据中心, 成都 610072;2. 高原与盆地暴雨旱涝灾害四川省重点研究室, 成都 610072

• 马振峰
• 1. 四川省气候中心, 成都 610072;2. 高原与盆地暴雨旱涝灾害四川省重点研究室, 成都 610072

• 刘佳
• 1. 四川省气候中心, 成都 610072;2. 高原与盆地暴雨旱涝灾害四川省重点研究室, 成都 610072

• 卿清涛
• 1. 四川省气候中心, 成都 610072;2. 高原与盆地暴雨旱涝灾害四川省重点研究室, 成都 610072

• 孙蕊
• 1. 四川省气候中心, 成都 610072;2. 高原与盆地暴雨旱涝灾害四川省重点研究室, 成都 610072

• 摘要：根据近34年（1981-2014年）156个四川地面站的地面观测资料，对比分析了在3种霾日判别方法下霾日的空间分布、季节变化、年代际变化特征以及气候趋势系数的分布特征.利用Morlet小波分析法比较了霾日的周期变化特点.结果表明：①原始观测霾日集中于四川盆地的北部；方法一霾日集中于盆地中南部和东部；方法二霾日集中于盆地地区，无代表高值区；方法三分布类似于方法一.②在季节变化上：不同判别方法下霾日的季节变化类似，秋冬霾日多于春夏.③霾日年代际变化明显.原始观测霾日2013年始出现突增；方法二霾日从2001年始有明显增长；方法一与方法三的年代际变化特征相似.④气候趋势系数上：原始观测变化特征不明显；方法一的系数呈现出南北反向变化；方法二与方法一的分布相似；方法三中系数的正负值区较为集中.⑤小波变化上：原始观测中霾日无明显周期变化，判别方法一、二、三均有各自的周期变化特点.

• Abstract：We compare and analyze the spatial distribution, seasonal and decadal variations, and the coefficient distribution of climate trend in haze days under three identification methods using 34 year (1981-2014) surface observation data from 156 stations in Sichuan. The periodic variation characteristics are analyzed by using the Morlet wavelet transformation. The results show that the original observation of haze days was concentrated in northern Sichuan Basin. Haze days were however concentrated in the south and east of the basin by using method I. Haze days were concentrated in the basin area without representative high value area by using method II. The distribution characteristics from method III are similar to them from method I. Different identification methods show similar seasonal variation of haze days higher in fall and winter than spring and summer. Haze days in Sichuan show obvious decadal variation. The original observation of haze days suddenly increased in 2013. Haze days have significant increase from the beginning of 2001. Inter-decadal variations from method I are similar to them from method III. The climate trend of original observation is not obvious. The coefficients obtained from method I show a reverse change from north to south. The distribution from method II is similar to that from method I. The positive and negative zones are more concentrated using method III. The wavelet transform results indicate insignificant periodic change in the original observation of haze days. Discrimination method I, II and III have their own periodic variations characteristics.

• Key words：discrimination method  haze day  spatial and temporal distribution  wavelet transformation analysis

• 
  摘要点击次数： 1311 全文下载次数： 1583