白山流域春季径流影响因素及作用机理识别
|
摘要
我国北方流域春汛来水水源组成多元,影响径流过程的因素众多,且各因素的作用过程复杂,给春汛来水预报带来了挑战。春汛是白山流域第二个集中来水期,准确的春汛预报是科学合理制定水库调度方案的前提和基础,既可以保证农业供水需求、航运用水需求,又能为后期防汛预留库容,实现水资源高效利用。依托白山水库1958—2016年春季长系列日入库流量过程资料,通过Eckhardt递归数字滤波法进行基流分割并绘制基流比过程,根据基流比过程演化趋势将春季径流的水源组成划分为融雪产流、冻土条件下融雪、降雨产流和冻土条件降雨产流等3个阶段。多年平均状况下,白山流域融雪产流开始日期为3月28日,冻土条件产流开始日期为4月28日,融雪产流结束日期为5月20日,冻土消融日期为6月15日;多年平均状况下,融雪产流历时为28 d,冻土条件下融雪、降雨产流历时为26 d,冻土条件下降雨产流历时为24 d。以白山流域内3个气象站1960—2016年日降雨、温度、辐射和风速数据为影响因素,以白山水库日入库径流量(径流深)为目标变量,在融雪产流期(3月28日—5月20日)内,采用全局灵敏度分析法识别出日总辐射和平均风速无时滞效应、最低温度具有1 d的时滞效应、平均温度和最高温度具有2 d的滞后效应,继而识别出基于时滞的日总辐射、平均风速、平均温度和降水为融雪产流的关键影响因素;采用遗传算法拟合融雪产流经验公式,1960—2010年校准期内模拟精度较好,在径流系数小于等于1时和大于1时,拟合优度分别为97.1%和77.5%,平均相对误差为7.5%和22.5%,效率系数分别为96.8%和71.2%;在2011—2016验证期内,在径流系数小于等于1时和大于1时,拟合优度分别为99.3%和99.8%,平均相对误差分别为7.4%和16.8%,效率系数分别为97.8%和94.6%。
There are many formitions of water source in the spring flood of the northern basin of China and factors affecting the runoff process, and the process of each factor is complex, which brings challenges to the spring flood forecasting.Spring flood is the second centralized inflow period in Baishan basin. Spring flood forecasting can scientifically and reasonably formulate reservoir dispatching plan, which can not only guarantee the demand of agricultural water supply and navigation water use, but also reserve reservoir capacity for flood control in the later period, so as to realize efficient utilization of water resources.Based on the long series daily-scale inflow process data of Baishan Reservoir in spring from 1958 to 2016, the base flow is segmented by Eckhardt recursive digital filtering method and basic flow ratio is plotted. According to the evolution trend of the basic flow rate process, the water source composition of spring runoff is divided into three stages, snowmelt runoff, snowmelt-rainfall runoff under frozen soil conditions and rainfall runoff under frozen soil condition.Under the multi-year average condition, the commencing date of snowmelt runoff in Baishan Basin is March 28, the commencing date of runoff under frozen soil conditions is April 28; the end date of snowmelt runoff is May 20, and the date of frozen soil ablation is June 15~(th).Under the multi-year average condition, the snowmelt runoff duration lasted for 28 days, the snowmelt-rainfall runoff under frozen soil conditions duration lasted for 26 days, the rainfall runoff under frozen soil conditions duration lasted for 24 days.Taking the daily rainfall, temperature, radiation and wind speed data of three meteorological stations in Baishan watershed from 1960 to 2016 as the influencing factors, and taking the daily inflow of Baishan Reservoir(runoff depth) as the objective variable, the global sensitivity analysis method was used to identify the total daily radiation and average wind speed without time lag effect during the snowmelt runoff period(March 28-May 20), and the minimum temperature with time lag effect of 1 day and level effect. Mean temperature and maximum temperature have a two-day lag effect, and then the key influencing factors of snowmelt runoff are identified based on the time lag, such as daily total radiation, average wind speed, average temperature and precipitation.The empirical formula of snowmelt runoff is fitted by genetic algorithm, and the simulation accuracy is better during the calibration period from 1960 to 2010. When the runoff coefficient is less than or equal to 1 and larger than 1, the goodness of fit is 97.1% and 77%. The average relative error is 7.5% and 22.5%, and the efficiency coefficients are 96.8% and 71.2% respectively. During the verification period of 2011-2016, when the runoff coefficient is less than or equal to 1 and greater than 1, the goodness of fit is 99.3% and 99.8%, the average relative error is 7.4% and 16.8%, and the efficiency coefficient is 97.8% and 94.6%, respectively.
