Abstract : This study aims to address the problem of excessive discrepancies in temperature data between primary and backup stations to ensure the accuracy and reliability of observation data. Based on the 2023 temperature observations from the primary and backup stations at the benchmark meteorological station in Guiyang , this study analyzes the cases in which data differences between dual sets of national-level automatic meteorological stations surpass the allowable range . A method is developed to effectively identify and assess data anomalies by employing spatial regression diagnostics , temporal consistency verification , and linear relationships between automatic stations . The proposed method enables rapid and accurate identification of erroneous data , and can support the accurate judgment and timely correction of observational data. This method can assist operation personnel in promptly detecting sensor performance fluctuations , and in repairing or replacing malfunctioning equipment , thereby safeguarding the quality of meteorological observation operations .