1、广西水利水电 GUANGXI WATER RESOURCES&HYDROPOWER ENGINEERING 2023(3)1流域概况漓江上游流域位于桂林水文站以上,集水面积2785 km2,平均海拔150 m,河段干流长105 km,其中,流经市区河段长4.93 km。流域地势由西北向东南倾斜,呈北高南低,特殊的地势使得流域内降水比较丰沛,该流域上游是广西三大暴雨中心之一。研究区属亚热带季风气候,多年平均降雨量1 889.4 mm,汛期在38月,汛期降雨量占年降水量的75%左右,汛期内流域降水较多,从而导致流域内水灾频繁。本文以漓江上游流域为研究对象,基于HEC-HMS模型构建漓江流域上游科学
2、合理的洪水预报方案,为该流域防洪减灾提供理论与方法支持。2研究方法与模型构建HEC-HMS模型是美国陆军工程师团水文工程中心开发的一个基于物理机制的半分布式水文模型。该模型充分考虑了降雨时空分布不均匀性和流域下垫面空间变异性,可以更好地反映产汇流规律,在中小流域有很好的适用性1。HEC-HMS模型克服了概念性流域水文模型在产流、坡面汇流、基流与河道汇流计算方面的单一性,提供了多种模拟方法,可根据实际情况进行优化组合,提高模拟计算的效率和精确度;同时具有一系列强大的水文模拟功能,能够模拟不同时间尺度的降雨径流过程,在洪水预警系统中得到了广泛应用2,3。HEC-HMS模拟要素分为产流计算、坡面汇流
3、、基流计算和河段上的洪水演算。本文产流计算、坡面汇流分别采用SCS-CN曲线法和斯奈德(Snyder)单位法,基流计算和河段上的洪水演算分别采用退水曲线法和马斯京根法。2.1研究方法2.1.1产流计算HEC-HMS提供的径流计算方法有Green&Ampt入渗曲线法、SMA入渗曲线法,以及Deficit and Constant入渗曲线法等,但这些方法涉及的参数多,而且参数不易获取,限制了它们的广泛应用。美国农业部土壤保持局提出的径流曲线数法(SCS-CN),该方法参数少,计算过程简单,所需的资料易于通过多种手段获取,且能反映不同土壤类型、不同土地利用方式、坡度、前期土壤湿润程度等下垫面因素以及
4、地表参数变化对水文过程的影响,对降雨观测的要求不严格,因此被广泛应用于估算各种地表覆被条件下的地表径流4。相关计算公式如下:收稿日期2023-01-03基金项目广西重点研发计划项目(桂科AB19259015、桂科AB20159003)作者简介冯世伟(1989-),男,广西玉林人,广西壮族自治区水利科学研究院工程师,硕士,主要从事水文水资源研究、规划与咨询工作。水文水能 HEC-HMS模型在漓江上游流域洪水预报中的应用冯世伟1,邓芳芳2,李静3(1.广西壮族自治区水利科学研究院广西水工程材料与结构重点实验室,南宁530023;2.广西科源工程咨询有限责任公司,南宁530023;3.广西外国语学院
5、,南宁530023)摘要漓江上游流域洪水灾害频发,是广西山洪灾害多发区中较严重的流域之一。本文以漓江上游流域为研究对象,通过HEC-HMS水文模型,采用SCS-CN法计算降水损失、Snyder单位线法计算坡面汇流、退水曲线法模拟基流、马斯京根法进行河道洪水演算等4种研究方法模拟降雨-径流过程,基于实测水文、气象数据资料选取20082013年的27场洪水作为率定期洪水,20142015年8场洪水作为验证期洪水进行降雨径流模拟,经过模拟计算得到结果:率定期27场洪水合格率为81.5%,平均确定性系数为0.809;验证期8场洪水合格率为75%达到乙级精度,平均确定性系数为0.76,达到乙级精度。模拟
6、结果的分析表明,HEC-HMS模型适用于漓江流域上游且取得较好的模拟结果,可为漓江流域上游洪水作业预报提供参考。关键词洪水预报;漓江上游流域;HEC-HMS洪水预报模型中图分类号P338文献标识码A文章编号1003-1510(2023)03-0019-0519R=(P-0.2S)2P+0.8S,P 0.2SR=0,P 0.2S(1)式中:R为径流量,mm;S为流域当时最大可能滞留量,mm;P为降雨量,mm。2.1.2坡面汇流HEC-HMS模型坡面汇流计算的基本方法有两种,即运动波法和单位线法,本文采用斯奈德(sny-der)综合单位线法。标准单位线洪峰流量Up与标准单位线的滞后时间tp存在如下
7、关系:UpA=CCptp(2)其中,tp=(L0.8(S+1)0.7)(1900 Y0.5)(3)式中:Up为标准单位线洪峰流量;A为流域面积;C为转换常数,在国际制单位下取2.75;Cp为单位线洪峰系数;tp为单位线的滞后时间;L为主河道长度;Y为流域坡度;S=1000/CN-10。2.1.3基流汁算本文选择退水曲线理论法来模拟基流,如式(4)所示。Qt=Q0kt(4)式中:Q0为初始流量(t=0);Qt为t时刻基流量;k为衰减系数。2.1.4洪水演算该法是基于水量平衡方程和槽蓄曲线方程的河道流量演算方法。水量平衡方程:t2(I1+I2)-t2(Q1+Q2)=S2-S1(5)槽蓄曲线方程:S
8、=KxI+(1-x)Q(6)联解(5)、(6)两式,可得马斯京根演算方程如下:Q2=C0I2+C1I1+C2Q1(7)C0=0.5t-Kx0.5t+K-KxC1=0.5t+Kx0.5t+K-KxC2=-0.5t+Kx-Kx0.5t+K-Kx(8)C0+C1+C2=1(9)式中:S为河段总蓄量,hm3/s;S1、S2分别为时段起、止河段蓄水量,hm3/s;I1、I2分别为时段起、止上断面入流量,m3/s;Q1、Q2为时段起、止下断面出流量,m3/s;K为稳定流情况下河段传播时间;x为流量比重因素。2.2模型构建根据漓江上游流域数字高程模型(DEM)、土壤和土地利用等数据,利用基于ARCGIS的拓
