采用中国陆域1989年-2009年3小时0.71°格点的ERA-Interim地表2 m气温预报数据，运用累积分布函数与分位数对研究区的极端低温和极端高温进行分析。结果表明：东北、青藏高原及西北地区为极端低温的集聚区，极端低温逐步沿西北-东南方向递增；青藏高原为全国极端高温中的唯一冷中心，西北盆地为最热中心；ERA-Interim再分析资料能够较好地反映中国极端低温和极端高温空间分布规律，整体上具有较好的可信度。结合分位数方法定量分析了极端气温的持续时间，为极端气候事件的定量研究提供了方法借鉴。

In this paper, the 3-hour and 0.71o grid temperature forecast data of 2 m above ground surface during the period of 1989 to 2009 derived from ERA-Interim reanalysis dataset were applied for the analysis of extremely low and high temperatures in China. Based on the cumulative distribution function and quantile function, the results indicated that the Northeast China, Tibetan Plateau, and Northwest China are the main extremely low-temperature areas, and the extremely low temperature rises along the northwest-southeast direction. Tibetan Plateau is the sole cold area with the extremely high temperatures, while the basins located in Northwest China are the warmest areas. ERA-Interim reanalysis data can represent the spatial distribution of extremely low/high-temperature areas in China. The duration of extreme temperatures can be quantitatively analyzed using the quantile function, which provides reference for the quantitative analysis of extreme climate events.

福建省公益类项目（2013R04）；福建省科技厅产学研重大项目(2012Y4001)；欧洲中期气象预报中心(ECMWF)自主研究项目