Dust aerosols pose a major environmental challenge to photovoltaic (PV) power generation in arid and semi-arid regions, where accurate quantification is essential for optimizing PV resource allocation. This study develops an integrated framework linking historical conditions with future projections by combining reanalysis datasets and CMIP6 multi-model ensembles to evaluate dust impacts on monthly PV potential, spatial suitability, and long-term trends across Central Asia from 1980 to 2024. Results show a clear south-north gradient in ideal PV potential, with monthly maxima above 45 kWh m(-2) month(-1) and strong seasonal variability (summer: 35.41; winter: 13.38 kWh m(-2) month(-1)). Dust deposition consistently reduces PV output, with attenuation reaching 14.48% in winter and 3.20% in summer along the southern Tarim Basin. Mechanistic analyses reveal nonlinear influences of temperature, potential evapotranspiration, and aridity; LightGBM identifies temperature and evapotranspiration as dominant drivers, while Geodetector highlights thermo-hydrological interactions in winter. Spatial suitability assessments indicate generally high PV potential (scores >0.5 in Uzbekistan, Turkmenistan, and most of Xinjiang), although dust reduces suitability to 0.2-0.3 in northwestern and southern Xinjiang. The southeastern Tarim Basin remains a favorable deployment zone. Future projections show SSP1-2.6 yielding the highest PV potentials (near-term: 37.89; long-term: 38.48 kWh m(-2) month(-1)), whereas SSP3-7.0 produces the lowest (36.89 kWh m(-2) month(-1)), driven by warming and weakened winds. While results clarify dust-induced attenuation and spatial heterogeneity, the analysis relies on satellite-derived products, and further ground-based validation is needed to assess applicability at finer spatial scales.