Application of AI Method to Quality Control in Surface Temperature Observation Data

A method of spatial quality control (called AI for short), based on auto-regression and inverse distance weighting (IDW), is proposed. The method enables quality control of meteorological data in both the temporal and spatial dimension. Aimed at assessing the applicability of the method, in this study, hourly temperature observational data for the year 2007 from four surface meteorological stations in different regions (Nanjing station 58238, Lianyungang station 58044, Wuxi station 58353, and Xuzhou station 58027) were selected as controlled objects to carry out quality control using the AI method. Compared with IDW and the spatial regression test (SRT) in discriminating artificial errors, it is shown that the proposed method can mark suspicious data effectively. Furthermore, it is highly effective, stable, adaptable, and applicable in both plain and hilly areas.