*4.2. Changes in Discharge*

For this study, data from four gauging stations with at least 20 years of data in the KRB have been analyzed. The highland Doab station (see Figure 8, Table 3) shows a strong and significant increase in the mean and minimum annual streamflow, with over 100%, whereas the maximum flow is still strong, but due to the limited data not significant.

**Figure 8.** Mean, maximum, and minimum annual discharge for Doab gauging station. Significant trends (α = 0.05) are depicted as solid red line.

**Table 3.** Trends in mean, maximum, and minimum annual discharge for the gauging stations Doab, Pul-i-Khumri, Chahar Dara, and Kulokh Tepa. Statistically significant trends are bold (α = 0.05).


The Puli-Khumri station, (Figure 9, Table 3) further downstream in the lowland of the KRB, shows inhomogeneous trends with an again strong and significant increase in minimum flow with over 50% decrease, whereas the maximum annual discharge is significantly decreasing by over 20% and the mean annual flow is consequently levelled out without a significant trend.

**Figure 9.** Mean, maximum, and minimum annual discharge for Doab gauging station. Significant trends (α = 0.05) are depicted as solid red line.

The Chahar Dara gauging station (Figure 10, Table 3) further downstream shows strong decreasing trends throughout the year; however, only for the maximum flow this decrease is significant with over 40% reduction.

**Figure 10.** Mean, maximum, and minimum annual discharge for Chahar Dara gauging station. Significant trends (α = 0.05) are depicted as solid red line.

The Kulokh Tepa station (Figure 11, Table 3) at the confluence of Kunduz the Amu Darya River shows similar decreasing patterns with a significant decrease in the maximum flow by almost 60% and a slightly less significant decrease (α = 0.1) for the mean annual discharge by around 25%.

**Figure 11.** Mean, maximum, and minimum annual discharge for the Kulokh-Tepa gauging station. The significant trend with α = 0.1 is depicted as dashed, the significant trend with α = 0.05 is depicted as solid red line.

*4.3. Change in Landcover*

LULC trends in the KRB are assessed by comparing changes between the years 1992 and 2019 (Figure 12). Figure 13 shows the areal changes of the ten defined LULC types. Since 1992, irrigated agriculture, forest/trees, shrubland, urban coverage, as well as barren land and water surfaces, have increased substantially. At the same time, rainfed agriculture, grasslands, and snow/glacier coverage drastically decreased. Table 4 shows landcover classification area in Km<sup>2</sup> and the change in landcover percentage between 1992 and 2019.


**Table 4.** Comparison of 1992 and 2016 Landcover areas [23,28].

Forest/tree area also includes fruit trees and the doubling of this coverage can be explained by a massive expansion of fruit tree plantations, such as almond and pistachio trees. Grassland was mainly degraded to barren land or shrub land. There is also a shift from rainfed to irrigated agriculture, even though the decrease in rainfed agriculture cannot fully be explained by this shift. Large areas of rainfed agriculture seemed to shift into shrublands and barren land.

**Figure 12.** Land cover maps of 1992 and 2019 of the Kunduz River Basin derived from Landsat data.

**Figure 13.** Changes in land use and land cover between 1992 and 2019 in the Kunduz River Basin.
