潮河流域TOPMODEL模型网格尺度研究

TOPMODEL模型首要也是关键的问题就是地形指数的计算问题.除计算方法外,地形指数的分布还与DEM网格尺度息息相关.较大的网格推出的地形指数不具有显示水流路径的物理含义,会丢失一些物理信息.太细的分辨率可能引起流向和坡度的混乱.导致水流路径不连续.对于流域尺度而言,合适的网格尺度在很大程度上影响模拟结果的精度,但目前关于这方面的研究却很少.本文在集水面积5 340km2的中尺度潮河流域上进行地形分析计算,深入的分析了从50 m～600 m以50 m为间隔的不同网格尺度对地形指数分布的影响.并以平水年为例,研究了这些地形指数对流域径流模拟的影响.得出的结论对于TOPMODEL模型在中尺度流域上的应用具有指导意义.     

基于结构网格的溃坝水流数值模拟

针对溃坝水流数值模拟面临的复杂地形和不规则边界等问题,基于结构网格建立了适应复杂地形和不规则边界的溃坝水流数值模拟有限体积模型（HydroM2D）。模型基于具有守恒特性的二维浅水方程,利用HLLC格式的近似Riemann解计算网格界面通量,利用MUSCL-Hancock法不断向前积分,使模型在时空上具有二阶精度;对源项进行离散处理确保模型的稳定性;模型引入有效干湿边界和不规则地形边界处理方法,准确模拟了干湿单元的动态交替和复杂边界上的水流特性。最后分别利用水槽试验、物理模型和实际算例对模型进行验证。结果表明,该模型对不同情景下的溃坝洪水模拟结果和实测资料以及现有模型模拟结果具有较高的一致性,模拟精度较高,稳定性较好,具有推广应用价值。     

基于GIS/RS的流域水文过程分布式模拟——Ⅰ模型的原理与结构

面向流域水资源管理,提出了一个基于GIS/RS的流域分布式水文模型,模型主要包括单元水文模型与河网汇流模型两大部分。单元水文模型涉及到冠层截留、融雪、蒸散发、坡面流、非饱和土壤水运动和地下水出流等水文物理过程。产流计算考虑到地形坡度的影响采用基于地形指数的计算方法。汇流演算基于河网结构采用分段马斯京根方法。模型的大部分参数与输入信息可以利用GIS和RS技术获取,能够对气候变化和人类活动对下垫面的改变,做出快速的模拟与响应。     

DEM空间分辨率对TOPMODEL径流模拟的影响研究

本文以褒河流域为例，引入信息熵的概念量化分析了DEM分辨率对地形特征（坡度、单宽上坡集水面积和地形指数）计算的影响，发现随着DEM分辨率的降低，计算得到的坡度值偏小，单宽上坡集水面积值偏大．从而导致地形指数计算值偏大。在此基础上，再基于不同分辨率的DEM数据得到的地形指数分布，采用TOPMODEL对褒河流域的9场次洪进行模拟，结果表明DEM分辨率降低会使得洪峰流量和径流深增大，确定性系数降低，模型的模拟精度变低．但其对峰现时差的影响不明显。     

基于下垫面特征的包气带厚度估算及水文模拟

从新安江模型中的包气带厚度的物理意义出发,考虑土地覆盖因素、地形坡度因素、土壤质地结构因素以及气候条件等因素,分别采用了土地覆被相关法、地形指数回归法以及综合要素法对研究区域的包气带厚度进行了估算。据此计算模型参数流域自由水蓄水容量,并在金沙江下游横江流域进行模拟对比试验。相比之下,综合要素法能更为准确的计算出特定空间尺度下的包气带厚度,模型模拟效果更好,更适用于此区域的模型参数定量计算及水文模拟。     

