【作者】 王彪；

【导师】 朱建荣；

【作者基本信息】 华东师范大学 ， 河口海岸学， 2011， 博士

【摘要】 珠江是中国南方的最大河系,径流量全国第二,仅次于长江。珠江径流包括了西江、北江、东江三条主要河流,其中西江是其主干流。三江径流在珠江三角洲汇聚,经三角洲河网分流后由八大口门分流入海,形成独特的“三江汇集,八口分流”的水系特征。改革开放以来,珠江三角洲地区社会经济发展迅速,三角洲地区对环境资源的需求越来越大,其中淡水资源供应逐渐趋于紧张,枯季珠江河口发生盐水入侵,使得淡水供应更为紧缺。本论文依托国家海洋公益项目(项目名称：珠江口咸潮数值预报技术研究；项目编号：200705019)对珠江河口水动力、盐水入侵及其动力机制开展研究。为此,收集、整理了大量的基础数据,主要包括多批次的水文观测资料以及较为完整的河口、河网区的水深、岸线等地形数据。基于河口盐水入侵强度主要决定于径流量和潮汐这个物理机制,对珠江河口磨刀门水道平岗泵站的盐度、潮位以及上游径流量资料进行了处理、分析,建立了平岗泵站日平均盐度与潮差和径流量的经验统计模型。通过分离短时间尺度潮差变化和长时间尺度径流量变化对盐水入侵的不同影响,按潮差、径流影响的主次逐步回归建立一个盐度-潮差模型和一个盐度-流量模型,从而合成得到盐度统计模型。建立统计模型时,通过潮形标准化以及模型预估修正等,提高了模型精度。本文提出的盐度统计模型建立方法简单,可操作性强,而且物理意义明确,但由于统计模型更多的是从变量间的相关性角度出发,它虽然能得出变量间的定量关系式,但对变量间的内在因果联系揭示的不够。为了能从动力学上对珠江河口咸潮入侵的活动规律及其物理机制作出解释,建立一个基于河口动力学的数值模式是十分必要的。考虑到珠江河口的复杂地形条件以及盐水入侵在河口区域明显的三维特征,本文基于无结构网格FVCOM模型,在珠江河口建立了一个完全三维的盐水入侵数值模式,模式计算区域包括了整个珠江河网、河口以及邻近海域,模式中较为完整的考虑了径流、潮汐、风、斜压以及陆架环流等各种动力因子作用。模式验证结果表明,实测和模式计算的潮位、断面水通量、断面盐度以及野外船测定点的流速、流向、盐度等吻合良好,说明建立的珠江河口三维盐水入侵数值模式能模拟珠江河口的水动力及盐水入侵过程。基于这一模式,本文对枯季珠江河口的水动力及盐水入侵过程进行了数值模拟,并针对各主要动力因子设置了敏感性试验。基于数值试验结果,对珠江河口潮汐、潮流、河口环流以及盐水入侵的动力过程及相应动力机制进行了分析。珠江河口区域及口外浅海陆架海域8个主要分潮的同潮图分析表明,浅海陆架海域潮波总体上自东向西传播,等位相线呈东南-西北走向,进入河口区域,潮波从向西传播,逐渐转为向北,等位相线更加接近东西方向分布。在浅海陆架区域,K1和O1分潮振幅较大,而珠江河口区域则以M2分潮为主,浅水分潮成分很小。珠江河口潮汐属于混合潮类型,潮型系数介于0.8-1.5。珠江河口大潮期间,潮差介于2.2-3.1m,小潮期间减小到0.6-1.1m。珠江河口潮流特征在不同海区变化有所不同,口外海域潮流较弱,伶仃洋海域潮流较强,虎门处落急流速可达200cm/s。总体上,珠江河口潮流具有不规则半日潮变化特征,在一个大小潮周期过程中,潮流流态在小潮后的中潮期间出现异常,由不规则半日潮转变成不规则全日潮,此时涨潮历时远大于落潮历时。伶仃洋海域受东北季风影响,存在一个表层向西、底层向东的横向环流。在伶仃洋深槽区域,因潮流和盐度分层的大小潮周期变化,余流也有较为独特的大小潮变化特征,即在小潮期间存在较为明显的重力环流,在小潮后的中潮期间重力环流最强,大潮期间向海余流开始增强(或向陆余流减弱),最终在大潮后的中潮期间向海余流达到最强。受下泄径流、科氏力、盐度锋面以及风等作用,出伶仃洋向西有一股较强的西向沿岸流,这股强流在底层迅速减弱。对于珠江河口的盐度时空变化,总体上,珠江河口(主要为伶仃洋海域)的盐度随时间变化较不规则,与潮汐、潮流特征相似,在一个大小潮周期中盐度变化总体呈现为不规则半日周期变化,但在小潮后中潮期间出现较为明显的不规则全日周期变化。大小潮周期尺度上,伶仃洋由西至东逐渐出现盐度-潮差相位差。空间分布上,枯季因下泄径流受科氏力以及东北风的作用,珠江河口等盐度线具有沿岸分布特征,即在伶仃洋内总体呈东北-西南走向,出伶仃洋后等盐度线沿西南西-东北东方向向西延伸。伶仃洋中,盐度由西至东逐渐增大,深槽区盐水入侵相对更为明显,而且垂向分层总体上在小潮后的中潮期间最为明显。珠江河口的盐度分布对风应力的变化非常敏感,不同风向的风对珠江河口盐度的影响不同。径流增减使得珠江河口盐水入侵发生相应减弱和增强；相对而言,盐水入侵对径流减小响应比对径流增大响应更为明显。河口盐度对海平面变化相对较不敏感,海平面上升仅在磨刀门等局部区域影响较大。对于珠江河口磨刀门区域,资料分析表明,磨刀门水道盐水入侵存在异常特征,即盐度峰值与潮差峰值之间存在明显的相位差,半月周期尺度上的盐度峰值主要出现在小潮后的中潮。数值模拟结果分析表明,磨刀门盐水入侵的异常变化规律与洪湾水道关系密切。受地形和东北风影响,洪湾水道成为向磨刀门水道输送盐分的一个重要通道。东北风作用能显著加强磨刀门盐水入侵,但不是盐水入侵异常的根本原因。小潮期间,潮汐动力减弱,东北风作用相对增强,使得洪湾水道盐水入侵更为严重,最终导致磨刀门水道上游段的盐度峰值出现在小潮后的中潮期间。洪湾水道、风应力和潮汐的相互作用,是磨刀门水道盐水入侵异常特征的动力成因。更多还原

【Abstract】 The Pearl River is the largest river system in South China and its river runoff ranks second in China only to the Changjiang River. The Pearl River includes Xijiang, Beijiang and Dongjiang River with Xijiang as the main stream. Characterized by "runoff comes from three rivers and goes through eight gates", the river runoff from the three rivers meets in the Pearl River Delta, run through a complicated river net system and finally gets into the South China Sea through eight gates. Since the Reform and Opening, the economy of the Pearl River Delta develops rapidly and the demand for the environmental resource also grows fast. The fresh water supply is in shortage especially in dry season when saltwater intrusion occurs. Supported by the Marine Special Program for Scientific Research on Public Causes (Program name:Research on the numerical prediction of salt tide in the PRE, Grant