Abstract：In order to solve the problem of insufficient realtime performance of automatic weather station data and unstable data transmission in the weak network environment,a cloud edge end realtime processing system is constructed,and multisource fusion,flow anomaly detection and quality control compression methods are designed at the edge side to verify typical stations.The results show that edge processing can reduce the average delay to the order of 100 milliseconds,improve the accuracy of anomaly recognition to about 97%,and reduce the amount of uplink data by about 60%,significantly enhancing the data quality and transmission stability.The research shows that the system is suitable for the engineering deployment of automatic station service,and it is recommended to further expand the edge intelligent application in complex scenarios.
窦建虎, 王雯燕, 高山, 李晓冬. 基于边缘计算的自动气象站数据实时处理方法研究[J]. 气象水文海洋仪器, 2026, 43(5): 22-26.
Dou Jianhu, Wang Wenyan, Gao Shan, Li Xiaodong. Research on real time processing method of automatic weather station data based on edge calculation. Meteorological Hydrological and Marine Instrument, 2026, 43(5): 22-26.
2026, Vol. 43 