重庆山地月平均气温空间分布模拟研究

利用月平均气温物理经验统计模型,结合重庆市及周边51个气象站1961-2000年常规气象观测资料和重庆市1:25万DEM数据,充分考虑地形(海拔、坡度、坡向和地形遮蔽)、太阳辐射、长波有效辐射等因素对气温的影响,完成了复杂地形下重庆市山地100 m×100 m分辨率月平均气温空间分布的模拟.模拟结果能较好地反映气温的宏观分布趋势和局地分布特征,月气温模拟平均绝对误差最大为0.40℃,全年平均为0.21℃.采用交叉验证、个例年验证和野外考察资料验证对模拟结果和模型性能进行了多方面的考察,结果表明:交叉验证误差全年平均为0.26℃;个例年验证误差全年平均为0.30℃;8个野外考察点验证各月误差总体小于1.0℃.     

基于水库模型的扎龙湿地水循环模拟

基于水库模型的概念,根据地形地物将扎龙湿地划分为相互联系的9个水库单元,对每个水库单元进行水量平衡模拟,建立了扎龙湿地水循环系统模拟模型。以月为时间步长,利用建立的模型对扎龙湿地1971-2003年33年的水循环过程进行模拟,应用1971-1981年出口径流量进行模型参数标定,应用1982-1987年出口径流量和12个水面面积对模型模拟效果进行检验。模拟的出口径流量效率系数和相关系数接近0.90,相对误差小于3%;模拟的湿地水面面积也具有满意的精度。模拟结果表明,建立的模型对湿地水循环系统具有较好的模拟能力。     

基于对偶加权后验误差估计的2.5维直流电阻率自适应有限元正演

采用易于模拟复杂地形起伏和倾斜界面的非结构三角单元剖分网格,并利用对偶加权后验误差估计指导网格自动细化过程,实现了2.5维直流电阻率法自适应有限元数值模拟。在实例模型分析中,分别计算了层状模型和垂直岩脉模型的直流电阻率响应,并与其解析解进行了比较。对比结果表明,该算法所得数值解精度很高,解的相对误差小于0.5%。最后,计算了起伏地形2.5维地电模型视电阻率异常,并利用比较法进行了地形改正。地形改正结果与水平地形时的结果对比表明,比较法可以较好地消除地形影响,突出局部地质体异常。     

TOPMODEL模型的空间尺度分析

以东江流域的岳城、胜前、东坑、蓝塘和九州五个子流域为例,应用前四个子流域分析DEM分辨率和流域大小对地形指数和TOPMODEL模型模拟结果的影响,探讨TOPMODEL模型对DEM分辨率的依赖性.结果表明:DEM分辨率对地形指数有着显著的影响,且径流模拟精度依赖于DEM空间分辨率,随着DEM空间分辨率的降低模拟得到的确定性系数逐渐减小.为了克服TOPMODEL模型难以考虑降雨空间分布不均对径流过程的影响,建立基于子流域的TOPMODEL模型,将基于子流域的TOPMODEL模型和整个流域的TOPMODEL模型应用九州子流域进行模拟比较,发现基于子流域的TOPMODEL模型能够得到精度更好的模拟效果,而且可以得到不同子流域对流域出口流量过程的贡献度,进而能够分析不同降雨情况下的流域出口洪水过程.     

滁州花山水文实验流域径流与氮素耦合模拟研究

采用SWAT模型方法对滁州花山水文实验流域进行了径流与氮素耦合模拟,探讨了流域土地利用对氮素输移的影响。构建了花山水文实验流域SWAT模型,采用2008年1月~2013年12月的降雨、径流等数据对径流模型进行率定和验证,以2012年5月~2013年12月的氮浓度监测数据对氮负荷模型进行验证。利用所建立的模型对现状土地利用情形和土地利用变化情景假设下进行了径流与氮素耦合模拟,计算了现状和不同土地利用情景假设下流域的径流深和氮负荷。结果表明:汛期的氮流失量占全年氮流失的比重较大,每年的汛期是控制氮流失的关键时期;旱地的氮流失潜力较水田大,林地的增加对氮负荷具有一定的消减作用;控制氮肥的使用,加强水土保持能力,可以有效减少流域氮负荷。     

