宁夏自动站与人工观测气温的差异对比分析

Comparative Analysis on Temperature Difference of AWS and Man-observed Station in Ningxia

利用2003—2004年宁夏17个自动站和人工站的气温平行观测资料,进行了对比差值及其均方根、标准差、粗差率和一致率等方面的对比分析,并对自动站观测气温序列进行了显著性检验。结果表明:由于观测原理、时次、方式等方面的不同,造成了两种仪器的气温偏差有明显的日变化,季节性和地域性差异不明显,自动站观测气温与历史序列无显著性差异。虽然自动站有很多人工站无法比拟的优越性,但两者之间的偏差较大,特别是白天,偏差幅度超过±0.2℃,离正常业务使用有一定的距离,所以需要一定时间的平行观测,对其数据序列进行均一性分析和客观订正,使其在天气预报和气象服务工作中更好地发挥作用。

Comparing with conventional observation, the automatic weather station （AWS） observation is less interfered by people and has high time resolution, its operational use will bring more active contribution to people＇s life and property and local economy development. However, it is found that in AWS operational use, the records of two AWS are different in the same observation environment just because of the different apparatus and different observation mode. Both national and international researches show that the apparatus＇s change is the important factor which causes the inconsistency in the climate data sequence. In real atmosphere, the air temperature＇s fluctuation is obvious, and the radiation error caused by solar radiation can＇t be ignored, so the individual datum can＇t be compared simply, instead the whole data sequence is to be used. In order to compare the differences of two data sequences, the average discrepancy of two data sequences and the exceptional data in AWS data record and the differences with historic data sequence must be analyzed. According to the principle of observation data＇s contrast, by using the observed data of hourly, daily maximum, minimum and average temperatures and the corresponding parallel data in 02：00, 08：00, 14：00 and 20：00 （BT） of 17 automatic and conventional weather stations in Ningxia from 2003 to 2004, the difference, average square root, standard error, coarse difference and consistency are analyzed and the significance test of AWS air temperature records are conducted. Results show that, differences exist between the apparatus because of the different sensors and working theories, the asynchronism of the observation and the different methods caused the discrepancy between the 2 sets, there are obvious system discrepancy and diurnal change of errors between AWS and man-observed station because the AWS is more sensitive than manbserved one, and the values of man-observed are lower than that of the AWS observation. The temperature observation at night is stable, and the errors are within the normal value, at daytime, the errors are bigger than the normal value, especially at sunrise, the maximum and minimum temperature periods. Most of the diurnal maximum, minimum and average temperatures exceed the stan- dard range. The AWS observation has little difference among seasons and regions, but it has obvious discrepancies with man-observed station, so it shows that AWS observation of air temperature isn＇t steady, and it has obvious discrepancy with manned observation station especially at daytime, the discrepancy of air temperature record of AWS and the historic record are not prominence. The error by the comparison between AWS and manned observing apparatus is caused by the difference of two different apparatus, it is not sure that if it is caused by manned apparatus or AWS apparatus, it＇s only a relative comparable results. It＇s more complicated actually, the error is not only caused by observing apparatus but also＇caused by the random fluctuation of the atmosphere. The manned observing apparatus is recognized by meteorologists, if the difference between AWS and manned observing apparatus is not obvious, then the AWS is as good as manned observing apparatus, but research shows that even the AWS is superb than the manned station, there are big errors by AWS especially in daytime, the range may be more than ±0.2 ℃ and it is not suitable for operational use and have to get parallel observations, objective calibration must be conducted to provide the scientific basis for the utilization of the AWS data.