作 者：Liu, SaiyanHuang, ShengzhiXie, YangyangWang, HaoLeng, GuoyongHuang, QiangWei, XiaotingWang, Lu
作者机构：State Key Laboratory of Eco-hydraulics in Northwest Arid Region of ChinaXi’an University of Technology Xi’an710048 ChinaSchool of Hydrologic energy and power engineeringResearch Institute of Modern Rural Water Conservancy Yangzhou225009 ChinaState Key Laboratory of Simulation and Regulation of Water Cycle in River BasinChina Institute of Water Resources and Hydropower Research Beijing100038 ChinaEnvironmental Change InstituteUniversity of Oxford OxfordOX1 3QY United Kingdom
出 版 物：《Water Resources Management》
年 卷 期：2019年第33卷第3期
页 面：939-953
核心收录：
中图分类：TV[工业技术-水利工程]
学科分类：08[工学]0815[工学-水利工程]
基 金：Acknowledgements This research was supported by the National Natural Science Foundation of China (51709221)the Planning Project of Science and Technology of Water Resources of Shaanxi (2015slkj-272017slkj-19)Key laboratory research projects of education department of Shaanxi province (17JS104)and China Scholarship Council (201708610118).
主 题：FloodsAtmospheric pressureClimate changeFlood controlWater resourcesChangepointsFlood mitigationInformation criterionMean change pointsNonstationaritiesSouthern oscillationVegetation coverageWater resources management
摘 要：The assumption of stationarity in the flood time series is the basis for flood design and forecasting. Therefore, identification of the non-stationarity of flood series and the underlying causes is necessary for flood risk and water resources management. The Wei River Basin (WRB) of China was selected as the case study. Nonstationary flood behavior was examined comprehensively in terms of trends and the mean and variance change point. Then, the implications of the nonstationary flood series were explored. Furthermore, the impacts of antecedent precipitation, El Niño Southern Oscillation/Pacific Decadal Oscillation and vegetation coverage on floods were investigated. The results indicated following: (1) There is a non-significant delay in the timing of the annual maximum flood peak and seasonal floods across the basin; (2) the assumption of stationarity in the flood series is invalid, with significant downward trends and change points identified; (3) bias arising from the variance change point is much more significant than that of the mean change point in estimating floods; and (4) changing climate and human activities are jointly responsible for nonstationary floods in the WRB. These findings provide new insights into nonstationary flood behavior by emphasizing the importance of identifying the potential variance change points in the flood series, which is important for flood mitigation and water resources management. © 2018, Springer Nature B.V.