No.:200705019), this paper investigates the hydrodynamic process as well as the saltwater intrusion in the Pearl River Estuary. To this aim, a great amount of basic data was collected and processed such as hydrologic data and the bathymetry data of river network and estuary area.Based on the mechanism that salt intrusion is mostly affected by river discharge and tidal forcing, a statistical regression model of the daily averaged salinity, with tidal range and river discharge as variables, was established by processing the Pinggang’s salinity and tidal level as well as the upstream river discharge data. Separating the different influence from short-timescale tidal range variation and long-timescale river discharge variation, a salinity-tidal range model and a salinity-discharge model could be obtained through stepwise regression analysis and then an integral salinity regression model was achieved by the combination operation. The model was improved by the means of tidal standardization as well as predict-revise operation. The approach in establishing salinity regression model is practical and of clear physical concept. However, due to the statistical nature, the model can reveal the quantity relations among salinity, tidal range and river discharge but it is not able to reveal the dynamics in them.To explain the dynamic mechanism of salt intrusion in the PRE, it is necessary to build a dynamical numerical model. Considering the bathymetry complexity of the PRE and the 3D feature of salt intrusion, a fully 3D salt intrusion numerical model was built in the PRE based on the unstructured model, FVCOM. The salt intrusion model considered relatively comprehensive dynamic factors such as river runoff, tide, wind, baroclinic effect and shelf circulation etc. and had a model domain covering the whole river network, estuary and the adjacent shallow shelf. The model was validated against tidal level, sectional water flux, section-averaged salinity, as well as current speed, direction and salinity, and validation results showed that the model is of an acceptable precision and could be used to simulate and investigate the hydrodynamic and salt intrusion processes.Based on the salt intrusion model, simulations of the hydrodynamics and salt intrusion in dry season condition were conducted and more sensitivity experiments against major dynamic factors were carried out. Then, tides, tidal current, circulation and salt intrusion in the PRE were analyzed and discussed according to results from those experiments.Cotidal charts of 8 major tidal constituents in the PRE and adjacent shallow shelf showed that the tide propagated from east to west in the shallow shelf with co-phase lines at southeast-northwest. It turned north when entering into the PRE with co-phase lines at west-east. In the shallow shelf, the Ki and O1 tidal component had relatively large amplitude while the M2 component dominated in the PRE. The shallow water tides were trivial. It was of a mixed tide pattern in the PRE indicated by a tidal form number ranging from 0.8-1.5. In the PRE the tidal range varied between 2.2m and 3.1m during the spring tide and between 0.6m and 1.1m during the neap tide.Tidal current varied at