Maize production in arid Northwestern China is increasingly constrained by water scarcity and rising carbon emissions. Inefficient water and nitrogen management practices exacerbate resource waste and greenhouse gas (GHG) emissions, hindering progress toward sustainable agriculture and carbon neutrality. Consequently, there is an urgent need to optimize integrated water-nitrogen management strategies to simultaneously enhance crop yield and water use efficiency (WUE) while mitigating GHG emissions. In this study, field experiments combined with the DeNitrification-DeComposition (DNDC) model were used to assess the effects of water-nitrogen coupling on maize yield and GHG emissions. A field experiment was conducted with three irrigation gradients: severe water deficit (W1: 45-60 % Of), moderate water deficit (W2: 60-75 % Of), and mild water deficit (W3: 75-90 % Of, where Of denotes field capacity); and three nitrogen application rates: low (F1: 120 kg/ha), medium (F2: 240 kg/ ha), and high (F3: 360 kg/ha). The DNDC model was calibrated and validated using field data from 2023 to 2024, and was then linked with four Shared Socioeconomic Pathways (SSPs) to project maize yield and GHG emissions from 2025 to 2100. Results indicated that N2O and CO2 emissions were significantly affected by water-nitrogen interactions (P < 0.05), whereas CH4 fluxes remained a weak sink and showed no significant response to the treatments (P > 0.05). The higher cumulative N2O and CO2 emissions observed in the second year were primarily attributed to variations in water-filled pore space (WFPS), whereas soil temperature showed no significant correlation with N2O emissions. Compared with high nitrogen input, a moderate nitrogen application rate significantly reduced N2O and CO2 emissions. Across irrigation regimes, global warming potential (GWP) increased progressively with increasing water and nitrogen inputs. Nitrogen application rate was the dominant controlling factor for greenhouse gas intensity (GHGI) under wet conditions, whereas water deficit severity was dominant under drought. The F2W3 treatment achieved the highest maize yield and significantly enhanced WUE and key growth traits. This management strategy is projected to support yield increases and emissions reduction under the SSP1-2.6 and SSP2-4.5 scenarios. However, under high-emission scenarios, maize yield is projected to decline significantly alongside increased GHG emissions, with soils shifting from a CH4 sink to a source. Based on the combined results of the Mann-Kendall trend test and VIKOR multi-criteria decision analysis, F2W3 was identified as the optimal strategy, simultaneously achieving high yield, improved WUE, and lower GHG emissions. These findings provide a robust scientific basis for sustainable maize production and carbon-neutral agricultural practices in arid Northwestern China.