内陆河流域分布式日出山径流模型——以黑河干流山区流域为例

应用常规的气象水文数据并结合GIS,建立了一个适合西北干旱区内陆河山区流域的以日为步长的分布式径流模型,并对黑河干流山区出山径流进行了模拟计算和讨论。模型以子流域作为最小的产流、汇流单元,根据植被覆盖将各子流域分为裸地区、乔木区、牧草区和冰川区,并根据实际调查将土壤分为 3层,各分区单独进行水量平衡计算。产流过程以土壤储水能力和储水量表征,而储水能力和储水量等则由土壤的孔隙度、干密度和厚度等表征。入渗原理基于土壤储水率平衡原理,并考虑重力势的作用。实际蒸散发与蒸发力和土壤体积含水量的乘积成正比,不同的土壤和植被具有不同的调节参数。模拟结果表明,模拟效果较差的原因是区域日降水过程具有较大的随机性,难以用有限的站点合理计算区域日降水量。寻找一个合适的区域日降水量计算方法是目前少资料大型流域分布式水文模型模拟成功的关键。     

APHRODITE降水数据驱动的融雪径流模拟

针对流域内气象观测站点稀少和融雪径流过程的特点,利用APHRODITE降水数据进行插值,应用日有效活动温度改进度日数;依据季节性冻土受有效活动积温影响的特点,建立有效活动积温与径流系数的关系,提高模型中融雪速率和径流系数的计算精度。结合气象、水文资料和MODIS遥感积雪产品,应用改进的融雪径流模型（SRM）对开都河流域2000年与2006年融雪期的径流进行了率定和验证模拟。改进模型在率定期和验证期的模拟结果远远优于用日平均温度作为度日数的结果。结果表明,用APHRODITE降水数据及改进的度日数和径流系数作为SRM模型参数输入,能够较好模拟开都河流域融雪径流过程,大大提高模型模拟精度。     

Utilizing the MODIS-derived leaf area index to investigate the impact of vegetation processes on hydrological simulation of macroscale catchment

Integration of vegetation processes in rain–runoff (RR) models significantly affects runoff response by influencing evapotranspiration in mesoscale catchments. However, it is impossible to interpret the impacts of vegetation processes on runoff simulations in macroscale catchments using results from mesoscale catchments. Few studies involved vegetation process impacts on hydrological simulations by integrating daily vegetation information into conceptual RR models of macroscale catchments. In this study, we integrated the remotely sensed leaf area index (LAI) from the Moderate Resolution Imaging Spectroradiometer (MODIS) into a daily Distributed Time-Variant Gain Model (DTVGM). Then, this study assessed the performances of two DTVGM versions, with and without vegetation processes, in the Wei River catchment, China. The results showed that: (1) Integration of MODIS-LAI into the DTVGM model improved the calibration and runoff simulation results of the initial DTVGM model. (2) Inclusion of vegetation processes in the DTVGM changed the simulated proportions of water balance components in the hydrological model and made the simulation of water balance components more accurate. (3) The fact that inclusion of vegetation processes could improve the hydrological simulation performance of the daily conceptual RR model in the macroscale catchment was consistent with studies in mesoscale catchment.     

A distributed water–heat coupled model for mountainous watershed of an inland river basin in Northwest China (II) using meteorological and hydrological data

