Multi-scale groundwater dynamics and irrigation-induced depletion in arid oasis regions: Insights from the Shiyang River Basin, China

In arid inland oases, groundwater serves as the primary water resource sustaining both agricultural irrigation and ecosystem stability, yet its multi-scale temporal dynamics and depth-dependent response patterns remain insufficiently understood, hindering effective adaptive water management. Taking the Shiyang River Basin in northwestern China—a typical intensively irrigated inland river basin—as a case study, this research integrates long-term monitoring data from 52 wells (2010–2024 monthly observations and 2022–2024 daily records) with climatic variables (precipitation, evaporation), irrigated area expansion, and population growth records. Trend analysis, cross-correlation methods, and spatiotemporal modeling are employed to systematically investigate vertical response mechanisms, spatial heterogeneity, and the relative contributions of anthropogenic versus natural drivers across different aquifer depths and sub-regions. The results reveal that: (1) Approximately 75% of wells exhibited significant declining trends primarily driven by intensive agricultural exploitation, with maximum depletion rates reaching − 0.5 m/yr in downstream irrigated areas; (2) A novel depth-stratified response pattern was identified: shallow aquifers (< 10 m) respond immediately to irrigation events with daily fluctuations up to 0.41 m, enabling real-time monitoring of agricultural water use; mid-depth aquifers (10–30 m) function as natural buffers showing 1–2 month lagged responses to irrigation demand; deep aquifers (> 60 m) demonstrate sustained, irreversible declines exceeding − 2.7 m/yr, indicating unsustainable groundwater mining; (3) Evaporation exhibited the strongest correlation with groundwater variations at 1-month lag (r = 0.724), while irrigated area expansion and population growth exerted significant cumulative impacts within 2–3 months; (4) Clear spatial differentiation exists: human activities (primarily irrigation) dominate groundwater dynamics in downstream agricultural regions with 1–3 month time lags, whereas upstream groundwater remains primarily governed by natural recharge processes. Based on these depth-stratified and spatially heterogeneous response patterns, we propose a differentiated management framework tailored to irrigation systems: "real-time regulation for shallow aquifers—buffer storage management for intermediate aquifers—strategic protection for deep aquifers," coupled with zonally differentiated strategies that prioritize irrigation efficiency improvements downstream and natural recharge conservation upstream. These findings provide scientific evidence and actionable technical support for dynamic groundwater regulation, optimized irrigation scheduling, and sustainable water allocation in arid agricultural regions facing water scarcity challenges.