Coincidence Risk Analysis of Floods Using Multivariate Copulas: Case Study of Jinsha River and Min River, China

Journal Article · Fri Feb 01 00:00:00 EST 2019 · Journal of Hydrologic Engineering

DOI:https://doi.org/10.1061/(ASCE)HE.1943-5584.0001744· OSTI ID:1497016

Peng, Yang ^[1]; Shi, Yulong ^[2]; Yan, Hongxiang ^[3]; Chen, Kai ^[2]; Zhang, Jipeng ^[2]

---

1. Professor, School of Renewable Energy, North China Electric Power Univ., Beijing 102206, China (corresponding author).
2. Postgraduate Student, School of Renewable Energy, North China Electric Power Univ., Beijing 102206, China.
3. Hydrology Technical Group, Pacific Northwest National Laboratory, Richland, WA 99354.

The coincidence of floods on the mainstream and its tributary may cause significant flood damages at downstream of the confluence. In this study, the coincidence risks of flood magnitudes and occurrence dates on Jinsha River and Min River, China were analyzed using multivariate copulas. The Pearson Type III (P-III) distribution and the Mixed von Mises distribution were selected to establish the marginal distributions of flood magnitude and occurrence date, respectively. The Clayton copula and Gumbel-Hougaard (GH) copula functions were then used to establish the joint probability distributions of flood magnitudes and occurrence dates of Jinsha River and Min River, respectively. The coincidence probabilities of flood magnitudes and occurrence dates on Jinsha River and Min River were investigated, and the interval of Annual Maximum (AM) flood occurrence dates of the two rivers was also analyzed. The results show that the higher coincidence probabilities of flood occurrence dates of the two rivers occur in the period from mid-June to early September, with three peaks of coincidence probability occurring on 28^th July, 14^th August, and 31^st August, respectively. The coincidence probability of flood magnitudes for 0.1%, 0.2%, 1%, 5%, and 10% design floods are 1.651×10-8, 6.601×10-8, 1.647×10-6, 4.071×10-5, and 1.606×10-4, respectively. The probability of AM flood occurrence date of Jinsha River and Min River with a time interval of ±3 days is 10.16%. These results can provide guidance for the reservoir operation of Xiluodu-Xiangjiaba cascade reservoirs and offer reference for flood control and disaster reduction at downstream of the confluence.

• Research Organization: Pacific Northwest National Lab. (PNNL), Richland, WA (United States)
• Sponsoring Organization: USDOE
• DOE Contract Number: AC05-76RL01830
• OSTI ID: 1497016
• Report Number(s): PNNL-SA-127655
• Journal Information: Journal of Hydrologic Engineering, Vol. 24, Issue 2; ISSN 1084-0699
• Publisher: American Society of Civil Engineers
• Country of Publication: United States
• Language: English