基于GIS考虑准动态湿度指数的滑坡危险性预测水文-力学耦合模型研究

降雨在区域浅层滑坡灾害危险性预测研究中起着关键作用。为综合斜坡地形因素和由降雨引起的地表水下渗、地下水径流对斜坡稳定性的影响,结合基于物理过程的无限斜坡模型和基于数字高程模型 DEM 的简化运动波模型,提出考虑准动态湿度指数的滑坡危险性预测水文-力学耦合模型。该模型基于地理信息系统GIS平台,采用矢量-栅格复合单元对斜坡进行危险性分析(即以斜坡单元为基本研究对象,用栅格数据进行单个斜坡单元稳定性分析)。首先,基于极限平衡理论推导无限斜坡模型;然后,在考虑降雨强度以及持续时间的情况下应用简化的运动波模型计算降雨入渗和地下水径流作用过程中的地形准动态湿度指数,得到滑坡土体饱和因子在时间和空间上的分布情况,并在此基础上实现水文模型与无限斜坡模型的耦合;最后,以三峡库区巴东新城区滑坡灾害危险性预测为例,验证模型在区域浅层滑坡灾害危险性预测中具有较高的精度。     

基于DM-LSTM的城市降雨径流预测研究

智慧水务背景下,如何基于人工智能理论与技术深化城市降雨径流模型研究,是一项值得探索的课题。由于城市降雨径流时间分辨率高且样本特征分布不具有规律性,直接采用长短期记忆（LSTM）模型进行预测面临着挑战。基于此,提出用数据挖掘（DM）算法及规则对城市降雨径流时序数据集进行聚类和重构,并基于深度学习算法对LSTM模型的结构和参数进行优化,构建了DM-LSTM耦合模型,并用于研究区域的降雨径流模拟。结果表明,对于各类降雨事件,与LSTM模型相比,DM-LSTM耦合模型的均方根误差（RMSE）降低了2.1%～41.9%,纳什效率系数（NSE）提高了0.4%～56.4%,决定系数（R2）提高了0.3%～65.6%。DM-LSTM耦合模型不仅对各类降雨事件均表现出更好的预测性,而且模型运行时间仅为2.044 s,能够很好地满足城市降雨径流预测对实时性、准确性和稳定性的需求。     

低影响开发设施组合的水文模拟及不确定性分析

在海绵城市建设中,通过多种低影响开发(LID)设施的组合应用可实现对城市雨水径流的调控。针对LID设施组合的水文模拟存在高维参数估值不确定性分析的难题,以深圳环保产业园LID示范区为例,构建LID组合设施的暴雨洪水管理模型(SWMM),结合Morris筛选法与DREAM算法,对不同降雨条件下模型中不同设施、不同类型参数的灵敏度和不确定性进行分析。结果表明,位于下游的设施参数灵敏度普遍高于上游设施,排水及导水相关参数灵敏度相对较高;下游设施的排水参数后验分布峰值明显,参数不确定性小;中等强度降雨事件模拟结果的不确定性比高强度降雨事件要小。     

降雨影响下的区域滑坡危险性动态评价研究——以三峡库区万州主城区为例

 降雨引发的滑坡具有区域性的群发效应,能够在短时间内造成大量的灾难性损失。基于此,提出一种可考虑不同降雨期影响的区域滑坡危险性评价方法。该方法以瞬态降雨入渗的区域斜坡稳定性计算模型为基础,将滑坡危险性定义为在一定持续降雨期内各栅格单元体失稳的概率。通过岩土体物理力学参数的不确定性进行各栅格单元体失稳概率的求解,继而获得区域内滑坡的危险性分布。基于ArcGIS软件开发出区域滑坡危险性动态评价工具。以三峡库区万州主城区为例,详细介绍危险性评价工具的数据处理过程以及参数选取方法,并以2种不同的降雨工况进行比较计算。现场斜坡稳定性的调查与计算结果的对比及统计分析表明：滑坡的危险性分布图与真实滑坡的稳定性情况基本一致,并在一定程度上反应了该地区斜坡稳定性的时空分布特征,测试并验证了评价工具的正确性。     

城市雨水水质模拟研究进展

降雨径流污染是城市面临的重要环境问题之一,雨水水质模型可用于支持雨水管理政策和技术的实施。雨水水质的模拟与预测需要更好地理解与污染物相关的物理、化学和生物过程以及模型的复杂性。笔者分析了雨水主要污染物、污染过程及BMP/LID去除效果的模拟特征与函数表达,介绍和比较了6种常用确定性雨水水质模型的特点和复杂性。提出了模型研究的发展空间:扩大模拟污染物的范围;逐步细化代表不同污染物和BMP/LID的函数表达;建立适合不同区域、场景的参数数据库和识别方法;减小监测数据和参数校验对模拟结果的影响。     

