Preliminary study on temperature quality control and the fields of experts correction based on FCM
Liu Wenjing1,Liu Chengxiao1,Chen Hao1,Wang Gen2,3
1.Anhui Public Meteorological Service Center,Hefei 230031; 2.Anhui Meteorological Information Centre,Hefei 230031; 3.Atmospheric Environment Shenyang Institute of China Meteorological Administration,Shenyang 110016
摘要 本着提高安徽省地面气温资料质量的宗旨,文中采用模糊C均值(Fuzzy C-Means,FCM)聚类法进行地面气温资料的质量控制研究。具体执行过程中,通过FCM将区域内各测温划分为若干气温相似的聚类,定义离群率(空间尺度)和离群速度(时间尺度)识别出气温资料中的离群值。进一步基于专家场模型(Fields of Experts,FoEs)对识别出的气温离群值进行订正,FoEs考虑了邻近站和本站前后时次的气温信息。与传统方法相比,文中算法从整体气温出发,不需要设置气温参考值,FoEs不仅能够订正离群资料,还能对连续缺测资料进行插补。因而文中的方法具有实用性和科学性,较适合计算大样本的气温数据集。
Abstract：Aiming to improve the quality of ground temperature data in Anhui Province,the Fuzzy C-Means clustering algorithm is used in this paper to investigate quality control on temperature data.During the specific implementation process,similar clustering results of the area are outputted through FCM algorithm.The definition of stray rate (space scale) and the group velocity (time scale) of temperature can identify the outliers of temperature data.Further,outliers are corrected by using the method of fields of experts (FoEs),which have already considered the temperature information of adjacent and self station before and after the time.Compared with traditional methods,the algorithm is based on the overall temperature,which does not need to set the temperature reference value.FoEs can not only correct the outliers away from the group,but also can be used in interpolation processing for observational data that consecutive missing.The methods are more practical and scientific,which is suitable for calculating the temperature of large sample set.
刘文静,刘承晓,陈浩,王根. 基于FCM的气温质量控制和专家场订正初步研究[J]. 气象水文海洋仪器, 2017, 34(2): 9-15.
Liu Wenjing,Liu Chengxiao,Chen Hao,Wang Gen. Preliminary study on temperature quality control and the fields of experts correction based on FCM. Meteorological Hydrological and Marine Instrument, 2017, 34(2): 9-15.
WANG G,ZHANG J W.Generalised variational assimilation of cloud-affected brightness temperature using simulated hyper-spectral atmospheric infrared sounder data[J].Advances in Space Research,2014,54(1):49-58.
SHAFER M A,FIEBRICH C A,ARNDT D S,et al.Quality assurance procedures in the Oklahoma Mesonetwork[J].Journal of Atmospheric & Oceanic Technology,2000,17(4):474-494.
[7]
REEK T,DOTY S R,OWEN T W.A deterministic approach to the validation of historical daily temperature and precipitation data from the cooperative network[J].Bulletin of the American Meteorological Society,1992,73(6):753-765.
[8]
EISCHEID J K,BRUCE B C,KARL T R,et al.The quality control of long-term climatologically data using objective data analysis[J].Journal of Applied Meteorology,1995,34(12):2787-2795.
[9]
GANDIN L S.Complex quality control of meteorological observations[J].Monthly Weather Review,1988,116(5):1137-1156.
[10]
COLLINS W G.The operational complex quality control of radiosonde heights and temperatures at the national centers for environmental prediction.Part I:Description of the method[J].Journal of Applied Meteorology,2001,40(2):137-151.
[11]
FENG S,HU S Q,QIAN W H.Quality control of daily meteorological data in China,1951—2000:A new dataset[J].International Journal of Climatology,2004,24(7):853-870.
ZOU X,QIN Z K.Time dependence of diurnal cycle errors in surface temperature analyses[J].Monthly Weather Review,2010,138(138):2469-2475.
[15]
QIN Z K,ZOU X,LI G,et al.Quality control of surface station temperature data with non-Gaussian observation-minus-background distributions[J].Journal of Geophysical Research Atmospheres,2010,115(D16):751-763.
ROTH S,BLACK M J.Fields of experts:A framework for learning image priors[C]//Proc of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR),2005,2:860-867.
2017, Vol. 34 
