Effects of Climate Extremes on Spring Phenology of Temperate Vegetation in China

Abstract

The response of vegetation spring phenology to climate warming has received extensive attention. However, there are few studies on the response of vegetation spring phenology to extreme climate events. In this study, we determined the start of the growing season (SOS) for three vegetation types in temperate China from 1982 to 2015 using the Global Inventory Modeling and Mapping Study’s third-generation normalized difference vegetation index and estimated 25 extreme climate events. We analyzed the temporal trends of the SOS and extreme climate events and quantified the relationships between the SOS and extreme climate events using all-subsets regression methods. We found that the SOS was significantly advanced, with an average rate of 0.97 days per decade in China over the study period. Interestingly, we found that the SOS was mainly associated with temperature extremes rather than extreme precipitation events. The SOS was mainly influenced by the frost days (FD, r = 0.83) and mean daily minimum temperature (TMINMEAN, r = 0.34) for all three vegetation types. However, the dominant influencing factors were vegetation-type-specific. For mixed forests, the SOS was most influenced by TMINMEAN (r = 0.32), while for grasslands and barren or sparsely vegetated land, the SOS was most influenced by FD (r > 0.8). Our results show that spring phenology was substantially affected by extreme climate events but mainly by extreme temperature events rather than precipitation events, and that low temperature extremes likely drive spring phenology.

Keywords

• spring phenology
• extreme climate events
• remote sensing
• partial correlation analysis
• all-subsets regression