以径流污染为指标的降雨重要性分类

根据径流冲刷污染量,对场次降雨重要性提出划分标准,可以有重点地开展降雨径流污染监测工作,也可降低实际工程中对多年详细场次降雨资料的依赖。在长时间序列降雨资料的基础上,根据降雨径流污染程度的不同对降雨分类,找出径流污染较大的降雨(即重要降雨)。根据各类降雨特征参数(干旱时间和场次降雨量)的不同,用模糊C均值聚类法把降雨分为3类,分别为重要降雨、一般降雨和不重要降雨。研究提出降雨重要性分类方法,并应用于深圳市WH河流域案例。结果表明,绘制重要性不同降雨分布图,可以相对划分各类降雨分布区域,建立分类标准,处于模糊区域的降雨则需结合其他因素判别。     

排水管网模型参数率定及分析

选择JZ市老城区某区域为研究对象,建立排水管网Info Works CS模型,基于实际降雨事件数据,率定降雨径流模型参数,并对主要参数进行敏感度分析。结果表明,影响径流洪峰量的最敏感参数是Horton模型的稳渗率,不透水表层的固定径流系数次之,且初渗率和初期损失值的敏感度最小。研究成果可为城市排水管网模型建立及参数选择提供理论依据和技术支撑。     

景观指数评价小区LID设计方案效果的适用性研究

构建某小区SWMM模型,基于年径流总量控制率计算小区低影响开发(LID)设施总调蓄容积,并按不同比例分配至LID设施,形成多套LID设计方案,模拟计算设计方案在不同降雨情景下的场次降雨径流控制效果。选取6种景观指数,分析其与设计方案场次降雨径流控制率的相关关系,探讨景观指数评价LID设计方案效果的适用性。结果表明,香农多样性指数、香农均匀度指数与场次降雨径流控制率有较强的正相关性,斑块密度、分割度、最大斑块指数、斑块面积变异系数在特定情形下与场次降雨径流控制率有相关性,该结论可为小区尺度LID设计方案径流控制效果的简易预测提供参考。     

基于GIS和SWMM的精细化水文表征方法

SWMM是模拟城市降雨径流响应的动态模型，其汇水区的空间元素细分方式和水文表征直接影响模拟结果。为此，提出了基于GIS对汇水区的土地利用进行精细化分类的方法，并将下垫面信息通过物理水文定义，反馈、模拟到SWMM的汇水区水文表征和低影响开发（LID）模块描述中，直接影响水文汇流过程。与常规水文构建方法相比，精细化模型中增加的缓冲渗透区可以接收来自间接不透水区的径流，更符合实际径流走向，提高了模型精度。在相同降雨重现期下，精细化模型模拟得到的径流总量、径流峰值和溢流总量均比常规模型要小。同时，该方法可以更合理地模拟LID措施的水文作用，可为LID措施提供因地制宜的布设空间和比例。     

基于SWMM的海绵城市小区雨洪控制效果评价

"海绵城市"的提出为解决城市内涝问题提供了新的思路和方法。武汉市为全国"海绵城市"建设首批试点城市之一,试点建设还处于探索阶段,雨洪控制效果未得到验证。笔者以武汉市青山某小区为研究对象,通过SWMM建立海绵城市小区模型,针对汇水区的划分、海绵设施的定义、参数率定及模型验证、设计降雨的选择等方面作了研究。结果表明:相比传统小区,海绵城市小区能够有效削减雨水年径流总量和典型场降雨的峰值流量、径流总量,但对洪峰迟滞作用并不明显。     

Generalising relationships between runoff–rainfall coefficients and impervious areas: an integration of data from case studies in Israel with data sets from Australia and the USA

This study explores the generalised relationships between the proportion of impervious areas and storm runoff coefficient (RR) that characterises the current global trends of expansion and densification of urban zones. For this purpose there is an integrated database representing about 800 urban storm events in diverse locations around the world: from Israel and from detailed rainfall–runoff measurements in Australia and USA. Special attention is given here to the experience gathered in two case studies in Israel with information of surface imperviousness derived from remote sensing data, including air photographs and satellite images. Despite differences in measurement systems and methodologies between sites, there was evidential support for a new generalised three-phase model where in up to 20% of impervious area there are no expected major runoff events, a major enhancement in runoff–rainfall response to impervious area increase between 20% and 40%, and 1:1 relationship between 40% and 100% impervious area. This model may be implemented on a wide regional scale, based on remote sensing derived impervious areas for mapping urban areas with high flooding risk. Urban planning may incorporate this model for determining expected storm runoff levels and incorporating this information to derive appropriate hydrological solutions.     

