Closure to 'Quantifying the Impacts of Climate Change and Human Activities on Runoff Variation: Case Study of the Upstream of Minjiang River, China' by Shuqi Liang,Wensheng Wang, Dan Zhang, Yueqing Li,and Guoqing Wang

摘要

The writers are very grateful to the comments by the discusser on our paper. We are also very grateful to the editor giving us an opportunity to take part in the closure. We have a debate on the mentioned concerns about trend analysis in hydrologic and clima-tologic time series, and the related explanations to the queries are as follows. As the discusser pointed out, the Mann-Kendall test requires the data to be independent since positive and negative autocorrelations induce overestimated and underestimated significance of trends, respectively (Dinpashoh et al. 2014; Lacombe et al. 2012; Sa'adi et al. 2019; Yue et al. 2002). The lag-1 autocorrelation coefficients (r_1) of annual hydrometeorological series and corresponding critical values of significance level 0.05 in the upstream of Minjiang River (UMR) are listed in Table 1. Only the r_1 of mean temperature from 1960 to 2012 and 1970 to 2012, and the r_1 of runoff from 1960 to 2012 exceeded the critical value of significance level 0.05. Three trend test methods, Mann-Kendall (MK) test, prewhit-ening Mann-Kendall (PW-MK) test, and trend-free prewhitening Mann-Kendall (TFPW-MK) test, were employed to analyze the trends of annual hydrometeorological time series in the UMR. The results of Z statistics are presented in Table 2. Obtained results showed that (1) the significant trends of mean temperature from 1960 to 2012 and 1970 to 2012 in the UMR were accepted by the MK test and TFPW-MK test, but were rejected by the PW-MK test; (2) the significant trend of runoff from 1960 to 2012 was accepted by the MK test and TFPW-MK test, but was rejected by the PW-MK test; and (3) except for the runoff of 1960-1969 and 1970-2012, the results of significant trends by the MK test were consistent with the results of the significant trend by the TFPW-MK test. With the preceding analysis, we could draw the conclusion that the different methods had different trend results and the effect of data autocorrelation on trend analysis was not significant in the UMR.