理想和实际资料情况下Nash模型参数异参同效性研究
|
摘要
【目的】研究异参同效问题对水文模型参数率定的影响。【方法】以黑龙江农垦红兴隆分局所辖的双鸭山农场和友谊农场为研究区域,在理想资料和实际资料2种情况下,构建了基于SCE-UA算法的模拟优化模型,分别对Nash模型参数进行了优化并分析了参数率定过程中的异参同效现象的表现特征。理想资料排除了水文模型输入资料、模型结构以及模型输出等方面的误差对模型参数率定过程的干扰。【结果】在2种情况下,异参同效现象在Nash模型参数率定过程中的表现特征为对较优似然值对应的参数组取值范围的影响和最优似然值对应的参数组数量的影响。对不同场次洪水而言,2种情况下异参同效的2方面表现特征均与输入洪水的洪量以及洪峰的大小有关,即随着输入洪水的洪量和洪峰的增加,异参同效现象越明显。对同一洪水而言,在理想资料情况下参数率定过程中的异参同效显著程度小于实际资料情况下异参同效显著程度。【结论】在水文模型参数率定过程中,应基于不同输入资料情况下参数异参同效特征分析,选取合理的参数率定方法,尽量避免或消除异参同效对参数率定的不利影响,提高模型预报精度。
【Objective】The purpose of this work is to study the influence of equifinality phenomenon on accuracy of the Nash model parameter calibration.【Method】The Nash model parameters were calculated by the simulation-optimization model based on SCE-UA, using both idealized data and measured data. Taking Shuangyashan Farm and Youyi Farm in Hongxinglong Land Reclamation Administration in Heilongjiang province as an example, the performance of the equifinality phenomenon in parameter calibration was analyzed. The idealized data can eliminate the uncertainty caused by errors in the input data, model inaccuracy and model parameters in the calibration.【Result】Performance of the equifinality phenomenon in the Nash model parameter calibration was represented in value range and number of optimal parameters for both the idealized data and the measured data.For different flooding events, the performance of the equifinality phenomenon in terms of the two aspects was associated with the flooding volume and flooding peak for the two case studies. Specifically, the equifinality phenomenon was more noticeable with the increases in flooding volume and flooding peak. Under the same flooding event, the equifinality phenomenon was more noticeable in the parameter calibration for the idealized data than for the measured data.【Conclusion】In parameter calibration of hydrological model based on the equifinality phenomenon using different data sources, a proper parameter calibration method should be used to eliminate the unfavorable influence of the equifinality so as to improve accuracy of the model.
【Objective】The purpose of this work is to study the influence of equifinality phenomenon on accuracy of the Nash model parameter calibration.【Method】The Nash model parameters were calculated by the simulation-optimization model based on SCE-UA, using both idealized data and measured data. Taking Shuangyashan Farm and Youyi Farm in Hongxinglong Land Reclamation Administration in Heilongjiang province as an example, the performance of the equifinality phenomenon in parameter calibration was analyzed. The idealized data can eliminate the uncertainty caused by errors in the input data, model inaccuracy and model parameters in the calibration.【Result】Performance of the equifinality phenomenon in the Nash model parameter calibration was represented in value range and number of optimal parameters for both the idealized data and the measured data.For different flooding events, the performance of the equifinality phenomenon in terms of the two aspects was associated with the flooding volume and flooding peak for the two case studies. Specifically, the equifinality phenomenon was more noticeable with the increases in flooding volume and flooding peak. Under the same flooding event, the equifinality phenomenon was more noticeable in the parameter calibration for the idealized data than for the measured data.【Conclusion】In parameter calibration of hydrological model based on the equifinality phenomenon using different data sources, a proper parameter calibration method should be used to eliminate the unfavorable influence of the equifinality so as to improve accuracy of the model.
引文
[1]吴江,胡胜. DEM分辨率对SWAT模型水文模拟的影响研究[J].灌溉排水学报, 2016, 35(11):18-23.
[2]金鑫,陈琼,金彦香.土地利用/覆被数据精度对流域水文过程模拟的影响研究[J].灌溉排水学报, 2018, 37(2):108-115.
[3]李玉庆,张存,张文贤.水稻灌区农业面源污染物迁移转化规律模拟研究[J].灌溉排水学报, 2017, 36(11):29-35.
[4]王维,冯忠伦,杨伟,等.基于SCE-UA算法的新安江模型与垂向混合产流模型参数优化应用研究[J].中国农村水利水电, 2017(3):26-30.
[5]周建中,卢韦伟,孙娜,等.水文模型参数多目标率定及最优非劣解优选[J].水文, 2017, 37(2):1-7.
[6]芮孝芳.水文学前沿科学问题之我见[J].水利水电科技进展, 2015, 35(5):95-102.