Spatial variations of storm runoff pollution and their correlation with land-use in a rapidly urbanizing catchment in China

The composition of land use for a rapidly urbanizing catchment is usually heterogeneous, and this may result in significant spatial variations of storm runoff pollution and increase the difficulties of water quality management. The Shiyan Reservoir catchment, a typical rapidly urbanizing area in China, is chosen as a study area, and temporary monitoring sites were set at the downstream of its 6 sub-catchments to synchronously measure rainfall, runoff and water quality during 4 storm events in 2007 and 2009. Due to relatively low frequency monitoring, the IHACRES and exponential pollutant wash-off simulation models are used to interpolate the measured data to compensate for data insufficiency. Three indicators, event pollutant loads per unit area (EPL), event mean concentration (EMC) and pollutant loads transported by the first 50% of runoff volume (FF50), were used to describe the runoff pollution for different pollutants in each sub-catchment during the storm events, and the correlations between runoff pollution spatial variations and land-use patterns were tested by Spearman's rank correlation analysis. The results indicated that similar spatial variation trends were found for different pollutants (EPL or EMC) in light storm events, which strongly correlate with the proportion of residential land use; however, they have different trends in heavy storm events, which correlate with not only the residential land use, but also agricultural and bare land use. And some pairs of pollutants (such as COD/BOD, NH₃-N/TN) might have the similar source because they have strong or moderate positive spatial correlation. Moreover, the first flush intensity (FF50) varies with impervious land areas and different interception ratio of initial storm runoff volume should be adopted in different sub-catchments.     

KINEROS2 application for land use/cover change impact analysis at the Hulu Langat Basin,Malaysia

The impacts of land use/cover changes (LUCC) on a developed basin in Malaysia were evaluated. Three storm events in different intensities and durations were required for KINEROS2 (K2) calibration and LUCC impact analysis. K2 validation was performed using three other rainfall events. Calibration results showed excellent and very good fittings for runoff and sediment simulations based on the aggregated measure. Validation results demonstrated that the K2 is reliable for runoff modelling, while K2 application for sediment simulation was only valid for the period 1984–1997. LUCC impacts analysis revealed that direct runoff and sediment discharge increased with the progress of urban development and unmanaged agricultural activities. These observations were supported by the NDVI, landscape and hydrological trend analyses.     

Areal rainfall intensity distribution over an urban area and its effect on a combined sewerage system

The application of 21 pluviometers over the whole area of ?ód?, Poland, enabled an examination of areal distribution of intense rainfalls during 2010–2011. On the basis of collected data the distribution of rainfall intensity was analysed and a circular zone model with exponential shape in all directions has been proposed for describing rain storms. Also, the velocity of storm zone movement between 2 and 30 km/h was determined. Using the SWMM 5.0 software, adapted for non-standard use, runoff from three real catchments and combined storm overflow discharge was examined for selected measured rainfalls. An effect of storm zone range and movement was also studied for the same catchments using model storms. It was confirmed that using data from the city pluviometric network enables significantly better simulation results for CSO overflows to be obtained compared to an option based on a single pluviometric station, especially for large catchments greater than 200 hectares.     

Regression models for estimating storm runoff load and its application to Lake Kasumigaura

Regression models are proposed for accurate estimation of storm runoff load. Regression equations are obtained for the relationship between final cumulative load and flow during direct runoff by each storm event. The models are applied to estimate annual nutrients load by all influent rivers into the Lake Kasumigaura for a model year of average annual rainfall. The load in dry weather is calculated from observed data of all influent rivers. Total annual load for a year is the sum of loads in dry weather days and loads in wet weather days. The ratios of the load in a wet weather day to total annual loads are 29 percent for T‐N, 51 percent for T‐P, 53 percent for T‐COD, 22 percent for D‐N, 22 percent for D‐P and 30 percent for D‐COD.     

