Coupled spatiotemporal variability of temperature and spring phenology in the Eastern United States

While warmer spring temperatures frequently led to earlier vegetation phenology during the past decade in the eastern United States, detailed geographic patterns of this coupled relationship have not been investigated previously. Additionally, a record early greenup onset that occurred in the U.S. Northeast in 2010 and over a majority of the eastern United States in 2012 offered unique opportunities to assess spatiotemporal variations of phenology-climate interactions in the context of climate change. In this study, we examined interannual variations of land surface phenology (LSP) of the forested eastern U.S., coupled with both spring and winter temperature changes from 2000 to 2012, highlighting the different geographic patterns of phenological responses in 2010 and 2012. We employed S-mode canonical correlation analysis and extended principal component analysis to detect the coupled spatiotemporal variations of LSP and cumulative temperature variables. For the U.S. Northeast, these methods allowed us to identify potentially offsetting interactions in which reduced chilling during a warm 2011-2012 winter delayed dormancy break and lessened the advance in greenup onset during the very warm spring in 2012. In the Upper Midwest, however, winter chilling seemed sufficient and greenup closely tracked springtime warming. These geographic details of phenology-climate interactions unveiled how vegetation phenology is closely coupled with both spring and winter temperatures in time and space. This study further implied that both different temperature regimes and geographically varied climatic dependencies of vegetation could be responsible for spatially diverse phenological responses to climate change.