SWAT模型在土地利用/覆被变化剧烈地区的改进与应用

人口增长、气候变化、制度变迁、城市化等均会导致土地利用/覆被的变化,进而引起流域水文过程（截留、入渗、蒸散发和地下水补给等）和水循环过程的改变。当前,由于逐年土地利用/覆被数据获取困难、水文模型本身计算缺陷等问题,所有在流域尺度上开展的借助水文模型进行的土地利用/覆被变化影响下的水文模拟研究都存在一个共同缺点,就是采用的水文模型并不能逐年调用土地利用/覆被数据,即水文模型无法真实体现或模拟土地利用/覆被的时空变化。SWAT作为一个广泛应用的分布式水文模型,在其模拟期内,不能逐年调用土地利用/覆被数据,即在进行水文模拟时忽略了土地利用/覆被时间上的变化,这可能会影响其在土地利用/覆被变化剧烈地区（如黑河中游）的应用。黑河流域是典型的内陆河流域,也是中国西北地区第二大内陆河流域。黑河中游是黑河流域的径流耗散区。本文针对SWAT模型在考虑土地利用/覆被变化时的缺点,对其进行了改进并开发出能够逐年调用土地利用/覆被数据的LU-SWAT模型。在土地利用/覆被变化剧烈的黑河中游对SWAT和LU-SWAT模型的径流模拟效果进行比较,发现LU-SWAT模型更适用于黑河中游水循环模拟。     

TRMM及GPM降水数据在高寒内陆河流域的准确性评估

降水数据的准确性和时空分辨率成为水文过程模拟的关键。卫星遥感降水资料的日益丰富为资料缺乏区的水文模拟带来了新的突破。本研究拟在资料缺乏、下垫面复杂,观测难、建模难的柴达木盆地高寒内陆河流域—巴音河中上游,基于近5年的TMPA 3B42、GPM IMERG V5及GPM IMERG V6逐日降水数据和气象站点观测数据建立SWAT模型,采用流域出口径流数据及不同的参数化方案分别率定4个模型,比较和探究不同数据在巴音河流域的适用性。结果表明：① 在月、年尺度上,实测降水数据及TMPA 3B42 V07对应的SWAT模型径流模拟效果较好,前者模拟精度较后者仅高6%~12%,且二者对于流域水量平衡的刻画均较准确。说明TMPA 3B42数据对应的径流模拟结果误差相对较小,可直接用于高寒内陆河流域水文模拟;② GPMIMERG V5数据对应的那什系数N_SE值为0.13（月）、-1.58（年）,误差百分数P_BIAS值为41.2%（月、年）,均方根误差与标准误差比率R_SR值为0.93（月）、1.61（年）,其径流模拟误差较大,模拟效果不可信,说明GPMIMERG V5数据集并不适用于巴音河流域水文模拟;③ GPMIMERG V6-F对应的月径流模拟结果明显优于GPMIMERG V5-F,前者模拟精度较后者提高4倍,但其模拟的年径流对应的N_SE值为-0.12,R_SR值为1.09。该研究可为资料缺乏的高寒内陆河流域生态水文过程模拟提供参考。     

基于元胞自动机模型的河道汇流过程模拟

对河道汇流过程进行模拟可为洪水灾害预警预报提供参考。利用水力水文学方法能很好地模拟河道汇流过程,但需要输入的参数多,运算过程复杂,对数据精度要求高,而且在无资料区流域无法确定河道上断面流量情况下,该方法具有一定局限性。本文将元胞自动机模型与水文模型相结合,构建了河道汇流过程中的元胞自动机模型和产流汇流规则。通过建立河道坡面拓扑关系,利用SCS-CN（Soil Conservation Service-Curve Number）模型逐个计算河道元胞上的坡面入流,并利用曼宁方程模拟河道汇流过程,最后在ArcEngine平台下进行二次开发,实现了河道汇流可视化。本文以厦门市茂林溪流域为研究区,对1997年5月6日至7日的一场降雨进行了模拟。将本文模拟结果与该流域其他学者的研究进行了对比分析,结果表明在输入数据与水文模型参数相同的情况下,本文不仅模拟出每次降雨间隔产生的较小洪峰,并且整场降雨产生的最大洪峰流量精度与时间精度均提高了5倍,可以更准确地模拟河道汇流过程,适用于河道汇流可视化,该模拟可以为洪水灾害预警预报提供一定参考。     