A distributed water–heat coupled model (DWHC) is calibrated by using daily precipitation data from 26 hydrological and meteorological stations: daily averaged air temperature data from the 11 stations and daily pan evaporation data (E601) from the 15 stations in 2000. Six tests by using different spatial interpolation methods to calculate the above daily meteorological data in each 1 km × 1 km grid, are designed to simulate the mean daily runoff generated from the research Heihe mountainous watershed in 2000. Due to spatial sparseness and asymmetry of the hydrological and meteorological stations, the results of the six tests have little differences. The interpolation method in 3-D mode considering altitude is not better than those taking no account of altitude, nor are the model results when the daily meteorological data at the two stations far from the research watershed are complemented. At last, a nearest neighbor interpolation method in 2-D mode is used to calibrate the DWHC model, in which the revised Nash-Sutcliffe Efficiency NSE, balance error B, determinate coefficient R ², root mean square error RMSE and average absolute error MAE is about 0.61, 0.08%, 0.73, 25.0 and 15.8 m³s⁻¹, respectively. However, by using the daily data in 1999 to validate the model, the NSE, B, R ², RMSE and MAE are, respectively, 0.63, −2.98%, 0.77, 34.9 and 20.3 m³s⁻¹. The reason that the model result is not favorable is mainly because of the lack of detailed soil information, meteorological data and vegetation data; even worse, the basic equations for runoff generation processes are mainly derived from the research results in other regions and meanwhile, its flow concentration method should be improved too. The water balance of the research watershed in 2000 is also discussed in this paper. Though the runoff simulation results are not favorable, the estimated evapotranspiration and runoff components are in accordance with the usual knowledge qualitatively, parts of which meet with the field measurements. According to the model results, the runoff is mainly generated from the land surfaces and shallow soil layers in this cold mountainous watershed. The alpine meadow has evident water conservation function based on the model results, field investigation and field observation results. The DWHC model also reproduces the formation processes of the thick-layered ground ice to some extent, though it is suppositional due to lack of detailed soil, vegetation and meteorological information.     

Regulation of vegetation pattern on the hydrodynamic processes of erosion on hillslope in Loess Plateau,China

As vegetation are closely related to soil erosion, hydrodynamic parameter changes under various vegetation pattern conditions can be used as an important basis for the research of the soil erosion mechanism. Through upstream water inflow experiments conducted on a loess hillslope, how the vegetation pattern influences the hydrodynamic processes of sediment transport was analyzed. The results show that the placement of a grass strip on the lower upslope can effectively reduce runoff erosion by 69...     

喀斯特流域分布式水文模型及植被生态水文效应

根据喀斯特流域多孔介质与裂隙水流特征,建立了达西流、裂隙渗流与槽蓄汇流演算相结合的混合汇流演算模式,实现了对分布式水文-植被-土壤模型(DHSVM)的改进。利用贵州普定喀斯特生态水文试验站陈旗小流域观测资料,对模型计算的流量过程及植被截流、蒸散发及土壤含水率时空分布进行验证。结果表明,模型能较好地模拟喀斯特流域陡涨、陡落的流量过程。同时,模型能模拟出土壤含水率、实际蒸散发与降雨、下垫面岩溶裂隙、植被覆盖的响应关系,对分析中国南方喀斯特地区下垫面变化条件下的生态水文效应具有重要意义。     

内陆河高寒山区流域分布式水热耦合模型（Ⅰ）：模型原理

在全球变暖的背景下，我国多数大江大河源区存在冰川退缩、雪线上升以及多年冻土和季节冻土明显退化等现象，并由此造成河源区产流量减少以及生态环境恶化等诸多问题，这在内陆河山区流域体现的较为明显，但目前分布式水文模型中很少涉及冻土水热耦合问题。文章以黑河干流山区流域为例，构建了一个内陆河高寒山区流域分布式水热耦合模型（DWHC）。模型基于土壤水热连续性方程将流域产流、入渗和蒸散发过程融合起来，在植被截留、入渗、产流和蒸散发计算方面也有所改进和创新，部分模块具有多个可选择方案。模型设计了与中尺度大气模式MM5的嵌套接口，也可以用地面气象资料驱动。模型在1 km×1 km网格基础上，以日为时间步长，将流域土壤分为18类，土壤剖面分为3～5层不等，流域植被概化为9类。模型只需要土壤初始含水量、初始地温和常规气象资料，以及土壤和植被物理参数，就能够连续演算各层土壤的温度、液态含水量、固态含水量、感热传导、潜热变化、水势梯度、导水率以及水分入渗和毛细上升量等水文循环要素。主要介绍了模型的基本原理和构建思路，有关模型的地面资料驱动结果和与MM5嵌套结果部分，参见后续文章(Ⅱ)、(Ⅲ)。     