There are many formitions of water source in the spring flood of the northern basin of China and factors affecting the runoff process, and the process of each factor is complex, which brings challenges to the spring flood forecasting.Spring flood is the second centralized inflow period in Baishan basin. Spring flood forecasting can scientifically and reasonably formulate reservoir dispatching plan, which can not only guarantee the demand of agricultural water supply and navigation water use, but also reserve reservoir capacity for flood control in the later period, so as to realize efficient utilization of water resources.Based on the long series daily-scale inflow process data of Baishan Reservoir in spring from 1958 to 2016, the base flow is segmented by Eckhardt recursive digital filtering method and basic flow ratio is plotted. According to the evolution trend of the basic flow rate process, the water source composition of spring runoff is divided into three stages, snowmelt runoff, snowmelt-rainfall runoff under frozen soil conditions and rainfall runoff under frozen soil condition.Under the multi-year average condition, the commencing date of snowmelt runoff in Baishan Basin is March 28, the commencing date of runoff under frozen soil conditions is April 28; the end date of snowmelt runoff is May 20, and the date of frozen soil ablation is June 15~(th).Under the multi-year average condition, the snowmelt runoff duration lasted for 28 days, the snowmelt-rainfall runoff under frozen soil conditions duration lasted for 26 days, the rainfall runoff under frozen soil conditions duration lasted for 24 days.Taking the daily rainfall, temperature, radiation and wind speed data of three meteorological stations in Baishan watershed from 1960 to 2016 as the influencing factors, and taking the daily inflow of Baishan Reservoir(runoff depth) as the objective variable, the global sensitivity analysis method was used to identify the total daily radiation and average wind speed without time lag effect during the snowmelt runoff period(March 28-May 20), and the minimum temperature with time lag effect of 1 day and level effect. Mean temperature and maximum temperature have a two-day lag effect, and then the key influencing factors of snowmelt runoff are identified based on the time lag, such as daily total radiation, average wind speed, average temperature and precipitation.The empirical formula of snowmelt runoff is fitted by genetic algorithm, and the simulation accuracy is better during the calibration period from 1960 to 2010. When the runoff coefficient is less than or equal to 1 and larger than 1, the goodness of fit is 97.1% and 77%. The average relative error is 7.5% and 22.5%, and the efficiency coefficients are 96.8% and 71.2% respectively. During the verification period of 2011-2016, when the runoff coefficient is less than or equal to 1 and greater than 1, the goodness of fit is 99.3% and 99.8%, the average relative error is 7.4% and 16.8%, and the efficiency coefficient is 97.8% and 94.6%, respectively.
引文
[1] 裴欢,房世峰,刘志辉,等.分布式融雪径流模型的设计及应用[J].资源科学,2008,30(3):454- 459.
[2] 朱景亮,齐非非,穆兴民,等.松花江流域融雪径流及其影响因素[J].水土保持通报,2015,35(2):125- 130.
[3] HORTON R E.The melting of snow [J].Monthly Weather Review,1915,57(3):599- 605.
[4] ANDERSON E A.A point energy and mass balance model of snow cover[R].Washington,DC:US Department of Commerce,1976.
[5] 怀保娟,李忠勤,孙美平,等.SRM融雪径流模型在乌鲁木齐河源区的应用研究[J].干旱区地理,2013,36(1):41- 48.
[6] 孟现勇,吉晓楠,刘志辉,等.SWAT模型融雪模块的改进与应用研究[J].自然资源学报,2014,29(3):528- 539.
[7] 季山,张德伟,宋文生.黑龙江省积雪和融雪径流的初步研究[J].黑龙江水专学报,1994(3):7- 17.
[8] KUMA R M.Eimating evapotranspiration using artificial neural network[J].Journal of Irrigation and Drainage Engineering,2002,128(4):224- 232.
[9] AKGUL M,YURTSEVEN H,AKBURAK S,et al.Short term monitoring of forest road pavement degradation using terrestrial laser scanning[J].Measurement,2017,103:283- 293.
[10] 倪雅茜,张文华,郭生练.流量过程线分割方法的分析探讨[J].水文,2005(3):10- 14.
[11] 董晓华,邓霞,薄会娟,等.平滑最小值法与数字滤波法在流域径流分割中的应用比较[J].三峡大学学报(自然科学版),2010,32(2):1- 4.
[12] ECKHARD T K.A comparison of baseflow indices,which were calculated with seven different baseflow separation methods[J].Journal of Hydrology,2008,352(1/2):168- 173.
[13] 蔡煜东,姚林声.径流长期预报的人工神经网络方法[J].水科学进展,1995,6(1):61- 65.
[14] LI Hongyan,TIAN Qi,WANG Xiaojun,et al.Multivariate coupling sensitivity analysis method based on a back-propagation network and its application[J].Journal of Hydrologic Engineering,2015,20(8):06014013.
[15] IPCC 1995.Climate change:Working Group II:Impacts,adaptations and mitigation of climate change:scientific-technical analyses [M].Cambridge,UK:Cambridge University Press,1996.
[16] DAVIS L.Handbook of genetic algorithms[M].New York:Van Nostrand Reinhold,1991.
[17] MICHALEWIC Z Z.Genetic algorithms+data structures=evolution program[M].Berlin:Springer-Verlag,1996.
[18] 苑希民,李鸿雁,刘树坤,等.神经网络和遗传算法在水科学领域的应用[M].北京:中国水利水电出版社,2002.
[19] 陈国良,王煦法,庄镇泉,等.遗传算法及其应用[M].北京:人民邮电出版社,1996.
