【作者】 龚政；

【导师】 张东生；

【作者基本信息】 河海大学 ， 港口、海岸及近海工程， 2002， 博士

【摘要】 推导了σ坐标系下三维非线性斜压流场、盐度场数学模式，建立了长江口平面二维、三维正压、三维斜压诊断及三维斜压预报模式的流场、盐度场数学模型，采用先进的K-KL二方程紊流闭合模型求解垂向涡粘系数。通过对计算结果的分析比较，探讨了长江口流场、盐度场数值模式的选用问题。 对模型的水陆边界条件的处理做了深入探讨。采用东中国海潮波模型推算出来的开边界潮位过程作为外海边界条件，保证了所模拟的长江口潮流与东中国海潮流的整体一致性；对上游边界条件做了认真的校验，指出应该用上游断面的流量进行边界控制，在缺乏边界实测流量过程的情况下，提出了“流量过程线放大法”进行流量控制，保证了边界流量的真实性；在三维模型中借鉴了二维模型中“干湿法”的思想进行了动边界处理，在时间步长较大的情况下仍然可以保证计算的稳定性，拓宽了三维模型在河口和近岸水域的应用范围。 通过流体微团运动轨迹跟踪法对长江口平面二、三维Lagrange余流场进行了计算，给出了长江口余流的宏观特征；计算并分析了径流、风、及密度斜压效应等对Lagrange余流场的影响；首次对各种模式得到的流速和Lagrange余流结果进行了系统分析。 首次将误差逆传播神经网络(简称：BP神经网络)方法运用到河口盐水入侵强度预测研究中，结果令人满意，拓宽了研究河口盐水入侵的思路和途径，也提高了预测河口盐水入侵强度的效率。更多还原

【Abstract】 This paper presents a series of numerical models and simulated results of Yangtze Estuary. First of all, 3-D nonlinear baroclinic current and salinity numerical models with vertical cr coordinate were derived. 2-D, 3-D barotropic, 3-D baroclinic-diagnosis and 3-D baroclinic-prediction numerical models of Yangtze Estuary were established, in which vertical eddy coefficient was obtained from the advanced K-KL turbulence closure model. Based on analysis and comparison of results, it was discussed on choosing of current and salinity numerical models.Water and land boundary conditions for numerical models of Yangtze Estuary were dealt especially. Water level process obtained from tidal wave model of China East Sea was used as open water boundary condition of Yangtze Estuary models, which ensures simulated tidal current feature in Yangtze Estuary consistent with that in China East Sea. It is indicated that runoff at upstream section should be used as velocity boundary condition. If no field data of boundary runoff available, the "runoff process magnifying method" presented is suggested, which ensures the facticity of boundary runoff. Wetting and drying boundary condition in 3-D model was dealt with the idea of "wetting and drying method" in 2-D model, which ensures the simulation stability with longer time step and also widens the application area of 3-D model in shallow water zone.2-D and 3-D residual currents in Yangtze Estuary were simulated with tracking method of water particles, and macro characteristic of residual current was presented. The effects of runoff, wind and water density gradient on Lagrange residual current were simulated and analyzed. It’s the first time to analyze the simulated current and Lagrange residual current fields obtained from a series of models systemically.Error Back-Propagation Neural Network (BP Neural Network) method was used to predicate Salinity Intrusion in estuary for the first time, and the predicated results are satisfying. It not only widens research ideas and approaches on Salinity Intrusion in estuary, but also improves predication efficiency.更多还原