4.4 Uncertainties
Among the uncertainty sources discussed in the corresponding section (section 2.6), we suppose that uncertainty in trend results is mainly attributed to data limitations. As suggested by previous studies (e.g., Onyutha, 2016), the uncertainty in the trend reduces as the series record length increases. These difficulties may be to some extent be solved by considering extended suspended sediment time series. However, it can be impossible for gauging stations in the Terek basin as many closed or stopped to measure suspended sediment yield. Likewise, future research may still benefit from considering a larger number of catchments from neighboring regions of the Greater Caucasus (e.g., Georgia, Armenia, Azerbaijan).
Even though we used data from 33 catchments (Fig. 1 ), which is significantly more than many other contemporary studies in that region (Gusarov et al. , 2021), this number may still be too small to identify the impact of all relevant factors. However, especially considering more high-altitude gauging stations (higher than 1000 m) in the analyses could help since that altitude group is less populated with prolonged timeseries.
Furthermore, apart from potential cropland, no other factors could be identified that significantly correlated with the SSL trend in high-mountain gauging stations (Fig. 8c ). In other words, our analyses provide no evidence that SSL trend values for high mountain catchments (with a gauging station altitude higher than 1000 m) depend on landuse/landcover or spatial scale. However, it should be emphasized that this does not necessarily imply that these factors cannot influence the SSL change. The lack of significant correlations between SSL trend values and the considered variables may also be attributed to the inherent limitations of our approach and a small number of gauging stations in this altitude group subset.