感潮河段水位预报误差辨识与系统响应修正方法研究

In tidal rivers, flow movement is complicated under combined effects of river dynamics and ocean dynamics, and many error sources can affect the accuracy of real-time water stage forecasting. Error identification and correction is important for improving the accuracy of water stage forecasting in tidal rivers. According to the complex error mechanism of water stage forecasting in tidal rivers, an error identification approach based on Fourier series is proposed to analyze the effect of different source errors on stage routing. According to the problems of weak physical basis, poor performance and loss of forecast lead period existed in current error correction, an error correction method based on the system differential response theory is proposed to correct the key error sources on water stage forecasting in tidal rivers. Further, a flood forecasting model system with error correction capacity for typical estuary is developed to evaluate the performance of the proposed methods on error identification and correction. The characteristic and innovation of this research will include the development of key error sources identification method based on the formation process of tidal river stage, system response error correction method based on multi-source error incremental contribution, and flood-tide-correction combined model for flood forecasting in tidal rivers. This study will improve the accuracy of real-time flood forecasting in estuaries, and provide a new approach for error study in other fields.

感潮河段水流运动规律复杂，实时水位预报误差来源多，精度不高，误差诊断与修正对提高感潮河段水位预报精度具有重要意义。本项目针对感潮河段水位预报误差影响机理复杂的难题，提出基于傅里叶级数的误差辨识方法，分析不同误差源对感潮河段水位计算的影响；针对现有误差修正存在的物理基础不强、预见期损失和修正效果不理想等问题，构建基于系统微分响应理论的误差修正方法，对感潮河段水位预报的关键误差源进行修正；建立典型河口具有误差修正能力的洪水预报耦合模型系统，评估所提方法对误差源的辨识能力和多误差源的修正能力。与现有类似成果相比，在基于感潮河段水位预报误差形成全过程的关键误差源溯源反演识别方法、基于洪水预报多源误差增量贡献的系统响应修正方法、感潮河段洪-潮-修正一体化耦合模型等方面具有特色和创新。研究成果将有效提高河口地区实时洪水预报精度，也为其它领域误差研究提供新的思路。

感潮河段水流受上游洪水波和下游潮波的共同作用，洪水预报误差来源多且影响复杂，感潮河段水位预报一直是水情预报的难点。项目围绕感潮河段水位演算、洪水预报误差辨识与修正、洪-潮-修正一体化耦合等问题开展深入研究，主要获得以下成果：(1)以水量平衡方程和槽蓄方程为基础建立了水位演算基本方程；将感潮河段水位视为洪水波和潮波双向运动后的叠加，基于水位演算基本方程构建了双向波水位演算模型；考虑区间入流及参数变化，构建了考虑区间入流的变参数双向波水位演算模型，在钱塘江感潮河段和曹娥江感潮河段进行应用检验，获得了较好的模拟效果。(2)基于产汇流模型的物理基础，建立了水文系统微分响应误差修正方法；分项建立了降雨、产流等关键因素与流域出口流量误差之间的系统响应关系，基于系统微分响应分别构建了全流域平均修正和分单元修正；通过理想模型和实际流域进行了方法验证，采用系统响应修正后的水文模型计算效果优于原模型效果，分单元修正效果总体上优于流域平均修正；分单元修正可以将流域出口断面计算流量误差进行分解用于修正各单元误差，考虑了洪水预报误差的时空变化。(3)以新安江模型、感潮河段水位演算模型和潮位模型为基础建立了河口洪潮组合预报模型系统；将误差系统响应修正方法耦合到预报模型系统中构建了感潮河段洪-潮-修正一体化耦合模型系统，并进行了应用检验。项目研究达到了预期研究目标，研究成果进一步提高了河口地区洪水预报精度，可为其它领域误差研究提供新的思路。