【作者】 仕玉治；

【导师】 周惠成；

【作者基本信息】 大连理工大学 ， 水文学及水资源， 2011， 博士

【摘要】 气候变化和人类活动影响是当前流域水文水资源演变规律研究的两大科学课题。水循环系统是气候系统的重要组成部分,而气候变化必然引起流域水资源的时空变化。人类活动对流域的水循环影响,一方面通过改变土地利用、以及修建大型水利工程等改变流域下垫面,破坏了流域内天然的产汇流机制；另一方面通过工农业取用水以及跨流域调水等活动直接影响流域水资源的时空分布。当前流域水资源俨然已经发生巨大变化,气候变化和人类活动是两个主要的驱动因子,量化分析气候变化和人类活动对流域水文水资源的影响,对认知水资源演变规律和实现水资源可持续利用发展具有重要的实践意义。鉴于此,本文以东北地区两个流域为研究实例,以大尺度流域水文模型SWAT为基础,开展气候变化及人类活动对流域水资源的影响研究。本论文的研究内容和成果如下：(1)流域水文气象要素变化特征分析。基于多种趋势检验方法和变点分析方法,分析了东北两个实例流域水文气象要素的变化趋势和变点,通过水文气象要素的特征识别,可以初步判断流域气候变化和人类活动影响的大小及变化特征。研究结果表明多种方法可以定性识别流域的降雨、温度等要素的变化特征。(2)大尺度流域水循环模型SWAT构建。本文阐述了模型结构及其原理方法,并详细说明模型相应气象、土壤、土地利用、土壤等资料数据库构建,为下一步进行流域水循环模拟以及量化分析气候变化和人类活动影响研究提供有力支撑。(3)针对历史气候变化和人类活动影响的贡献量化问题,提出一种基于SWAT的新的分项量化方法。该方法为削弱人类活动时期内降雨丰枯变化对分离气候变化和人类活动影响量的干扰,对长序列降雨、温度进行还原计算,利用长序列降雨、温度数据资料分析气候线性变幅,然后将年降雨的线性变化量根据年内月分配比例对实际降雨量序列逐年进行还原,依据各月份温度线性变化量对实际温度序列逐年逐月进行还原；构建有、无降雨和温度还原计算的SWAT模型,利用基准期数据资料进行模型参数率定,并分别模拟人类活动时期有、无降雨和温度还原计算的径流量,通过对比分析人类活动时期相对于基准期的还原和实际径流的相对变化量,分离出降雨丰枯变化量、气候变化量以及人类活动影响量,即可实现历史气候变化及人类活动对河川径流影响量化分析。本文以辉发河流域为研究对象,对模型方法进行了验证说明,结果显示本文提出的方法能够合理地评估气候变化和人类活动影响的贡献量。(4)针对灌区农业灌溉用水人类活动影响量化问题,建立基于改进的SWAT灌区水循环模型。该模型在原模型基础上,修改了灌区水源灌溉模式、灌区田间水量平衡方程计算项、渠道渗漏计算以及水田灌溉制度,分别为泡田期和水稻生育期定义灌溉判断准则,在两个时期均可实现多水源灌溉模式；以浑太河流域灌区为研究对象,对改进后的灌区模型进行验证,通过重构有、无灌溉条件下的还原模拟,量化分析浑太河流域灌溉回归水量和灌溉影响比率,以及农业灌溉对灌区水循环要素的影响。(5)为减弱单一气候模型预测的不确定性,本文提出基于贝叶斯加权平均(REV-BMA)的GCM多模式气候变化预测方法,并耦合SWAT大尺度流域分布式水文模型研究未来气候变化对流域水资源的影响。首先,利用自动降尺度方法(ASD)和距离倒数权重插值降尺度方法(IDW)分别对GCM模型(HadCM3、CGCM3、BCCR、CSIRO)进行降尺度,采用提出的REA-BMA方法计算多模型的可靠性因子,获得未来时期(2011～2040年)降雨、温度要素的加权变化量；然后,依据降雨、温度未来变化量对基准期(1961～2000年)实际降雨、温度数据进行比例缩放后作为SWAT模型输入,并构建未来气候变化对流域水资源影响的评估模型(REV-BMA-SWAT);最后,选取浑太河流域为研究对象,分析未来时期(2011～2040年)浑河干流上大伙房水库入库径流变化,以及浑太河流域水循环要素的相对变化状况,并与单一GCM模型、RVM模型以及基于历史趋势变化的模拟结果进行了比较分析。更多还原

