Global warming is accelerating the global hydrological cycle, yet current assessments remain constrained by static water-balance methods that overlook dynamic flux rates and yield fragmented insights. To address this gap, we propose a novel rate of hydrological change (RHC) index that systematically quantifies hydrological acceleration in arid inland river basins of Northwest China. By integrating multi-source observations, models, and reanalysis data, our method provides a unified perspective on the coupled atmosphere-to-runoff system, including water-phase changes. Results demonstrated a consistent acceleration of the regional water cycle. Process-based variables, including glacial mass loss (+58.36%) and evapotranspiration (+26.83%), exhibited more pronounced increases than state variables, reflecting a stronger response of phase-change processes within the accelerated hydrological cycle. The spatially weighted hydrological change rates reached 22.58% based on state variables (RHCs) and 25.61% based on process metrics (RHCp), with clear spatial gradients observed across mountain systems. Glacial melt and precipitation-to-runoff conversion were identified as the primary drivers of acceleration. Large-scale climate forcings, particularly atmospheric circulation patterns and ocean-atmosphere interactions, dominated hydrological variability, accounting for 73% to 80% of the explained variance, significantly surpassing the influence of local climate factors and human activities. Causal loop analysis further revealed bidirectional feedback mechanisms between water cycle acceleration and drought subsystems, showing that intensification of process variables amplified meteorological drought while buffering hydrological drought. Our study provides a systematic and mechanism-based perspective on hydrological cycle intensification in arid regions. The RHC framework offers a transferable approach to diagnosing complex hydrological responses to climate change, underscoring the need for process-based metrics in future water resource vulnerability assessments.