Spatiotemporal analysis of ground-based woody plant leafing in response to temperature in temperate eastern China.

The analysis of woody plant leafing in response to regional-scale temperature variation using ground-based phenology is usually limited by the sparse coverage and missing data of ground observation. In this study, a station-based multispecies method was proposed to generate spatiotemporal variation of woody plant leafing date using ground observations from the Chinese Phenological Observation Network during 1974-1996. The results show that the leafing date had slightly insignificant advance (-0.56 day decade(-1)), and the Arctic Oscillation (AO) index could explain 36% variance of the spring leafing date anomaly. The leafing date had been substantially delayed (4 days) when AO shifted from an extreme high index state (2) in 1989-1990 to a relatively low state (0.1) in 1991-1996. The canonical correlation analysis (CCA) was used to demonstrate the temporal evolutions and spatial structures of interannual variations of the spring temperature and leafing date anomalies. The three CCA spatial patterns of leafing date anomaly are similar to those of spring temperature anomaly. The first spatial pattern shows ubiquitous warming, which is consistent with the ubiquitous advance in leafing date across the study area. The second and third spatial patterns present the regional differences featured by advanced (delayed) leafing associated with high (low) temperature. The results suggest that the spring leafing date anomaly is spatiotemporally coherent with the regional-scale temperature variations. Although we focus here on woody plant leafing in a historical period in temperate eastern China, our station-based multispecies method may be applicable to analysis of the ground-based phenology in response to regional-scale climatic variation in other regions.