Decoding Drought: The Full-Chain Propagation Process from Atmosphere to Groundwater in Arid Ecosystems

Drought is the primary ecological-hydrological stress in arid regions, and its impacts are often not confined to a single hydrological component but are transmitted through various land surface system components, leading to cumulative effects. To explore the continuous coupling between different types of droughts and their differential propagation mechanisms across various ecosystems, this study focuses on the Three-North region. Based on the Standardized Precipitation Evapotranspiration Index (SPEI), Standardized Runoff Index (SRI), Surface Soil Moisture Drought Index (SMDIs), Root-zone Soil Moisture Drought Index (SMDIrz), and Groundwater Drought Index (SGI), the study introduces Convergent Cross Mapping (CCM) to systematically characterize the causal coupling relationships between different types of droughts and their propagation structural characteristics across different vegetation types.

The results show that significant continuous coupling relationships exist between different types of droughts, generally presenting the structural pattern of SPEI ↔ SRI ↔ SMDIs ↔ SMDIrz ↔ SGI. This indicates that drought signals in the “meteorological—runoff—soil—root-zone—groundwater” system do not evolve in isolation but are continuously transmitted through various components of the land surface system, forming a typical long-chain drought propagation process. At the regional level, root-zone soil moisture plays the most crucial role in the drought propagation network. It not only significantly responds to upstream meteorological and hydrological droughts but also has an important modulation effect on groundwater drought, acting as a key intermediary between surface hydrological processes and the groundwater system. Significant differences in drought propagation structures exist under different vegetation types: in forest and shrubland areas, the causal effect of runoff drought on root-zone soil moisture drought is the strongest, reflecting the critical control of deep moisture processes on plant-available water. The coupling intensity is relatively weak in grassland systems, exhibiting drought dynamics with rapid response and weak memory. In areas with sparse vegetation and bare soil, there is a high degree of bidirectional coupling between runoff drought and surface soil moisture drought, indicating that the drought process is mainly driven by surface hydrological processes.