川中紫色丘陵区径流泥沙SWAT模型的模拟应用分析

针对川中丘陵区紫色土坡耕地严重水土流失,选取典型代表李子溪流域为研究区,构建了其SWAT的模型数据库,包括地形、土壤、气象和土地利用数据库。并利用赵家祠水文站1970-1979年的实测径流和泥沙资料,对该流域的SWAT模型参数进行率定再采用1980-1986年的实测资料,对模型的适用性进行验证,同时用相对误差Re和Nash确定性系数Ens评价模拟效果。结果显示,径流和泥沙模拟相对误差均在±15%范围以内,Nash确定性系数均大于等于0.70,说明SWAT模型对李子溪流域年、月径流和年泥沙量的模拟精度较高。同时模拟值与实测值和降雨量的变化趋于一致。可见,用SWAT模型模拟和预测雨量较为丰沛、土壤侵蚀较严重的紫色丘陵地区的产流产沙是实用、可行的。     

黑河山区流域平均坡长的计算与径流模拟

 流域平均坡长是侧向流和汇流时间计算的重要参数,其会影响地表径流的计算。应用SWAT 2005和ArcView GIS 3.2集成的AVSWAT模型,对黑河干流上游山区流域莺落峡出山口径流进行模拟,发现其计算的流域平均坡长存在较大误差,进而影响到模拟结果。利用子流域内已知的平均坡度和平均坡长建立回归方程,计算各子流域的平均坡长,替换AVSWAT计算的不合理值,在保持其他参数不变的情况下,模拟的月径流纳什系数从0.60提高到0.75,模拟结果得到显著提高。敏感性分析结果和径流曲线数(CN2)的分析也间接验证了流域平均坡长修正方法的可行性。在修正流域平均坡长后,对AVSWAT模型的其他参数CN2等进行优化,模拟的月径流的纳什系数达到0.81,表明本文建议的流域平均坡长计算方法是可以应用到实际的干旱区黑河流域并取得较好模拟效果。     

基于ArcMap的长期水文影响评价模型二次开发

长期水文影响评价模型(Long-term Hydrological Impact Assessment,L-THIA)是研究人类活动对流域非点源影响的重要评价模型之一。本文深入阐明了L-THIA模型的原理,探讨了SCS CN方法和污染物负荷的具体计算方法。在此基础上,提出了L-THIA模型存在的不足和改进方向,并以ArcMap为平台,对该模型进行了二次开发,二次开发后的模型主要包括六个模块,分别为土地利用分类模块、土壤分类模块、CN值计算模块、径流深度计算模块、径流量计算模块和污染物负荷计算模块。基于ArcMap平台的L-THIA模型利于用户方便处理数据,且提高了模型模拟的准确度,实现了模型本地化。     

基于小波神经网络的盲区重力数据预测

引入精度较高的小波神经网络预测局部盲区重力数据。通过实验,对比分析了不同地形数据分布状况下经度、纬度二维输入和经度、纬度、高程三维输入小波神经网络盲区重力数据预测精度,发现利用小波神经网络可以较好地实现高精度盲区重力数据的预测,同等地形条件下三维输入预测的数据精度更高,更有利于后期利用插值方法制备高精度重力基准图。同时,利用克里金法和小波神经网络对真实盲区进行预测。结果表明,两者的预测值相差不大,但小波神经网络的预测结果更为平滑,表明小波神经网络有良好的适应性。     

基于CUDA的地表水动态模拟并行方法

如何高效地动态监测、模拟和预测地表水过程是防灾减灾中亟待解决的问题,也是科学化的国土整治、区域规划、环境保护和水资源管理的基础。因此,本文利用统一计算设备架构(CUDA)对基于不规则三角网(TIN)的地表水动态模拟算法进行并行化改进,提出了一种基于CUDA的地表水动态模拟并行方法,旨在对任意时刻的地表水进行快速、高精度的动态模拟,从而满足实际的应用需求。该算法从高精度的数字高程模型(DEM)中提取地形特征点和流域线,生成受流域线约束的TIN。在此基础上,根据TIN表面的三角面坐标数据获取水流方向,再结合任意位置的降雨源点追踪得到流水线网络。基于曼宁公式,利用流水线流速计算核函数得到每条流水线上雨滴的流速,结合预设的时间,利用汇流量统计核函数得到该时刻的地表汇流量。具体的并行化过程包括数据的传输,CUDA中线程的划分和流水线流速计算核函数和汇流量统计核函数的实现。本文选取位于加拿大新布伦瑞克西北部的BBW(Black Brook Watershed)流域作为实验样区,将该算法与改进前的方法进行对比表明,该模型在保持精度的同时,大幅提高了模拟效率,加速比达到11.2;另外,通过与SWAT(Soil and Water Assessment Tool)模型进行对比,结果表明,该并行方法的Nash系数提高了88%,相关系数提高了5%。     