[20] LI Hongyan,LIU Xiaowei,LI Shiming.Application of the optimum empirical formula by GAS to the calculation of thesediment concentration process in the Lower Yellow River[J].Journal of Sediment Research,2009,30(3):30- 36.
[21] IPCC.Working Group II:Climate Change Impacts on Adaptation and Vulnerability[R].2001.
[22] 胡铁松,袁鹏,丁晶.人工神经网络在水文水资源中的应用[J].水科学进展,1995,6(1):76- 82.
[23] HSU K,GUPTA H V,SORROSHIAN S.Artificial neural network modeling of the rainfall-runoff process[J].Water Resource Research,1995,31 (10):2517- 2530.
[24] HUANG W,XU B,CHAN-HILTON A.Forecasting flows in Apalachicola River using neural networks[J].Hydrological Processes,2004,18(13):2545- 2564.
[25] 丁晶,邓育仁,安雪松.人工神经前馈(BP)网络模型用作过渡期径流预测的探索[J].水电站设计,1997,13(2):69- 74.
[2] 朱景亮,齐非非,穆兴民,等.松花江流域融雪径流及其影响因素[J].水土保持通报,2015,35(2):125- 130.
[3] HORTON R E.The melting of snow [J].Monthly Weather Review,1915,57(3):599- 605.
[4] ANDERSON E A.A point energy and mass balance model of snow cover[R].Washington,DC:US Department of Commerce,1976.
[5] 怀保娟,李忠勤,孙美平,等.SRM融雪径流模型在乌鲁木齐河源区的应用研究[J].干旱区地理,2013,36(1):41- 48.
[6] 孟现勇,吉晓楠,刘志辉,等.SWAT模型融雪模块的改进与应用研究[J].自然资源学报,2014,29(3):528- 539.
[7] 季山,张德伟,宋文生.黑龙江省积雪和融雪径流的初步研究[J].黑龙江水专学报,1994(3):7- 17.
[8] KUMA R M.Eimating evapotranspiration using artificial neural network[J].Journal of Irrigation and Drainage Engineering,2002,128(4):224- 232.
[9] AKGUL M,YURTSEVEN H,AKBURAK S,et al.Short term monitoring of forest road pavement degradation using terrestrial laser scanning[J].Measurement,2017,103:283- 293.
[10] 倪雅茜,张文华,郭生练.流量过程线分割方法的分析探讨[J].水文,2005(3):10- 14.
[11] 董晓华,邓霞,薄会娟,等.平滑最小值法与数字滤波法在流域径流分割中的应用比较[J].三峡大学学报(自然科学版),2010,32(2):1- 4.
[12] ECKHARD T K.A comparison of baseflow indices,which were calculated with seven different baseflow separation methods[J].Journal of Hydrology,2008,352(1/2):168- 173.
[13] 蔡煜东,姚林声.径流长期预报的人工神经网络方法[J].水科学进展,1995,6(1):61- 65.
[14] LI Hongyan,TIAN Qi,WANG Xiaojun,et al.Multivariate coupling sensitivity analysis method based on a back-propagation network and its application[J].Journal of Hydrologic Engineering,2015,20(8):06014013.
[15] IPCC 1995.Climate change:Working Group II:Impacts,adaptations and mitigation of climate change:scientific-technical analyses [M].Cambridge,UK:Cambridge University Press,1996.
[16] DAVIS L.Handbook of genetic algorithms[M].New York:Van Nostrand Reinhold,1991.
[17] MICHALEWIC Z Z.Genetic algorithms+data structures=evolution program[M].Berlin:Springer-Verlag,1996.
[18] 苑希民,李鸿雁,刘树坤,等.神经网络和遗传算法在水科学领域的应用[M].北京:中国水利水电出版社,2002.
[19] 陈国良,王煦法,庄镇泉,等.遗传算法及其应用[M].北京:人民邮电出版社,1996.
[20] LI Hongyan,LIU Xiaowei,LI Shiming.Application of the optimum empirical formula by GAS to the calculation of thesediment concentration process in the Lower Yellow River[J].Journal of Sediment Research,2009,30(3):30- 36.
[21] IPCC.Working Group II:Climate Change Impacts on Adaptation and Vulnerability[R].2001.
[22] 胡铁松,袁鹏,丁晶.人工神经网络在水文水资源中的应用[J].水科学进展,1995,6(1):76- 82.
[23] HSU K,GUPTA H V,SORROSHIAN S.Artificial neural network modeling of the rainfall-runoff process[J].Water Resource Research,1995,31 (10):2517- 2530.
[24] HUANG W,XU B,CHAN-HILTON A.Forecasting flows in Apalachicola River using neural networks[J].Hydrological Processes,2004,18(13):2545- 2564.
[25] 丁晶,邓育仁,安雪松.人工神经前馈(BP)网络模型用作过渡期径流预测的探索[J].水电站设计,1997,13(2):69- 74.
© 2004-2018 中国地质图书馆版权所有 京ICP备05064691号 京公网安备11010802017129号 地址：北京市海淀区学院路29号 邮编：100083 电话：办公室：(+86 10)66554848；文献借阅、咨询服务、科技查新：66554700 |