山地城市小流域暴雨径流模型的研发及应用

采用ArcEngine二次开发包结合水力学基础算法,建立山地城市暴雨径流模型,以模拟山地城市降雨径流,分析山地城市排水系统的排放能力。应用该模型对重庆市北部新区盘溪河流域排水系统进行了预测和评估,结果表明,模拟结果与实测数据的效率系数Nash-Suttcliffe系数Ens为0.56~0.76,相对误差8.82%~11.8%。模型关键参数率定及敏感性分析表明,最敏感参数为径流宽度、坡度、曼宁系数和不透水面积率。针对排水管道内流量、充满度和溢流检查井个数等方面进行了排水系统排放能力的评估,结果表明,一年一遇暴雨下有0.52%的排水管道长时间处于满流状态,五十年和百年一遇的超载排水管道占排水管道总数5.86%和 8.20%;百年一遇暴雨下有0.72%的检查井发生溢流,一年一遇暴雨下有0.31%的检查井发生溢流。该模型实现了对山地城市暴雨径流的产汇过程模拟以及内涝点位置和数量的识别,为城市雨水管网管理提供决策支持。     

A Microcomputer-Based Geographic Information System for Hydrologic Simulation

Abstract: A microcomputer-based geographic information system (GIS) supporting hydrologic simulation is described. The hydrologically-oriented GIs can automatically provide parameters required for prediction of flood hydroghraphs and physically-based synthetic flood frequency curves for ungaged watersheds. The hydrologic model uses the Soil Conservation Service (SCS) runoff curve number (CN) method and a synthetic unit hydrograph to represent the physical transformation of rainfall into runoff The standard SCS method was modified to account for randomness of key parameters whose probability distributions can be derived from regional rainfall and runoff data. The GIs, and hydrologic theory and techniques, including the modified SCS method, derivation of a regional dimensionless unit hydrograph, and Monte Carlo simulation procedure, are described. Results of hydrologic analysis and simulation by the system implemented for an area of 11000 km² in the Alberta foothills are presented.     

The role of inter-event time definition and recovery of initial/depression loss for the accuracy in quantitative simulations of highway runoff

A long-term runoff monitoring was carried out in a highway drainage system in Winterthur, Switzerland. Several runoff quantity simulations were carried out using the distributed model ‘InfoWorks-CS’. Serial rainfall monitoring data were used for investigation of water runoff behaviour from the highway. The inter-event time definition was calculated from the runoff simulation and applied for better accurate runoff volume determination. Under continuous rainfall condition, the quantity simulation showed a good agreement with the measured hydrograph. However, in some cases where rainfall was not continuous, overestimation of runoff volume and difference in peak height/timing was found after rainfall halting. It was believed that the initial/depression loss on road surface was recovered during the halting period of rainfall and caused the difference in the runoff volume and peak height/timing difference in simulation. The consideration of regenerated depression loss significantly improved runoff simulation results in the on-and-off type rainfall events. This result showed that it was essential to estimate the inter-event time definition to consider appropriateness of the initial or intermediate loss.     

Finite-Element Modeling of Storm Water Runoff Using GRASS GIS

Abstract: This paper presents methodologies used in the model r.water.fea for simulating storm water runoff using the finite-element method and a raster geographic information system (GIS). The internally integrated model r.water.fea simulates storm water runoff by interfacing the raster data structure with the finite-element solution of the kinematic wave equations. The integrated system offers many advantages because the GIS and simulation model exchange input and output data internally. Distributed simulation of the spatially and temporally variable hydrologic process is possible with r.water.fea. Storm water runoff simulated with r.water.fea compares favorably with analytic solutions for a simple domain. An example simulation of a catchment demonstrates the capabilities of r.water.fea.     

Assessing methods for predicting green roof rainfall capture: A comparison between full-scale observations and four hydrologic models

To optimize the application of green roof technology, there is a need to quantify stormwater mitigation in advance of green roof construction. This study contributes toward meeting this need by assessing the utility of four hydrologic models for predicting green roof rainfall capture, including the: (1) curve number method, (2) characteristic runoff equation, (3) Hydrological Evaluation of Landfill Performance (HELP V3.9D) model, and (4) Storm Water Management Model (SWMM V5.1). Modeling results were compared to over twenty-four months of observed runoff data, collected between June 2011 and December 2013, from two full-scale green roofs in New York City (NYC). Both the curve number method and characteristic runoff equation had the highest Nash-Sutcliffe efficiency index (NSEI) between modeled and observed cumulative runoff depth per event (NSEI = 0.97) due to parameter calibration requirements, where error was mainly due to variations in green roof antecedent moisture conditions. The HELP model was originally intended for evaluation of a continuous landfill cover. As a result, HELP's inability to account for the non-vegetated areas on green roofs caused underestimation of runoff depth for most events (NSEI = 0.84). Alternatively, the SWMM model tended to overestimate event runoff depth (NSEI = 0.94), thought to be the result of its storage term parameterization. Model assessments point to the need for more robust parameter estimation methods, particularly for inputs that are statistical or difficult to measure directly, to improve pre-development accuracy of green roof performance models.