降雨径流与洪水淹没模型耦合的应用研究

洪水研究包括径流与淹没两种模式。为了探究流域降雨产汇流与淹没情况、提高洪水预报精度,本研究在传统流域水文模型的基础上耦合二维水动力学模型,建立水文-水动力耦合模型。以我国吉林温德河流域为研究实例,模拟了2017年“7·13”洪水在下游口前镇所处子流域洪水淹没过程。首先对基础数据进行预处理,建立HEC-HMS水文模型并进行参数优化后,最终获得流量过程水文结果作为水动力学模型边界条件,之后建立HEC-RAS二维水动力学模型对重要子流域进行淹没模拟。耦合模型计算结果显示,水文模型经多参数优化流量模拟的NSE系数为0.988,水动力计算最大淹没水深达9.3 m相对误差为-5.2%。从泛洪模拟结果来看,子流域上游部分的农田大量被淹,淹没水深范围在0.5~2.0 m,平均流速基本在1 m/s以下。下游口前镇内最大淹没水深接近1 m,水流速度0.2 m/s至1.5 m/s,与实际的淹没情况相吻合。研究表明,所建水文水动力耦合模型模拟计算的结果准确率较高,对具有复杂水文、水力条件的流域的洪水预报具有重要的指导意义。     

改进的HASM-AD算法及在空间变量模拟的应用分析

高精度曲面建模(HASM)可以显著提高空间曲面模拟的精度,但是计算速度低限制了该模型的进一步应用。为了提高HASM模型的计算效率,本文对HASM-AD算法作了改进,通过在计算过程中为采样点添加索引,避免了计算过程中对采样点信息的重复查找操作;同时,在遍历独立计算单元时实时计算第一类、第二类基本量及克式符号,避免了全局存储上述变量所需要的额外内存消耗。数值试验表明,由于将全局线性方程组求解问题转化为局部独立计算单元(5×5栅格)内的方程组求解,改进的HASM-AD算法显著提高了计算效率,同时降低了模型运行过程中的内存消耗。最后,本文以全国陆地降水空间分布模拟分析作为实例,验证改进的HASM-AD算法模拟精度及计算效率,模拟结果表明,改进的HASM-AD算法模拟结果精度优于其他HASM算法(以HASM-PCG为例),并且计算效率优势更为明显,实现全国10km分辨率的降水分布模拟耗时仅为4s。表明改进的HASM-AD算法提高了计算速度,并且适于大尺度的空间变量模拟应用。     

Application of swat model in the upstream watershed of the Luohe River

1INTRODUCTIONIntheHuanghe(Yellow) Riverbasin, soilerosionisaseriousproblem,whilerunoffandsedimentyieldsim-ulation hasnotbeenextensivelystudiedonthebasisofGIS(GeographicInformationSystem) and dis-tributedhydrologicalmodel.Inthisstudy,theLushiwatershed,whichislocatedattheupstreamoftheLushiHydrologicalStationintheLuoheRiver—thebiggesttributary oftheHuanghe Riveranddown-streamofXiaolangdiDam,isselectedasthestudyarea.ThelevelofsoilerosioninLushiwatershedismoderatein theHuangheRiverbas…     

Energy Balance-Based SWAT Model to Simulate the Mountain Snowmelt and Runoff——Taking the Application in Juntanghu Watershed(China) as an Example

