本文選題：Cox回歸模型　切入點(diǎn)：POT抽樣　出處：《水科學(xué)進(jìn)展》2016年04期

【摘要】：運(yùn)用Cox回歸模型、月頻率法以及離散指數(shù)法,研究了新疆塔里木河(塔河)流域8個水文站點(diǎn)POT抽樣和5個區(qū)域洪水序列時間集聚性特征以及受低頻氣候變化的影響。結(jié)果表明:受氣候變化的影響,塔河流域洪水序列呈現(xiàn)顯著集聚性特征,洪水發(fā)生頻率高的時期往往也是大量級洪峰集中發(fā)生的時期,這是塔河流域洪災(zāi)損失居高不下的主要原因;Cox回歸模型擬合的氣候指標(biāo)系數(shù)值為正值的站點(diǎn)和區(qū)域,氣候指標(biāo)正相位導(dǎo)致相同超過概率的洪水發(fā)生時間提前,而相同發(fā)生時間的洪水發(fā)生超過概率降低,氣候指標(biāo)值為負(fù)值時則相反;塔河流域大部分水文站點(diǎn)和區(qū)域洪水發(fā)生的超過概率均對氣候指標(biāo)變化有較好響應(yīng),這一現(xiàn)象有利于塔河流域洪水風(fēng)險控制與洪災(zāi)管理;塔河流域站點(diǎn)洪水序列多無年際集聚性現(xiàn)象,而區(qū)域洪水序列的年際集聚性特征顯著。
[Abstract]:The application of Cox regression model on frequency method and the discrete index method of Xinjiang Tarim River Basin (Tahe) 8 hydrological stations POT sampling and 5 regional flood time series clustering characteristics and the influence of low-frequency climate change. The results showed that affected by climate change, Tahe basin flood sequence showed significant agglomeration characteristics flood, high frequency period is often a concentrated peak period, which is the main reason of Tahe River flood loss high; climate index value Cox regression model fitting for positive stand point and regional climate index, positive phase lead time had the same probability of flood in advance, and the same time the flood occurred more than reduce the probability, the climate index value is negative when the opposite; more than the occurrence probability of Tahe basin most hydrological stations and regional flood of gas This change is favorable for flood risk control and flood management in the Tahe River Basin. There is no interannual clustering phenomenon in the flood season of Tahe basin, and the interannual clustering characteristics of regional flood sequences are significant.

【作者單位】： 中山大學(xué)水資源與環(huán)境系;中山大學(xué)華南地區(qū)水循環(huán)與水安全廣東省普通高校重點(diǎn)實(shí)驗(yàn)室;安徽師范大學(xué)國土資源與旅游學(xué)院;
【基金】：國家杰出青年科學(xué)基金資助項(xiàng)目(51425903) 香港特別行政區(qū)研究資助局項(xiàng)目(CUHK441313)~~
【分類號】：TV122
，

本文編號：1720985