different area of the PRE, relatively weak in the adjacent sea and strong in the Lingdingyang Bay with a speed of 200cm/s at Humen outlet. In general, tidal current showed irregular semidiurnal variation but it turned to be irregular diurnal during the moderate tide after neap with flood duration far longer than ebb one.In the Lingdingyang Bay, the northeast wind drove a transverse circulation with westward flow at surface and eastward at bottom. In deep channels, the residual current varied irregularly that it showed gravitational circulation pattern during neap; this pattern became more obvious during moderate tide after neap; the seaward flow increased throughout the water column during the spring tide and reached its maximum during the moderate tide after spring. Due to the combination effect of river discharge, Coriolis forcing, salinity front and wind, a strong westward alongshore current was observed out of the Lingdingyang Bay at surface, which was greatly weakened at the bottom.As for the temporal and spatial variation, the salinity in the PRE changed irregularly similar to the tidal level and tidal current, that it was generally irregular semidiurnal and became irregular diurnal during the moderate tide after neap. In a spring-neap cycle, a salinity-tidal range phase difference was gradually emerged from west to east in the Lingdingyang Bay. As the river runoff was steered by the Coriolis forcing and northeast wind, the salinity contours lay along the coastlines, i.e. at NE-SW in the Lingdingyang and ENE-WSW out of the bay. In the Lingdingyang Bay, salinity increased eastward, the salt intrusion was more severe in deep channels, and the vertical stratification was most obvious in the moderate tide after neap. Salinity spatial distribution was sensitive to the wind condition and changed quite differently to different winds. The decrease and increase in river discharge would relieve and exaggerate the salt intrusion situation respectively, and the salt intrusion was more sensitive to the decrease than increase. Salinity was less sensitive to sea level rise in general except in local regions such as Modaomen area.In the Modaomen waterway, observed data showed that abnormal salt intrusion occurred there with an obvious salinity-tidal range phase difference, i.e. salinity peaks appeared mainly during the moderate tide after neap in a spring-neap cycle. Numerical simulations showed that the abnormal intrusion was closely related to the Hongwan waterway. Due to the combined effect of bathymetry and northeast wind, Hongwan waterway became an important channel transporting salt into the Modaomen waterway. Northeasterly wind could reinforce the salt intrusion in the Modaomen area however it was not the essential reason for the abnormal salinity variation. During the neap tide, the tidal force became weak while the wind force became relatively stronger and thus could generated a larger salt transportation through Hongwan waterway and finally resulted in the salinity maximum during the moderate tide after neap. The interactions among the Hongwan waterway, wind and tidal forcing were the dynamical mechanism of the abnormal salt intrusion in the Modaomen waterway.更多还原