In order to predict long-term flooding under extreme weather conditions in central Asia, an energy balance-based distributed snowmelt runoff model was developed and coupled with the Soil and Water Assessment Tool(SWAT) model. The model was tested at the Juntanghu watershed on the northern slope of the Tian Shan Mountains, Xinjiang,China. We compared the performances of temperature-index method and energy balanced method in SWAT model by taking Juntanghu river basin as an application example(as the simulation experiment was conducted in Juntanghu River, we call the energy balanced method as SWAT-JTH). The results suggest that the SWAT snowmelt model had overall Nash-Sutcliffe efficiency(NSE) coefficients ranging from 0.61 to 0.85 while the physical based approach had NSE coefficients ranging from 0.58 to0.69. Overall, on monthly scale, the SWAT model provides better results than that from the SWAT-JTH model. However, results generated from both methods seem to be fairly close at a daily scale. Thestructure of the temperature-index method is simple and produces reasonable simulation results if the parameters are well within empirical ranges. Although the data requirement for the energy balance method in current observation is difficult to meet and the existence of uncertainty is associated with the experimental approaches of physical processes, the SWAT-JTH model still produced a reasonably high NSE. We conclude that using temperature-index methods to simulate the snowmelt process is sufficient, but the energy balance-based model is still a good choice to simulate extreme weather conditions especially when the required data input for the model is acquired.     

Effects of spatial information of soil physical properties on hydrological modeling based on a distributed hydrological model

The spatial distribution of soil physical properties is essential for modeling and understanding hydrological processes. In this study, the different spatial information (the conventional soil types map-based spatial information (STMB) versus refined spatial information map (RSIM)) of soil physical properties, including field capacity, soil porosity and saturated hydraulic conductivity are used respectively as input data for Water Flow Model for Lake Catchment (WATLAC) to determine their effectiveness in simulating hydrological processes and to expound the effects on model performance in terms of estimating groundwater recharge, soil evaporation, runoff generation as well as partitioning of surface and subsurface water flow. The results show that: 1) the simulated stream flow hydrographs based on the STMB and RSIM soil data reproduce the observed hydrographs well. There is no significant increase in model accuracy as more precise soil physical properties information being used, but WATLAC model using the RSIM soil data could predict more runoff volume and reduce the relative runoff depth errors; 2) the groundwater recharges have a consistent trend for both cases, while the STMB soil data tend to produce higher groundwater recharges than the RSIM soil data. In addition, the spatial distribution of annual groundwater recharge is significantly affected by the spatial distribution of soil physical properties; 3) the soil evaporation simulated using the STMB and RSIM soil data are similar to each other, and the spatial distribution patterns are also insensitive to the spatial information of soil physical properties; and 4) although the different spatial information of soil physical properties does not cause apparent difference in overall stream flow, the partitioning of surface and subsurface water flow is distinct. The implications of this study are that the refined spatial information of soil physical properties does not necessarily contribute to a more accurate prediction of stream flow, and the selection of appropriate soil physical property data needs to consider the scale of watersheds and the level of accuracy required.     

A Modified Groundwater Module in SWAT for Improved Streamflow Simulation in a Large,Arid Endorheic River Watershed in Northwest China

Interactions between surface water and groundwater are dynamic and complex in large endorheic river watersheds in Northwest China due to the influence of both irrigation practices and the local terrain. These interactions interchange numerous times throughout the middle reaches, making streamflow simulation a challenge in endorheic river watersheds. In this study, we modified the linear-reservoir groundwater module in SWAT (Soil and Water Assessment Tools, a widely used hydrological model) with a new nonlinear relationship to better represent groundwater processes; we then applied the original SWAT and modified SWAT to the Heihe River Watershed, the second largest endorheic river watershed in Northwest China, to simulate streamflow. After calibrating both the original SWAT model and the modified SWAT model, we analyzed model performance during two periods: an irrigation period and a non-irrigation period. Our results show that the modified SWAT model with the nonlinear groundwater module performed significantly better during both the irrigation and non-irrigation periods. Moreover, after comparing different runoff components simulated by the two models, the results show that, after the implementation of the new nonlinear groundwater module in SWAT, proportions of runoff components changed-and the groundwater flow had significantly increased, dominating the discharge season. Therefore, SWAT coupled with the non-linear groundwater module represents the complex hydrological process in the study area more realistically. Moreover, the results for various runoff components simulated by the modified SWAT models can be used to describe the hydrological characteristics of lowland areas. This indicates that the modified SWAT model is applicable to simulate complex hydrological process of arid endorheic rivers.     

