鄱阳湖流域洪水变化特征及气候影响研究

为探讨鄱阳湖流域洪水过程的变化特征和规律,系统分析流域洪水量级、频率、发生时间的变化特征,利用核密度估计分析洪水发生率的非平稳性,运用月频率法评价洪水集聚性特征,并探讨低频气候因子对洪水变化的影响。研究表明：① 鄱阳湖流域各水文站点年、秋季和冬季最大洪水及POT超阈值洪水以增加趋势为主。② 洪水发生率年内集聚性显著,主要集中在4~7月;年际洪水发生次数呈现非平稳泊松分布,洪水发生率出现明显的年际集聚性特征。③ ENSO、IOD对下年洪水量级及洪水发生次数有明显影响,洪水发生次数与年最大洪水量级异常现象通常是ENSO和IOD协同作用结果。

Floods Characteristics and Impacts from Climate Indices in the Poyang Lake Basin

Changing properties of floods in the Poyang Lake have been thoroughly investigated in terms of flooding magnitude, occurrence rates and timing. These characteristics were obtained from annual and seasonal maximum flow (AMF) and sampling the flood series based on the Peak-over-Threshold (POT) method. Kernel estimation technique and Bootstrap method were used in detection of nonstationary of flooding processes. Monthly frequency have been used in analysis of changing properties of floods in terms of annual and international viewpoints. Moreover, impacts of ENSO, NAO, IOD and PDO on floods in the Poyang Lake Basin have been widely analyzed. Results indicated that: 1) Annual maximum streamflow and POT-based flood events are generally in increasing tendency at annual and seasonal (winter and autumn) time scales. The change points of AMF at hydrological stations of the lake basin were mainly in 1980s-1990s, except Lijiadu station where the change point of AMF was 1966. The flood magnitude increased by 19%, 13%, 22% and 16% after the change points at Lijiadu, Meigang, Hushan and Wanjibu stations. Autumn AMF and Winter AMF in all stations are increased or significantly increased. The average increase rate of Autumn AMF and Winter AMF are 113% and 40.5%, respectively, after the change points. 2) Clustering effects of floods were evident and floods were active during April and July, which were significant at 5% significance level, implying that the time interval from April to July is the active flood period.Massive occurrences of heavy floods during a certain period usually trigger high flooding risk and which poses serious challenges for human mitigation to flood hazards. 3) Annual occurrence rates of floods follow nonstationary poisson distribution with evident interannual clustering properties. Further, two periods having higher occurrence rates of floods were identified, i.e. in the late 1960s and the early 1970s, and the mid 1990s. However, the highest occurrence rates of floods are observed mainly during the 1990s. Active flood activities can usually be observed at most hydrological stations in the Poyang Lake Basin. 4) ENSO and IOD have coincident impacts on flooding magnitude and occurrence rates of floods of the subsequent years. ENSO usually have negative impacts on floods of the same year during Spring seasons and have a significant impact on the flood occurrence rate and on annual maximum streamflow during Spring in the subsequent years. IOD has a significant positive relation with the occurrence timing of floods in the same year. Positive IOD phase usually delayed occurrence timing of floods. PDO and NDO have no significant impacts on flood processes of the lake basin. the warm phase of ENSO together with the warm phase of IOD combine to intensify precipitation extremes of the next year in the Poyang Lake Basin and hence increase peak flood flow and flood occurrence rate across the basin. The results of this study are of great scientific and practice merits in terms of human mitigation to floods and also management of flood hazards, conservation of ecological environment of the Poyang Lake Basin.