9、展模块HEC-GeoHMS对DEM进行地形处理,提取漓江上游流域的地表水流路径、河网、流域和子流域边界,计算并提取各子流域地形特征和河流特征参数,并根据子流域划分情况建立研究地区内各水文单元的链接关系,最后生成hms文件。根据流域内的土壤状况,利用spaw软件计算获得土壤饱和导水率后进行水文分组,并结合经过重新分类的土地利用图求取各子流域的CN值5。结合上文提到的研究方法,建立漓江上游流域HEC-HMS水文模型(见图1)。模型将漓江上游流域划分为12个子流域,利用泰森多边形雨量权重法确定各子流域控制雨量站权重。各子流域控制雨量站权重见表1,各子流域特征值见表2。图1漓江上游流域模型概化图表1漓
10、江上游流域各子流域控制雨量站及权重W1W2桂林桂林庙头青狮潭三街10.1070.7570.0900.045子流域编号控制雨量站雨量站权重冯世伟,邓芳芳,李静:HEC-HMS模型在漓江上游流域洪水预报中的应用20广西水利水电 GUANGXI WATER RESOURCES&HYDROPOWER ENGINEERING 2023(3)W3W4W5W6W7W8W9W10W11W12庙头青狮潭砚田三街青狮潭砚田砚田川江华江司门三街砚田川江华江司门川江华江高寨鲤鱼塘清水江司门鲤鱼塘清水江司门清水江严关司门青狮潭砚田司门三街桂林三街桂林三街0.0200.5500.0400.3910.6430.3570.6
11、070.1120.0960.1780.0080.0340.7590.0440.1630.0180.2990.2800.3730.0120.0190.3730.6030.0240.2230.5900.1870.0060.0160.4270.5510.0400.9600.9230.077子流域编号控制雨量站雨量站权重表2漓江上游流域各子流域特征参数子流域编号W1W2W3W4W5W6W7W8W9W10W11W12面积/km215.078331.771310.351451.287270.753183.769330.085173.388233.451310.01386.20460.127占总面积比例/%
12、0.0050.1200.1130.1640.0980.0670.1200.0630.0850.1120.0310.022最大河长/km8.51959.46444.81934.88954.02229.71443.58636.70138.49429.46619.85618.447平均坡度/%9.70813.24115.34433.82439.37734.86136.41331.45918.05824.60913.8546.9493应用成果3.1预报误差与精度评定HEC-HMS模型模拟精度通过合格率、确定性系数、径流深误差、洪峰流量误差、峰现时间误差等5个指标进行综合评价,根据 水文情报预报规范(G
13、B/T 22482-2008),径流深预报以实测值的20%作为许可误差,降雨径流预报以实测洪峰流的20%作为许可误差,峰现时间以预报时间至实测洪峰出现时间之间时距的30%作为许可误差,当许可误差小于3 h或一个计算时段长,则以3 h或一个计算时段长作为许可误差。预报项目的精度按合格率QR或确定性系数DC的大小分为3个等级,预报项目精度等级按表3规定确定。表3预报项目精度等级表精度等级合格率/%确定性系数甲级QR85.0DC0.90乙级85.0QR70.00.90DC0.70丙级70.0QR60.00.70DC0.503.2预报误差与精度评定本文选择了20082015年率定期27场洪水及验证期8
14、场洪水进行降雨-径流模拟,HEC-HMS模型模拟结果见表4,再根据表3进行结果精度评定。从表4可以看出:选出的率定期27场洪水中3个场次未满足径流总深误差要求,1个场次未满足洪峰流量误差要求,2个场次未满足峰现时间误差要求,达到合格标准的有22场,不合格5场,合格率为81.5%,达到了乙级精度;27场洪水的平均确定性系数为0.809,达到乙级精度。验证期8场洪水中1个场次未满足洪峰流量误差要求,2个场次未满足峰现时间误差要求,径流总深误差均满足误差要求,达到合格标准的有6场,不合格2场,合格率为75%,达到乙级精度;8场洪水的平均确定性系数为0.763,达到乙级精度。4结语本文结合DEM与Ar
15、cGIS地理信息系统软件进行子流域划分、子流域面积计算、泰森多边形划分以及雨量权重的计算等,构建了HEC-HMS模型洪水预报方案。对20082013年率定期27场洪水及20142015年验证期8场洪水进行降雨径流模拟,率定期合格率达到了乙级精度,平均确定性系数达21表4HEC-HMS模型模拟结果率定期验证期序号12345678910111213141516171819202122232425262712345678场次20 080 50420 080 60820 080 73020 080 83020 081 10520 090 60820 090 62920 100 41820 100 51
16、820 100 51320 100 61220 100 62420 110 50720 110 52220 110 61220 120 30420 120 42420 120 50220 120 51020 120 60520 120 62320 130 32620 130 40520 130 42920 130 50720 130 51520 130 60520 140 40520 140 70120 140 61520 140 50820 150 50820 150 51120 150 51520 150 608模拟径流总深/mm101.00333.3524.3138.6274.6158.
