| 提孜那甫河流域融雪径流模拟及不确定性分析 |
| |
| 引用本文： | 张爽,曾献奎,吴吉春.提孜那甫河流域融雪径流模拟及不确定性分析[J].吉林大学学报(地球科学版),2019,49(5):1415-1424. |
| |
| 作者姓名： | 张爽 曾献奎 吴吉春 |
| |
| 作者单位： | 南京大学地球科学与工程学院, 南京 210023 |
| |
| 基金项目： | 国家重点研发计划项目（2016YFC0402802）；国家自然科学基金项目（41672233，41571017）；中央高校基本科研业务费专项资金（020614380040） |
| |
| 摘 要： | 为了开展寒旱山区典型流域融雪径流过程的研究，提高融雪径流模型（SRM）在山区融雪地区的水文过程模拟精度，本文选取新疆提孜那甫河流域作为典型研究区，在SRM径流计算基础上，加入合适的基流数据并进行不确定性分析。考虑4种常见的基流分割方法（数字滤波法、加里宁法、BFI法（滑动最小值法）和HYSEP（hydrograph separation program）法），基于贝叶斯理论，采用马尔科夫链蒙特卡洛（MCMC）模拟进行参数不确定性分析，对使用不同基流数据SRM的融雪径流模拟表现进行综合评价。分析结果表明，基于加里宁基流分割方法的模型（SRMK）能够最佳地模拟研究区融雪径流过程（纳什系数NSE在识别期和验证期分别为0.866和0.721，大于其他对比模型）。MCMC模拟能够较好地识别SRM参数，获得可靠的参数后验概率分布。当实测降水资料缺乏或其代表性较差时，TRMM（tropical rainfall measuring mission）卫星数据能够描述研究区的降水过程特征。
|
| 关 键 词： | 融雪径流模型 基流分割 马尔科夫链蒙特卡洛法 不确定性 |
| 收稿时间： | 2018-08-11 |
Snowmelt Runoff Simulation and Uncertainty Analysis in Tizinafu River Basin |
| |
| Zhang Shuang,Zeng Xiankui,Wu Jichun.Snowmelt Runoff Simulation and Uncertainty Analysis in Tizinafu River Basin[J].Journal of Jilin Unviersity:Earth Science Edition,2019,49(5):1415-1424. |
| |
| Authors: | Zhang Shuang Zeng Xiankui Wu Jichun |
| |
| Institution: | School of Earth Sciences and Engineering, Nanjing University, Nanjing 210023, China |
| |
| Abstract: | In order to improve the accuracy of the simulated daily streamflow by the snowmelt runoff model (SRM) and focus on the snowmelt runoff process in the typical watershed of cold and arid mountainous areas, the authors selected the Tizinafu River basin in Xinjiang as the study area, and conducted uncertainty analysis on the basis of streamflow calculated by SRM through adding base flow-data. Based on the Bayesian method, combined with the four commonly used base flow separation methods (digital filter,Kalinlin,BFI, and HYSEP(hydrograph separation program)methods), the parameter uncertainty analysis was carried out by using Markov chain Monte Carlo (MCMC) simulation, and the model performance was evaluated comprehensively. According to the results of the model evaluation, the model using Kalinlin base-flow data (SRMK) has higher accuracy in both calibration and validation periods (the values of NSE (Nash-Sutcliffe efficiency coefficient)during model calibration and prediction periods are 0.866 and 0.721, respectively). MCMC method can identify model parameters very well, and can accurately obtain the posterior probability distribution of parameters, while TRMM data can describe the characteristics of precipitation in the study area when the related data is lacking or poorly representative. |
| |
| Keywords: | snowmelt runoff model base flow separation Markov chain Monte Carlo uncertainty |
| 本文献已被 CNKI 等数据库收录！ |
| 点击此处可从《吉林大学学报(地球科学版)》浏览原始摘要信息 |
| 点击此处可从《吉林大学学报(地球科学版)》下载免费的PDF全文 |
