自动站小时气温数据质量控制系统研究

摘要

The hourly meteorological data from the automatic weather stations(AWS) is very vital to meteorological disaster warning,decision-making service and forecast and so on.Based on the hourly temperature data from the national AWS,the questionable and wrong hourly air temperature data were analyzed.According to the daily maximum/minimum air temperature in four times (02:00,08:00,14:00,20:00 Beijing time) from the national AWS,a quality control procedure of hourly air temperature data was developed for the temperature of AWS in China.The quality control procedure could be used for the regional and national AWS.It had been evaluated by the hourly air temperature data about 27000 weather stations from 2006 to 2010 in China.The results show that the accurate rates,questionable rates and wrong rates of hourly air temperature in the regional AWS are 99.431％,2.24‰ and 3.45‰,and the corresponding rates in the national AWS are 99.82％,1.27‰ and 0.49‰,respectively.Both questionable rates are similar,while there is a magnitude difference for both wrong rates.According to the analysis of quality control of long time series air temperature data,it suggests that the design of the procedure is reasonable,and it could check the questionable data and wrong data.Thus,it could be used in the meteorological operation.%逐小时自动站数据对于气象灾害预警、决策服务及预报预测等十分重要.以国家级自动站小时观测气温数据为基础,分析研究小时气温数据的疑误形式,针对各种疑误数据,利用国家级台站建站以来的日最高、日最低以及4时次(北京时02、08、14时和20时)定时观测气温数据,研制形成适用于中国自动站(区域站和国家站)逐小时气温数据质量控制系统,并将此系统应用到2006-2010年中国27000多自动站小时气温观测数据中.结果表明:区域站的正确率、可疑率、错误率分别为99.43％、2.24‰和3.45‰,国家站则分别为99.82％、1.27‰和0.49‰；区域站和国家站数据的可疑率相当,但错误率国家站明显比区域站低一个量级.通过历史数据质量控制分析,证明自动站气温质量控制系统设计合理,可以判断出错误数据和可疑数据,具有可用性.