Characteristics, Statistical Analysis and Forecast of Water Resource Changes in the Amu Darya Basin

This paper presents a statistical and econometric analysis of water use, collector discharge and drainage-water mineralization in the Uzbek part of the Amu Darya Basin for two observation periods, 1980–1995 and 2010–2025 (no data are available for 1996–2009). Methods include descriptive statistics, the Mann–Kendall test and Sen’s slope estimator (applied separately to each period), the Mann–Whitney U and Pettitt tests, multivariate OLS and GLS regressions, Pearson and Spearman correlations, hierarchical cluster analysis, ARIMA models fitted on a calendar-year grid with the data gap treated as missing values, sensitivity analysis and Monte Carlo simulation (N = 10,000). Basin-average mineralization rose from 3.13 ± 0.20 g/L in 1980–1995 to 3.51 ± 0.24 g/L in 2010–2025 and increased significantly within each period (Sen’s slope ≈ +0.04 g/L/yr; p ≤ 0.001), although the 2010 level was below the 1995 level. In the OLS model, the time trend was significant (β_T = +0.035 g/L/yr, p < 0.001), whereas collector discharge was not (β_Q = −0.010, p = 0.058; R2 = 0.736). Kashkadarya showed the highest mean mineralization (5.5 g/L) and the steepest increase (+0.075 g/L/yr), and Surkhandarya showed an abrupt rise in 2024–2025. Trend scenarios for 2050 range from 3.9 g/L (level persistence) to 4.4 g/L (continued drift) and are highly uncertain. Uzbekistan’s total water use from the basin (33.9 km3/yr, of which about 20.1 km3/yr was for irrigated agriculture) was stable and corresponded to about 43% of the basin’s renewable water resources, indicating high water stress. Stable water-use volumes therefore do not guarantee stable water quality.