格尔木河地表径流的趋势和突变特征分析

Trend and Abrupt Change Features of Surface Runoff of Golmud River

中文关键词:  格尔木河  地表径流  MTT方法  MC-ApEn方法  突变特征

英文关键词:Golmud river  Surface runoff  MTT method  MC-ApEn method  Abrupt change features

中文摘要:

      位于柴达木盆地南缘中部的格尔木河是盆地内第二大河流,利用收集到的格尔木水文站1991年至2018年共28年的地表径流日监测数据,采用线性倾向估计法、Mann-Kendall(MK)检验法和滑动t检验法(Moving t-test,MTT),以及滑动移除近似熵(Moving Cut data-Approximate Entropy,MC-ApEn)法对径流数据进行了趋势和突变特征分析。结果表明:1)格尔木河流域径流量年际变化相对稳定,呈现缓慢增加趋势,其速率为1.89×10⁸ m³·(10a)-1,但年内平均径流量变化极不均衡,夏秋季径流量相对较大,占年径流总量的53.86%,其中7月径流量最大,占年径流总量的14.6%。2)联合MC-ApEn方法和MTT方法能准确确定出突变点位置,突变探测能力较好。MK方法和MTT方法分别受时间序列长度和时间窗大小的影响,难以准确识别出突变特征,其中MK方法仅能识别出较少的突变点。3)根据MK方法得出的结果,格尔木河年月日尺度的流量序列数据的突变点位于2007年附近,可能与2007年的流量幅值偏小有关。联合MC-ApEn方法和滑动t检验法,在月尺度序列上给出的稳定突变点分布在1996年、2001年、2009年和2015年,结合2010年格尔木发生的大洪水事件,2009年的突变可能指示了该次事件的异常。

英文摘要:

      The Golmud river is the second largest river in the Qaidam basin located in the middle of the south edge of the basin.Based on the surface runoff daily monitoring data collected from Golmud Hydrological Station from 1991 to 2018,the trend and abrupt change features of runoff data are analyzed by linear tendency estimation method,Mann-Kendall(MK)test method,Moving t-test(MTT)method and Moving Cut data-Approximate Entropy(MC-ApEn)method.The results show that:1)The interannual variation of runoff of Golmud River was relatively stable,showing a slow increasing trend,with the rate of 1.89×10⁸ m³·(10a)-1.However,the change of the average runoff during the year was extremely uneven,and the runoff in summer and autumn was relatively large,accounting for 53.86% of the total annual runoff.The runoff in July was the largest,accounting for 14.6% of the total annual runoff.2)The combinated methods of MC-ApEn and MTT can accurately determine the location of the mutation point,and the mutation detection ability is better.MK and MTT methods are respectively affected by the length of time series and the size of time window,so it is difficult to accurately identify mutation features,and MK method can only identify a few mutation points.3)According to the results of the MK method,the abrupt change point of the flow series data on the annual,monthly and daily scales in the Golmud River was around 2007,which may be related to the small amplitude of the flow in 2007.The stable mutation points given by the combination of MC-ApEn and MTT methods on the monthly scale series are distributed in 1996,2001,2009 and 2015.Combined with the great flood in Golmud in 2010,the abrupt change in 2009 may indicate this anomalous event.