[7]任政,盛东.基于多目标GLUE算法的新安江模型参数不确定性研究[J].水电能源科学, 2016, 34(3):15-18.
[8]赵红杰,柏继云,马力.遗传扩展蚁群算法用于马斯京根模型参数估计[J].东北农业大学学报, 2014, 45(8):118-123.
[9] MERZ R, PARAJKA J, BLOSCHL G. Time stability of catchment model parameters:implications for climate impact analyses[J]. Water Resources Research, 2011, 47(2).
[10]江善虎,任立良,刘淑雅,等.基于贝叶斯模型平均的水文模型不确定性及集合模拟[J].中国农村水利水电, 2017(1):107-112.
[11]王书功.水文模型参数估计方法及参数估计不确定性研究[D].北京:中国科学院研究生院,2006.
[12]周建中,卢韦伟,孙娜,等.水文模型参数多目标率定及最优非劣解优选[J].水文, 2017, 37(2):1-7.
[13]刘苏宁,甘泓,魏国孝.粒子群算法在新安江模型参数率定中的应用[J].水利学报, 2010, 41(5):537-544.
[14]陈坰烽,张万昌,吴波.多目标遗传单纯形算法在概念性水文模型参数优化中的应用[J].水土保持通报, 2008(3):107-112.
[15]芮孝芳.论流域水文模型[J].水利水电科技进展, 2017, 37(4):1-7.
[16] DUAN Q, AJAMI N K, GAO X, et al. Multi-model ensemble hydrologic prediction using bayesian model averaging[J]. Advances in Water Resources,2007, 30(5):1 371-1 386.
[17]梁忠民,戴荣,李彬权.基于贝叶斯理论的水文不确定性分析研究进展[J].水科学进展, 2010, 21(2):274-281.
[18] BEVEN K, FREER J. Equifinality, data assimilation, and uncertainty estimation in mechanistic modelling of complex environmental systems using the glue methodology[J]. Journal of Hydrology, 2001, 249(1/4):11-29.
[19] DUAN Q Y, SOROOSHIAN S, GUPTA V. Effective and efficient global optimization for conceptual rainfall-runoff models[J]. Water Resources Research, 1992, 28(4):1 015-1 031.
[20]史旭栋,高岳林,韩俊茹.基于模糊推理的粒子群优化算法[J].河南师范大学学报(自然科学版), 2017,45(2):108-118.
[21] COELLO A C C, PULIDO T G, LECHUGA S M. Handling multiple objectives with particle swarm optimization[J]. Ieee Transactions on Evolutionary Computation, 2004, 8(3):256-279.
[22]张新明,王霞,涂强,等.融合榜样学习和反向学习的粒子群优化算法[J].河南师范大学学报(自然科学版), 2017, 45(6):91-99.
[23]康立山,谢云,尤矢勇.非数值并行算法:模拟退火算法[M].北京:科学出版社, 1994.
[24]岳振芳,高岳林.融合模拟退火的改进教与学优化算法[J].河南师范大学学报(自然科学版), 2016, 44(1):149-154.
[25]鲍振鑫,张建云,刘九夫,等.基于土壤属性的VIC模型基流参数估计框架[J].水科学进展, 2013, 24(2):169-176.
[26]刘昭,杨文元,查元源,等.基于田块尺度含水率观测的土壤水力参数多模型反演[J].农业工程学报, 2015, 31(6):135-144.
[27]张质明,王晓燕,潘润泽.一种改进的不确定性水质模型参数率定方法[J].中国环境科学, 2017, 37(3):956-962.
[28] HER Younggu, CHAUBEY Indrajeet. Impact of the numbers of observations and calibration parameters on equifinality, model performance, and output and parameter uncertainty[J]. Hydrological Processes, 2015, 29(19):4 220-4 237.
[29]包为民,李倩,李偲松,等.基于灵敏度的条件目标函数构建方法研究[J].水力发电学报, 2013, 32(2):27-34.
[30]姚蕾,梁忠民,王军,等. Nash汇流模型在无资料地区的应用[J].水电能源科学, 2014,32(2):23-26.
[31]包为民.水文预报[M].北京:中国水利水电出版社, 2011.
[32]朱君君,芮孝芳,刘氚.基于理想资料的Nash模型异参同效性研究[J].水电能源科学, 2014, 32(3):80-84.
[33]朱君君. Nash模型参数异参同效问题及参数确定方法研究[D].南京:河海大学, 2011.
[2]金鑫,陈琼,金彦香.土地利用/覆被数据精度对流域水文过程模拟的影响研究[J].灌溉排水学报, 2018, 37(2):108-115.
[3]李玉庆,张存,张文贤.水稻灌区农业面源污染物迁移转化规律模拟研究[J].灌溉排水学报, 2017, 36(11):29-35.
