1School of Geographic and Oceanographic Science, Nanjing University, Nanjing, P.R. China
2Key Laboratory of Surficial Geochemistry, Ministry of Education, Department of Hydrosciences, School of Earth Sciences and Engineering, State Key Laboratory of Pollution Control and Resource Reuse, Nanjing University, Nanjing, P.R. China
3Department of Biological and Agricultural Engineering, Zachry Department of Civil Engineering, Texas A&M University, College Station, TX77843, USA; and National Water Center, UAE University, Al Ain, UAE
4Nanjing Hydraulic Research Institute, Nanjing, P.R. China
1School of Geographic and Oceanographic Science, Nanjing University, Nanjing, P.R. China
2Key Laboratory of Surficial Geochemistry, Ministry of Education, Department of Hydrosciences, School of Earth Sciences and Engineering, State Key Laboratory of Pollution Control and Resource Reuse, Nanjing University, Nanjing, P.R. China
3Department of Biological and Agricultural Engineering, Zachry Department of Civil Engineering, Texas A&M University, College Station, TX77843, USA; and National Water Center, UAE University, Al Ain, UAE
4Nanjing Hydraulic Research Institute, Nanjing, P.R. China
Received: 10 Jul 2019 – Accepted for review: 19 Sep 2019 – Discussion started: 14 Oct 2019
Abstract. Due to global climate change and urbanization, more attention has been paid to decipher the nonstationary multivariate risk analysis from the perspective of probability distribution establishment. Because of the climate change, the exceedance probability belonging to a certain extreme rainfall event would not be time invariant any more, which impedes the widely-used return period method for the usual hydrological and hydraulic engineering practice, hence calling for a time dependent method. In this study, a multivariate nonstationary risk analysis of annual extreme rainfall events, extracted from daily precipitation data observed at six meteorological stations in Haihe River basin, China, was done in three phases: (1) Several statistical tests, such as Ljung-Box test, and univariate and multivariate Mann-Kendall and Pettist tests were applied to both the marginal distributions and the dependence structures to decipher different forms of nonstationarity; (2) Time-dependent Archimedean and elliptical copulas combined with the Generalized Extreme Value (GEV) distribution were adopted to model the distribution structure from marginal and dependence angles; (3) A design life level-based (DLL-based) risk analysis associated with Kendall's joint return period (JRPken)and AND's joint return period (JRPand) methods was done to compare stationary and nonstationary models. Results showed DLL-based risk analysis through the JRPken method exhibited more sensitivity to the nonstationarity of marginal and bivariate distribution models than that through the JRPand method.
In this study, a multivariate nonstationary risk analysis of annual extreme rainfall events, extracted from daily precipitation data observed at six meteorological stations in Haihe River basin, China, was done in three phases: (1) Several statistical tests, were applied to both the marginal distributions and the dependence structures to decipher different forms of nonstationarity; (2) Time-dependent copulas were adopted to model the distribution structure.
In this study, a multivariate nonstationary risk analysis of annual extreme rainfall events,...
- Printer-friendly version
- Supplement
