Extreme Value Analysis of Annual Maximum Water Levels in the Pearl River Delta, China

Qiang Zhang,Chong-Yu Xu,Yongqin David Chen,Chun-ling Liu

DOI: https://doi.org/10.1007/s11707-009-0025-5

2009-01-01

Frontiers of Earth Science in China

Abstract:We analyzed the statistical properties of water level extremes in the Pearl River Delta using five probability distribution functions. Estimation of parameters was performed using the L-moment technique. Goodness-of-fit was done based on Kolmogorov-Smirnov’s statistic D (K-S D). The research results indicate that Wakeby distribution is the best statistical model for description of statistical behaviors of water level extremes in the study region. Statistical analysis indicates that water levels corresponding to different return periods and associated variability tend to be larger in the landward side of the Pearl River Delta and vice versa. A ridge characterized by higher water level can be identified expanding along the West River and the Modaomen channel, showing the impacts of the hydrologic process of the West River basin. Trough and higher grades of water level changes can be detected in the region drained by Xi’nanyong channel, Dongping channel, and mainstream of Pearl River. The Pearl River Delta region is characterized by low-lying topography and a highly-advanced socio-economy, and is heavily populated, being prone to flood hazards and flood inundation due to rising sea level and typhoons. Therefore, sound and effective countermeasures should be made for human mitigation to natural hazards such as floods and typhoons.