17、17353.54181.53110.3768.62299.86138.87133.7530.53178.0041.1139.8572.27161.75161.00225.6548.1559.46119.68214.4841.70159.6574.25215.74149.33160.4022.2636.55103.82196.24实测径流总深(mm)97.14282.5021.4337.8759.0264.95328.76163.45109.3064.42291.96136.8995.0033.65168.1836.6251.1573.53150.42137.01188.7950.7467.33
18、113.91193.1148.06161.7683.64203.27139.02171.0524.6840.6295.98213.04径流深误差/%3.9718.0013.001.9826.41-10.447.5411.060.986.522.711.4540.79-9.295.8412.26-22.09-1.717.5317.5119.52-5.10-11.695.0711.07-13.23-1.30-11.236.137.42-6.23-9.81-10.028.17-7.89实测洪峰/(m3/s)87132504627211030613437036001030166036808902060
19、47922306958009861480872232011301260235025607812130639287011802180599113017402270预报洪峰/(m3/s)86032644626921010609377828961086150235011002209038721936627641003145310652223992109322242636777212066229601284165356794517032681洪峰流量误差/%1.26-0.430.004.021.940.6513.5519.56-5.449.524.86-12.58-1.4719.121.664.754
20、.54-1.741.82-22.134.1812.2113.255.36-2.970.540.46-3.52-3.15-8.8324.135.2916.392.13-18.12峰现时间误差/h014-1-1-3201210-43-1-321112101-1-20103125030确定性系数0.8070.8750.6810.9120.7920.8880.8790.9010.8610.8860.9280.7870.4610.8080.8690.7670.6110.7960.8230.6160.8090.7070.8500.9790.8280.7520.9620.7060.9160.8210.640
21、0.5420.8010.7720.903是否合格是是否是否是是是是是是是否是是是否是是否是是是是是是是是是是否否是是是注:表中峰现误差负值表示模拟峰现时间相对实测峰现时间提前,正值表示模拟峰现时间相对实测峰现时间滞后。到了乙级精度;验证期合格率达到乙级精度,平均确定性系数达到乙级精度。模拟结果分析表明,HEC-HMS模型适用于漓江流域上游且取得较好的模拟结果,可为漓江流域上游洪水作业预报提供参考。参考文献1李向新,和红强.HEC-HMS水文建模系统原理 方法 应用M.北京:中国水利水电出版社,2015.2梁睿.HEC-HMS水文模型在北张店流域的应用研究D.太原:太原理工大学,2012.3程文
22、飞,陈军锋,吴博,等.HEC-HMS水文模型在圪洞流域洪水模拟中的应用J.水电能源科学,2018,36(8):52-55.4李润奎,朱阿兴,陈腊娇,等.SCS-CN模型中土壤参数的作用机制研究J.自然资源学报,2013(10):1778-1787.5陈芬,陈兴伟,谢剑斌.HEC-HMS模型次洪模拟的参数敏感性分析及应用J.水资源与水工程学报.2012(5):119-122.(责任编辑:窦波元)冯世伟,邓芳芳,李静:HEC-HMS模型在漓江上游流域洪水预报中的应用22广西水利水电 GUANGXI WATER RESOURCES&HYDROPOWER ENGINEERING 2023(3)Appl
23、ication of HEC-HMS model in flood forecasting for upperreaches of Lijiang River BasinFENG Shi-wei1,DENG Fang-fang2,LI Jing3(1.Guangxi Hydraulic Research Institute,Guangxi Key Laboratory of Water Engineering Materials and Structures,Nanning 530023,China;2.Guangxi Keyuan Engineering Consultant Co.,Ltd
24、.,Nanning 530023,China;3.GuangxiUniversity of Foreign Languages,Nanning 530023,China)Abstract:Flood disasters occur frequently in the upper reaches of Lijiang River Basin,which is one of the relativelyserious river basins with frequent mountain flood disasters in Guangxi.Taking the upper reaches of
25、Lijiang River basin as research object,the rainfall-runoff process was simulated by HEC-HMS hydraulic modeling system,including calculating the precipitation loss by SCS-CN method,calculating the slope runoff by Snyder unit hydrographmethod,simulating the base flow by regression curve method,routing
