[Objectives] Large reservoirs in arid regions profoundly alter downstream flow regimes, yet systematic assessments across different operational stages, hydrological year types, and spatial scales remain limited. This study aims to quantify the regulation effects of the Dashixia Reservoir on the hydrological regime of the Aksu River Basin and to reveal the underlying mechanisms of intra-annual redistribution, time‑scale dependence, and spatial attenuation. [Methods] A coupled SWAT-optimization modeling framework was established for the Aksu River Basin. The SWAT model, calibrated and validated using daily streamflow data (2012-2025) at the basin outlet, achieved satisfactory performance (calibration: R² = 0.78, NSE = 0.77; validation: R² = 0.72, NSE = 0.71). A multi-objective optimization model, incorporating flood control, hydropower generation, and water supply objectives, was employed to simulate reservoir outflow hydrographs in the absence of historical operation records. Two scenarios (without and. with reservoir), three hydrological year types (25% wet, 50% normal, 75% dry), four downstream sections, and two operational stages (impoundment and stable operation) were analyzed. The Relative Change Rate (RCR) was adopted to quantify alterations across 33 Indicators of Hydrologic Alteration (IHA). [Results] Reservoir operation fundamentally restructured the intra-annual flow distribution: spring-summer (January-September) mean flows decreased by 62.6%, while November-December flows increased by up to 74.6%; the January-April mean flows converged from a natural range of 46.33-60.28 m³/s to approximately 23.23 m³/s after impoundment. A distinct time-scale dependence was identified: the 1-day minimum flow increased by 642% in dry years (from 4.84 to 35.92 m³/s), yet the 90-day minimum decreased by 20.2% (from 36.46 to 29.09 m³/s); flood peak reduction efficiency dropped from 59% (1-day maximum) to 19.6% (90-day maximum). The annual minimum flow occurrence was delayed by 30-104 days, with dry years experiencing the greatest delays (94-104 days). High and low pulse characteristics shifted from frequent, short pulses to rare, long pulses: in dry years, low pulse frequency decreased by 87.5% (from 16 to 2 events) while mean pulse duration increased by 698% (from 5.7 to 45.5 days). Daily flow rise and fall rates were suppressed by 74%-86%, and reversal frequency declined by 66%-72%. All reservoir-induced effects decayed exponentially downstream: near-dam peak reduction of 59% dropped to 7.9% at the basin outlet, and low pulse duration increase fell from 698% to 17%; the reservoir signal was largely diluted by unregulated tributary inflows contributing approximately 50% of outlet discharge. [Conclusions] The Dashixia Reservoir homogenizes the Aksu River flow regime through three systematic patterns: (i) intra-annual redistribution, shifting spring-summer runoff to autumn-winter and converting the natural summer flood and winter drought rhythm into an artificial summer drought and winter flood pattern; (ii) time-scale dependence, with positive effects on one to three day extreme minima but negative effects on seven to ninety day sustained low flows, along with stronger flood reduction for short durations than for long-duration flood volumes; and (iii) spatial attenuation, wherein reservoir effects decline exponentially downstream and are largely overwhelmed by unregulated tributary inflows. While the reservoir effectively prevents short-term river drying and reduces flood peaks, it simultaneously introduces prolonged low-flow stress, phenological mismatches, and loss of ecological disturbance functions. These findings provide quantitative evidence for developing ecologically informed reservoir operation strategies in arid regions.