Responding time scales of vegetation production to extreme droughts over China

Extreme drought events have caused extensive and severe impacts on terrestrial ecosystem in last decades in China. Given droughts may be more intense and frequent under future climate change, accurate assessment of the drought impact on vegetation primary production can provide reliably scientific supports for the carbon sink potential. Numerous existing studies have used Standardized Precipitation Evapotranspiration Index (SPEI) to discover the drought-production relationships, however, most of them just considered the strongest correlation between production and different time scales (i.e. correlation-based method), which may underestimate the production loss because of the asymmetric responses under dry and wet conditions. In this work, we proposed a new method which assumed that the dominant time scale should correspond to the lowest primary production during each drought year (extreme-based method). Based on six independent Gross Primary Productivity (GPP) products and SPEI dataset, it showed that the extreme-based method was more reasonable and robust (with a larger inter-consistency of 0.50 than that of 0.05 for correlation-based method) to determine at which time scale GPP predominantly responded to extreme droughts. And the GPP loss can be underestimated by 45 +/- 26% (mean +/- s.d.) if the time scale was randomly selected. Furthermore, spatial analysis suggested that vegetation type, water balance and soil textures mainly affected the spatial heterogeneity of the dominant time scales. In detail, forests, humid biomes, and vegetation planted in loam tended to be more sensitive to longer-term droughts. This study highlighted that optimal time-scale selection using extreme-based assumption can give more accurate estimation of the drought impacts on vegetation primary production.

Automated summary

Extreme drought events in China have had extensive and severe impacts on terrestrial ecosystems in recent decades. This study proposes a new method, called the extreme-based method, to determine the dominant time scale at which gross primary productivity (GPP) responds to extreme droughts. The results show that this method is more reasonable and robust compared to the correlation-based method. The study also highlights the importance of considering vegetation type, water balance, and soil textures in understanding the spatial heterogeneity of dominant time scales.