【Abstract】 Climate change and human activities impact are two topics in the current evolution law research of hydrology and water resources. The water circular system is an important part of the climate system, and climate change must cause water resources change in temporal and spatial. However, human activities are the most important factors in watershed hydrologic cycle system. On the one hand, land use changes and construction of large hydraulic engineering make the basin underlying surface changed, and then affect the runoff yield and concentration mechanism; on the other hand, industrial and agricultural water use and inter-basin water transfer affect the temporal and spatial distribution of water resources directly.Current water resources system has changed enormously. It has important significance to quantify the influence of climate change and human activities for the cognition of evolution law and sustainable development of water resources. In view of this situation, this paper carries out the evolution law study of water resources based on large-scale distributed hydrological model. Research contents and results are as follows:(1) Change characteristics analysis of hydrometeorological factors in typical basin. Many trend test and change-point analysis methods are used to analyze the change tendency and change-point recognition of hydrometeorological factors in the typical basin of northeast China. Based on the hydrometeorological factors recognition, the degree and change characteristics of the basin climate change and human activities can be required primarily. Results show that these methods can qualitatively recognize the change characteristics of precipitation, temperature and other factors in the typical basin.(2) Construction of large-scale hydrologic cycle model. This paper describes the structure and principle of SWAT model, and then displays the database construction for meteorology, soil and land use in detail. These works can provide vigorous support to the hydrological cycle simulation and to the quantitative analysis of climate change and human activities influence.(3) To deal with the quantitative analysis problem of historical climate change and the influence of human activities, a partial quantization method is put forward base on SWAT model. In order to weaken the interference of high and low variation of rainfall in a short period to the separation calculation of influence quantity of climate change, reducing calculation of the rainfall in the influence period of human activities is made, linear change range of climate is analyzed using complete series of rainfall data and temperature, linear change quantity of annual rainfall is reduced to practical rainfall series according to distribution proportion of every month, model parameters calibration is made using rainfall data in reference period, and runoff simulation is conducted under different rainfall condition in the period of human activities. By comparing the runoff change under different rainfall data condition in human activities influence period with that in reference period, high and low change quantity of rainfall, climate change quantity and influence quantity of human activities can be separated out, then quantitative analysis of historical climate change and human activities influence on river runoff can be realized. Taking the typical Huifa river basin as the study area to validate the model, the result shows that the method proposed in this paper is successful in separating the influence quantity of climate change and human activities more reasonably.(4) To deal with the quantitative problem of the influence of human activities such as agricultural irrigation water use in irrigation district basin, a advanced water cycle model base on SWAT model is proposed. Modifications are made to calculation methods of water source irrigation and field water balance equation to calculate paddy field runoff, the judgment criterions of irrigation in ponding period and rice growth period are defined respectively, so that the multi-water resources irrigation mode can be put into effect in both periods. Taking Huntai basin irrigation district as the study area to validate the irrigation district simulation model, by remodeling river flow under irrigation condition or without irrigation condition, calculating the ratio of irrigation water regression, irrigation influence coefficient, to quantize influencing of irrigation on river flow and water cycle factors.(5) To deal with the uncertainty of forecast results of single climate model in the study of future climate change influence on water resources, a REV-BMA based GCM Multi-mode weighted average method for climate data processing is proposed, which explicitly relates SWAT hydrological model in large-scale basin to research the influence of future climate change on water resources. First, two downscaling methods, automated statistical downscaling (ASD) and interpolation distance weight (IDW) are applied to downscale the GCMs(CGCM3、HadCM3、CSIRO、BCCR) respectively; Second, the REV-BMA method is employed to compute the weighted variable quantity of future climate in the GCM multi-model, then proportion scaling is made to the rainfall and temperature input data in reference period according to future change quantity of rainfall and temperature, and applied SWAT to simulation. Finally, taking Huntai River basin as the study area, establish climate change on water resources impact assessment model (REV-BMA-SWAT) to assess the change of dahuofang reservoir inflow and basin hydrological cycle factors during 2011～2040, and by comparing it with the single GCM simulation, RVM model simulation, and based on history data trend simulation.更多还原