System-level optimization of cultivar, sowing, and water–nitrogen management enhances cotton yield and resource efficiency under future warming in Xinjiang, China

System-level optimization of cultivar, sowing date, and water–nitrogen management for irrigated cotton in arid Xinjiang, China. Locally calibrated APSIM-Cotton with NSGA-III identifies robust climate-smart management strategies. Climate-smart management strategies are evaluated across Xinjiang, revealing regional potentials and informing adaptation of irrigated cotton to future warming. Cotton production in Xinjiang's irrigated arid regions faces growing challenges from climate-induced alterations in hydrothermal conditions, necessitating adjustments in cultivar selection, sowing time, and water–nitrogen management. However, most studies have focused on individual factors, with limited evaluation of integrated cultivar–sowing–water–nitrogen management. To assess the combined effects of irrigation and nitrogen management under future climates, a locally calibrated APSIM-Cotton model was driven by two CMIP6 scenarios (SSP2-4.5, SSP5-8.5) at two representative irrigated sites, Aral and Shihezi. The design included five maturity types, eight sowing dates (31 March–5 May), five irrigation thresholds (55–75% of field capacity, FC), and five fertigation levels (10–18 kg N ha −1 per irrigation). Multi-objective NSGA-III optimization identified optimal combinations for yield, water use efficiency (WUE), and nitrogen use efficiency (NUE), which were then extended to the regional scale across Xinjiang. Optimal strategies consistently converged on late-maturity cultivars with early-April sowing, high irrigation thresholds (0.65–0.75 FC), and relatively high per-event fertigation (14–18 kg N ha −1 ), a pattern robust across periods and both scenarios. Under the baseline climate, these optimal combinations increased yield by >32%, WUE by >21%, and NUE by >38% relative to conventional water–nitrogen management reported in previous field studies. Relative to the baseline (1981–2010) climate period, these strategies increased yield by >10%, WUE by >14%, and NUE by >21% under future scenarios. Regional simulations further revealed that southern Xinjiang holds greater potential for improving WUE, while northern Xinjiang is more advantageous in enhancing NUE. These findings highlight the synergistic effects between sowing time and water–nitrogen management, representing a key pathway to stabilizing yields and improving resource-use efficiency under future climate conditions, informing development of high-yield and efficient cotton systems in Xinjiang and offering insights for climate-adaptive management in similar arid regions worldwide.