Vegetation Water Stress Response-Driven Approach for Verifying Vegetation Drought Monitoring Results: A Case Study of the Yangtze River Basin, China

Drought is a major natural disaster in China, severely affecting agricultural output and the ecological environment. Timely and accurate vegetation drought monitoring is essential for early warning and ensuring agricultural production as well as ecological security. However, assessing the accuracy and seasonal applicability of vegetation drought monitoring models remains challenging due to the lack of clearly defined ground truth data. To address this issue, this study integrates soil moisture with a criterion reflecting the temporal response of vegetation growth, based on the vegetation's response to moisture stress, thereby deriving a reliable ground truth reference for validation. A Vegetation Water Stress Response-Driven Vegetation Drought Verification Approach (VWSR-VDVA) is proposed, and a case study in the Yangtze River Basin was done, with Surface Water Availability-Temperature Index (SWATI), Process-Cognizant Vegetation Drought Model (PCVDM), Temperature Vegetation Dryness Index (TVDI) and Standardized Soil Moisture Index (SSI) across different seasons from 2000 to 2022, for evaluation test. The results show that PCVDM achieved the highest overall accuracy, particularly in spring and autumn. SWATI tended to underestimate Vegetation Drought Zones (VDZs) during early wetting or drought recovery due to its sensitivity to short-term soil moisture variations and globally derived drought thresholds, but performed best in summer. SSI and TVDI generally overestimated VDZs due to coarse spatial resolution, fixed drought thresholds, and inadequate representation of vegetation response diversity. These findings are generally consistent with results from conventional correlation analyses. However, station-based validation often fails in specific areas—such as near-water areas, urban green spaces, and cultivated land—where vegetation conditions are influenced by artificial irrigation or localized water supplies, weakening the reliability of meteorological indicators alone. In contrast, the VWSR-VDVA captures the dynamic covariation between vegetation condition and water availability, effectively overcoming these limitations. It enables more accurate drought identification in heterogeneous environments. Overall, the VWSR-VDVA offers a more precise representation of vegetation drought conditions and provides valuable insights for optimizing drought monitoring approaches. It also lays a theoretical foundation for the development of future drought early warning and monitoring systems.