Assessment of the effects of vegetation on soil erosion risk by water: a case of study of the Batta watershed in Tunisia

Soil erosion by water is a serious environmental problem which affects particularly the agriculture of developing countries. Due to specific factors, such as high rainfall intensity, steep slopes and vegetation scarcity, Tunisia is prone to soil erosion. Taking this into account, the main objective of this study was to estimate the soil erosion risk in the Batta watershed in Tunisia using qualitative and quantitative erosion model with remote sensing data and geographic information system (GIS). Moreover, a developed method that deals with evaluating the impact of vegetation on soil erosion by water is also applied. This method used multi-temporal satellite images with seasonal variability and the transformed soil adjusted vegetation index (TSAVI) which is appropriate in arid and semi-arid areas. For both erosion models, the results show that a large area of the Batta watershed is seriously affected by erosion. This potentially high risk is due especially to severe slopes, poor vegetation coverage and high soil erodibility in this watershed. Furthermore, the use of multi-temporal satellite images and vegetation index show that the effect of vegetation is a significant factor to protect the soil against erosion. The risk is especially serious in the summer season, but it decreases with the growth of vegetation cover in spring. Erosion model, combined with a GIS and remote sensing, is an adequate method to evaluate the soil erosion risk by water. The findings can be used by decision makers as a guideline to plan appropriate strategies for soil and water conservation practices.     

Evolutionary trend of water cycle in Beichuan River Basin of China under the influence of vegetation restoration

To understand the influence of vegetation restoration on the water cycle in semiarid areas, the effects of vegetation restoration on evolution of the key elements of water cycle were clarified by analyzing the evolutionary trend of atmospheric precipitation, ecological consumption water, and surface runoff on a river basin scale on the basis of analytical results of the changes in vegetation coverage and the long-term meteorological and hydrological monitoring data of Beichuan River Basin. The results show that the vegetation cover in the Beichuan River basin has rapidly increased in the hilly and mountainous areas since the 1980s, especially from 2000 to 2019, with the maximum and average vegetation cover rates increased by 14.98% and 52.2%, respectively. During 1956-2016, the annual precipitation in the basin remained relatively stable; the annual surface runoff slightly declined, with an average attenuation rate of 20 million m~3/10 a. The main reason for the runoff decline is the increase in ecological water induced by the vegetation restoration, which has changed the spatial-temporal distribution of the water from atmospheric precipitation in the basin. Spatially, more precipitation was converted into ecological water. As a result, the remnant runoff supplied to the lower reaches reduced accordingly. Temporally, more precipitation participated in the soil water-groundwater cycle, thus prolonging the outward drainage period of the precipitation. Moreover, the large-scale vegetation restoration induced a significant decrease in the surface wind speed, evaporation from water surface and drought index. As a result, a virtuously mutual feedback relationship was formed between the vegetation and meteorological elements. Therefore, vegetation restoration is of great significance for the improvement in the water conservation capacity and semiarid climate conditions in the Beichuan River basin.     

蓄满—超渗兼容水文模型的改进及应用

在介绍蓄满—超渗兼容模型理论和自由水箱模型理论的基础上,针对其各自的结构特点和产流机理,对二水源蓄满—超渗兼容水文模型进行改进,模型能反映地表径流、壤中流和地下径流的动态变化,可应用于洪水预报、水情模拟、水资源开发和利用等。并将模型应用于伊河流域栾川站,采用单纯形法、罗森布瑞克法及基因法联合自动优选法对模型参数进行了率定,并与原模型进行了比较分析,结果表明,改进模型模拟精度高,且能反映不同径流成分的变化过程。