低频气候变化引起的珠江流域年均和洪峰流量变化特征及灵敏度分析

Low frequency climate variability induced annual mean and annual maxim streamflow variations and related sensitivity across the Pearl River basin

低频气候变化是引起内陆径流年际和年代际变化的一个重要驱动因子。通过分析El Nio/Southern Oscillation(ENSO)、North Atlantic Oscillation(NAO)、Indian Ocean Dipole(IOD)和Pacific Decadal Oscillation(PDO)等主要低频气候因子对珠江流域年均(Q_(ann))流量和洪峰流量(Q_(max))的影响及其影响量级,研究结果表明珠江流域流量受到低频气候因子的显著影响,但影响强度的时间平稳性与趋势性有显著区域差异。对相应区域具有持续显著影响及相关强度呈显著上升趋势的气候因子可以作为Q_(ann)和Q_(max)的预测信号。低频气候因子位于不同的相位,导致珠江流域流量发生相应的变化:负相位ENSO、NAO和PDO易致较低Q_(ann),导致水文干旱风险的增加;而正相位的ENSO、IOD及负相位NAO和PDO易引发较高Q_(max),导致极端洪灾风险增加。对比Q_(ann)和Q_(max),Q_(max)对于气候指标变化的灵敏度要高于Q_(ann),Q_(max)灵敏度高于Q_(ann)的面积比例分别为56%、59%、71%和36%。研究对于根据低频气候变化信号预测珠江流域Q_(ann)与Q_(max)及珠江流域洪旱灾害的预报与预警具有重要理论意义与实际应用价值。

Variability of the low-frequency climate change is the principle driver for the annual and inter-annual streamflow changes.The influences of low-frequency climate indices such as El Nino/Southern Oscillation （ENSO）,North Atlantic Oscillation （NAO）,Indian Ocean Dipole （IOD）and Pacific Dec-adal Oscillation （PDO）,on annual mean discharge （Qann ）and annual peak discharge （Qmax ）of the Pearl River basin have been quantitatively analyzed.The results show that：①Qann and Qmax were signifi-cantly affected by different climate indices in different regions with distinctly different spatial patterns in terms of correlation degrees and sensitivity and also the trends of impact strength.ENSO and IOD at the same year exerted a persistent significant impact on the Qann at the regions covering most of the West River basin,and the strength of the correlation increased significantly,which can be taken as the predictor for Qann .For the same reason,PDO at the same year can be taken as the predictor for Qann in the eastern parts of the Pearl River basin and the North River basin,and NAO,IOD a year earlier and NAO at the same year can be taken as the predictor for Qmax in the middle Pearl River basin;② As for the entire Pearl River basin,negative-phase ENSO,NAO and PDO tend to cause Qann at lower levels with an in-creasing risk of droughts.However,positive-phase ENSO,IOD and negative phase NAO and PDO tend to cause Qmax at higher levels with an increasing risk of extreme floods;③Qmax was more sensitive to vari-ability of atmospheric circulation than Qann .Qann varied between 0.3% and 24%,while Qmax varied be-tween 0.5%and 31%per unit index change.The NAO and IOD both at the same year and a year earli-er show a higher sensitivity in most area of the Pearl River basin,but the sensitivity difference between Qann and Qmax was significant,the latter had an area ratio of 56%,59%,71%and 36%sensitivity high-er than the former.