Drought-related cumulative and time-lag effects on vegetation dynamics across the Yellow River Basin, China

Abstract

• Accumulated (lagged) drought significantly affected 50% (60%) of the vegetated area in the YRB. • Vegetation in the arid zone tended to be more sensitive and resistant to drought. • Grasslands and crops were more vulnerable to accumulated and lagged drought than forests. • Cumulative effect was stronger than lag effect independent of climate zone and vegetation type. • Cumulative and lag effects were significantly correlated with annual water availability. Yellow River Basin (YRB), a climate-sensitive and ecologically compromised area in China, is increasingly affected by extreme climate events (especially droughts) resulting from climate change and frequent human activity. Vegetation responds asymmetrically to drought with cumulative and time-lag effects, whereas response across various climatic zones and diverse vegetation types in the YRB remains unclear. To address this deficiency, we examined the spatiotemporal patterns of accumulated and lagged drought effects on vegetation dynamics for the period 1982 to 2015. The examination was based on the long-term Normalized Difference Vegetation Index (NDVI) and multiscale dataset of the Standardized Precipitation Evapotranspiration Index (SPEI). Cumulative (time-lag) effects were determined via the maximum correlation between the NDVI and the one- to 12-month timescale SPEI (one-month timescale SPEI), as well as the corresponding months of optimal response to drought. The main findings were as follows: (1) Accumulated and lagged drought significantly affected approximately 50% and 60% of the vegetated area in the YRB, respectively, with the strongest effects and the corresponding optimal months varying across climatic zones and vegetation types. (2) In general, vegetation in the arid zone tended to be more sensitive and resistant to drought, as evidenced by the occurrence of lagged and accumulated drought effects mostly in the short-term (one–three months) and medium-term (six–eight months), respectively. This finding may be related to the vegetation’s strategy for coping with water deficits. (3) The biome-level effects of drought on grassland and cultivated vegetation were stronger than those on forests, which may be associated with differences in the functional characteristics of root systems. (4) Annual water availability significantly affected the spatiotemporal patterns by which the NDVI responded to droughts of multiple timescales, with the correlation coefficients and corresponding months of response decreasing with increasing average annual SPEI. These results indicate that vegetated areas with low water availability were more susceptible to cumulative (time-lag) droughts. (5) Independent of the climate zone or vegetation type, drought cumulatively affected vegetation more than time-lag effects. The study improves the knowledge of climate–vegetation relationships in the YRB and provides theoretical support for addressing drought risk in a changing climate.