3.4 Double Mass Curves
According to CUSUM analysis, there is evidence that mean annualSSD values significantly decline after 1988-1994 depending on the gauging station. Transition years for every gauging station identified by Taylor’s method are summarized in Table 4 . To further quantify the sediment discharge changes before and after the transition years, double mass curves, along with the linear regression lines, were plotted in Figure 5 . There are precise breakpoints between the two regression lines, suggesting that the selected transition years are correct and meaningful. The regression lines’ slopes were 1.5-4.5 times lower after the breakpoints or transition years (i.e., at higher cumulative precipitation values) than before (see equations inFigure 5 ).
To estimate the relative reduction of total sediment discharge for the period after the transition years, linear models describing the cumulative SSD before the transition years were used to further extrapolate the cumulative sediment up until 2018 (dashed line inFigure 5 ). Compared with the extrapolated cumulative sediment discharge (SSDC ), observed cumulative sediment discharge (SSDO ) reduced by 11-43% in various basins (Table 4 ).
### TABLE 4 ###
### FIGURE 5 ###
We further calculated suspended sediment discharge for the period after the transition years using the regression equations established from the double mass curve of precipitation-sediment before the transition years (cf. Figure 5 ). The difference between observed values\(\overset{\overline{}}{\text{SS}D_{O}}\) before the transition year and estimated \(\overset{\overline{}}{\text{SS}D_{C}}\) after the transition year is due to precipitation change. However, the difference between the estimated mean values \(\overset{\overline{}}{\text{SS}D_{C}}\) and observed values \(\overset{\overline{}}{\text{SS}D_{O}}\) in the same period is the result of other factors (e.g., human activities, glacier shrinkage, sediment source depletion, etc.). The results are shown inTable 5 .
The impact of additional factors was dominant in all cases for the sediment discharge reduction. Their average contribution rate is 87.8% which is more robust than the average precipitation rate (12.2 %). Indeed, the impact of precipitation varies from basin to basin and can explain up to 21% of sediment discharge reduction (e.g., Fia-Ta). Contrariwise, the Kam-Ol basin precipitation events almost did not alter the sediment discharge (5 %). Therefore, there is evidence that precipitation played a minor role in the suspended sediment discharge reduction in the upper Terek basin during the last decades.
### TABLE 5 ###