Abstract：In order to solve the problem of outliers in meteorological visibility observations,a dynamic threshold optimization algorithm for meteorological visibility observation outliers is proposed.Firstly,the influence factors of visibility anomaly threshold were deconstructed,and based on this,an adaptive algorithm optimization design scheme was proposed according to feature engineering,anomaly correlation factor coupling model,etc.The experimental results show that the proposed scheme can maintain efficient and stable outlier recognition capability under complex meteorological conditions,providing strong technical support for the intelligent development of meteorological operations.
2025, Vol. 42 
