洪潮分离-耦合的高潮位自动预报模型OA

To address the challenge of high-tide-level forecasting in frequently scoured and silted tidal river sec-tions,where stable underwater topographic data are lacking,this study proposes a"flood-tide separation-coupling"modeling framework and develops an automatic forecasting model that requires only routine hydrological data.First,based on the principle of bidirectional wave superposition,the total cross-sectional water level is decomposed into two components:the"flood-dominated water level",representing runoff effects,and the"tide-dominated water level,"representing tidal effects.Second,using hydrological data from 2015 to 2019 at Zhijiang and Zhapu stations in Qiantang River,four sub-models were constructed:tide-dominated low-tide level routing,tide-dominated tidal range routing,flood-dominated water level routing,and effective tidal range propagation coefficient.The core of the model lies in screening the"pure tide"and"pure runoff"samples through flow thresholds to calibrate the funda-mental relationships and in designing a real-time calibration mechanism:a sliding time window(30 days)is em-ployed to dynamically update the tide-dominated tidal range relationship,and an offset coefficient β is introduced to dynamically adjust the comprehensive curve and envelope curve,thereby adaptively responding to riverbed scour and siltation changes.Finally,data from 2020 to 2024 were used for real-time calibration and forecasting valida-tion.The results show that during flood periods with average daily discharge exceeding 10 000 m³/s,the forecast er-rors for high-tide levels at Zhijiang station are all less than 0.3 m,with a maximum error of-0.28 m,and a forecast-ing qualification rate of 100%,meeting the Class-A accuracy standard specified in the"Standard for Hydrological Information and Hydrological Forecasting"(GB/T 22482-2008).This model overcomes the dependence of tradi-tional methods on underwater topographic data,enhances forecasting stability in frequently scoured and silted river sections by learning recent hydrological patterns in real time,providing a reliable and practical technical solution for operational flood level forecasting in similar tidal river sections.