青藏高原兹格塘错流域50年来湖泊水量对气候变化响应的模拟研究

运用数值模拟建立青藏高原兹格塘错流域土壤、植被、气候等的空间和属性数据库;接着,借助分布式流域尺度水文模型(SWAT模型),对兹格塘错1956—2006年间的流量进行模拟实验;最后,反演50年来兹格塘错流域水文过程,测试流域温度、降水和蒸发组合的敏感因子对湖泊水量变化的效应,探讨50年来湖泊水量对气候变化的响应。模拟实验的边界条件设置为自然地形、土壤、植被覆盖,其中土壤资料包括有机质含量、粒径等理化参数。模拟结果表明:兹格塘错的年平均流量为6.3m3/s,流量高峰集中在8月至10月,并且由于融雪补给的关系,3月出现另一个流量高峰;模拟结果与遥感解译所得到的结果吻合较好。敏感实验表明:兹格塘错流域内温度、降水和蒸发组合的敏感因子实验具有高原特征,即高原湖泊的水文过程和湖泊流量变化有着较为敏感的响应关系;兹格塘错流量受降水的影响最大,随着降水的增加,流量有所增加;在温度升高的情况下,流域蒸发量增加速度大,兹格塘错流量增加的效应不明显,而在冷湿模式下,流域蒸发量降低,兹格塘错流量增加显著。     

城市暴雨内涝模拟模型优化与精度验证

本文采用“Vegetation-Impervious Surface-Soil”模型和线性光谱混合分解方法,获取像元中不透水面、植被、土壤覆盖信息,用于计算SCS模型产流参数综合CN（Curve Number）值;基于土地利用类型,采用经验值与数值实验逐步求精相结合的方法,确定水动力汇流模型参数曼宁系数,并用实测积水数据验证两次参数修正的模拟效果。以上海中心城区为例进行验证,研究结果表明：①将采用V-I-S模型得到的不透水面、植被、土壤信息设定CN值,能够降低积水分布的极值化现象,提高SCS产流模型产流量和产流分布精度;②采用经验法和数值模拟逐步求精法,按土地利用类型设定曼宁系数,使各时段最大积水深度高于原模型,说明曼宁系数是汇流模型的敏感参数。     

福建东圳库区淹没及水土流失的3D-VRGIS构建与应用

本文应用VRGIS(Virtual Reality Geographical Information System)技术,建立水库工程水土流失三维可视化动态仿真模型,实现库区的空间信息可视化。以福建省东圳水库为研究对象,建立水土流失区域三维虚拟地理信息系统(3D-VRGIS),生成水土流失现状以及库区生态环境三维场景。并结合水土流失中水淹分析要素与三维空间的数据可视化技术,实现水土流失监测预报及其区域定位监控的可视化处理,为减少水土流失危害,提供可靠的信息资源。     

Hydrological Simulation Using TRMM and CHIRPS Precipitation Estimates in the Lower Lancang-Mekong River Basin

Satellite-based products with high spatial and temporal resolution provide useful precipitation information for data-sparse or ungauged large-scale watersheds. In the Lower Lancang-Mekong River Basin, rainfall stations are sparse and unevenly distributed, and the transboundary characteristic makes the collection of precipitation data more difficult, which has restricted hydrological processes simulation. In this study, daily precipitation data from four datasets(gauge observations, inverse distance weighted(IDW) data, Tropical Rainfall Measuring Mission(TRMM) estimates, and Climate Hazards Group InfraRed Precipitation with Stations(CHIRPS) estimates), were applied to drive the Soil and Water Assessment Tool(SWAT) model, and then their capability for hydrological simulation in the Lower Lancang-Mekong River Basin were examined. TRMM and CHIRPS data showed good performances on precipitation estimation in the Lower Lancang-Mekong River Basin, with the better performance for TRMM product. The Nash-Sutcliffe efficiency(NSE) values of gauge, IDW, TRMM, and CHIRPS simulations during the calibration period were 0.87, 0.86, 0.95, and 0.93 for monthly flow, respectively, and those for daily flow were 0.75, 0.77, 0.86, and 0.84, respectively. TRMM and CHIRPS data were superior to rain gauge and IDW data for driving the hydrological model, and TRMM data produced the best simulation performance. Satellite-based precipitation estimates could be suitable data sources when simulating hydrological processes for large data-poor or ungauged watersheds, especially in international river basins for which precipitation observations are difficult to collect. CHIRPS data provide long precipitation time series from 1981 to near present and thus could be used as an alternative precipitation input for hydrological simulation, especially for the period without TRMM data. For satellite-based precipitation products, the differences in the occurrence frequencies and amounts of precipitation with different intensities would affect simulation results of water balance components, which should be comprehensively considered in water resources estimation and planning.