26、 the river flood by Muskingen method.Basedon measured hydrological and meteorological data,27 flood events from 2008 to 2013 were used as floods of calibration period and 8 flood events from 2014 to 2015 were used as floods of validation period.The results of rainfall-runoff simulation include the 2
27、7 floods of calibration period have a pass rate of 81.5%and an average deterministiccoefficient of 0.809;the 8 floods of validation period have a pass rate of 75%and an average deterministic coefficient of 0.76 up to class B accuracy.Analysis of the simulation results demonstrate HEC-HMS model is su
28、itable forthe upper reaches of Li River Basin and has rendered good simulation results,and is able to provide reference forflood forecasting in the upper reaches of Lijiang River Basin.Key words:Flood forecasting;upper reaches of Lijiang River Basin;HEC-HMS modelWater quality variation characteristi
29、cs and water treatment effectanalysis of typical water tanks in rocky mountainous area of GuangxiWU Chang-hong,HUANG Xu-sheng,FENG Shi-wei(Guangxi Hydraulic Research Institute,Guangxi Key Laboratory of Water Engineering Materials and Structures,Nanning 530023,China)Abstract:In order to win the battl
30、e of rural drinking water safety and poverty alleviation in rocky mountainous areaof Guangxi,aiming at the water quality problem of water tank,the counties and districts in rocky mountainous areaspurchased a large number of household water purifiers for the masses drinking water of tanks in 2020.Acc
31、ording tothe 14th Five-year Plan for rural water supply guarantee in Guangxi,the water quality guarantee level should be improved through water source protection,strengthening management,necessary and feasible purification treatment forwater tanks that will still exist in rocky mountainous area for
32、a long time.7 widely used types of water purifiers wereinstalled at two water tanks in Hechi City and Baise City respectively to study the water treatment effect of household water purifier;and the water quality seasonal variation characteristics of the water sources of tanks were analyzed at the same time,so as to provide technical support for solving the water quality problems of water tanks inrocky mountainous area.Key words:Rocky mountainous area;water tank;water treatment;water purifier(上接第 6 页)23