[4]王维,冯忠伦,杨伟,等.基于SCE-UA算法的新安江模型与垂向混合产流模型参数优化应用研究[J].中国农村水利水电, 2017(3):26-30.
[5]周建中,卢韦伟,孙娜,等.水文模型参数多目标率定及最优非劣解优选[J].水文, 2017, 37(2):1-7.
[6]芮孝芳.水文学前沿科学问题之我见[J].水利水电科技进展, 2015, 35(5):95-102.
[7]任政,盛东.基于多目标GLUE算法的新安江模型参数不确定性研究[J].水电能源科学, 2016, 34(3):15-18.
[8]赵红杰,柏继云,马力.遗传扩展蚁群算法用于马斯京根模型参数估计[J].东北农业大学学报, 2014, 45(8):118-123.
[9] MERZ R, PARAJKA J, BLOSCHL G. Time stability of catchment model parameters:implications for climate impact analyses[J]. Water Resources Research, 2011, 47(2).
[10]江善虎,任立良,刘淑雅,等.基于贝叶斯模型平均的水文模型不确定性及集合模拟[J].中国农村水利水电, 2017(1):107-112.
[11]王书功.水文模型参数估计方法及参数估计不确定性研究[D].北京:中国科学院研究生院,2006.
[12]周建中,卢韦伟,孙娜,等.水文模型参数多目标率定及最优非劣解优选[J].水文, 2017, 37(2):1-7.
[13]刘苏宁,甘泓,魏国孝.粒子群算法在新安江模型参数率定中的应用[J].水利学报, 2010, 41(5):537-544.
[14]陈坰烽,张万昌,吴波.多目标遗传单纯形算法在概念性水文模型参数优化中的应用[J].水土保持通报, 2008(3):107-112.
[15]芮孝芳.论流域水文模型[J].水利水电科技进展, 2017, 37(4):1-7.
[16] DUAN Q, AJAMI N K, GAO X, et al. Multi-model ensemble hydrologic prediction using bayesian model averaging[J]. Advances in Water Resources,2007, 30(5):1 371-1 386.
[17]梁忠民,戴荣,李彬权.基于贝叶斯理论的水文不确定性分析研究进展[J].水科学进展, 2010, 21(2):274-281.
[18] BEVEN K, FREER J. Equifinality, data assimilation, and uncertainty estimation in mechanistic modelling of complex environmental systems using the glue methodology[J]. Journal of Hydrology, 2001, 249(1/4):11-29.
[19] DUAN Q Y, SOROOSHIAN S, GUPTA V. Effective and efficient global optimization for conceptual rainfall-runoff models[J]. Water Resources Research, 1992, 28(4):1 015-1 031.
[20]史旭栋,高岳林,韩俊茹.基于模糊推理的粒子群优化算法[J].河南师范大学学报(自然科学版), 2017,45(2):108-118.
[21] COELLO A C C, PULIDO T G, LECHUGA S M. Handling multiple objectives with particle swarm optimization[J]. Ieee Transactions on Evolutionary Computation, 2004, 8(3):256-279.
[22]张新明,王霞,涂强,等.融合榜样学习和反向学习的粒子群优化算法[J].河南师范大学学报(自然科学版), 2017, 45(6):91-99.
[23]康立山,谢云,尤矢勇.非数值并行算法:模拟退火算法[M].北京:科学出版社, 1994.
[24]岳振芳,高岳林.融合模拟退火的改进教与学优化算法[J].河南师范大学学报(自然科学版), 2016, 44(1):149-154.
[25]鲍振鑫,张建云,刘九夫,等.基于土壤属性的VIC模型基流参数估计框架[J].水科学进展, 2013, 24(2):169-176.
[26]刘昭,杨文元,查元源,等.基于田块尺度含水率观测的土壤水力参数多模型反演[J].农业工程学报, 2015, 31(6):135-144.
[27]张质明,王晓燕,潘润泽.一种改进的不确定性水质模型参数率定方法[J].中国环境科学, 2017, 37(3):956-962.
[28] HER Younggu, CHAUBEY Indrajeet. Impact of the numbers of observations and calibration parameters on equifinality, model performance, and output and parameter uncertainty[J]. Hydrological Processes, 2015, 29(19):4 220-4 237.
[29]包为民,李倩,李偲松,等.基于灵敏度的条件目标函数构建方法研究[J].水力发电学报, 2013, 32(2):27-34.
[30]姚蕾,梁忠民,王军,等. Nash汇流模型在无资料地区的应用[J].水电能源科学, 2014,32(2):23-26.
[31]包为民.水文预报[M].北京:中国水利水电出版社, 2011.
[32]朱君君,芮孝芳,刘氚.基于理想资料的Nash模型异参同效性研究[J].水电能源科学, 2014, 32(3):80-84.
[33]朱君君. Nash模型参数异参同效问题及参数确定方法研究[D].南京:河海大学, 2011.