Geographically Variable Response of Land Surface Phenology to 2012 Anomalous Spring Warming

The spring of 2012 was marked by an unusually warm spring followed by historically earlier than normal plant phenology across a large part of the United States. While it is understood that warmer spring advances plant phenology, the spatial variation of the phenological response to warming and the underlying geographic controls of such variations remain unclear. The significant anomaly in 2012 also provides a natural experiment to reveal how plant phenology would respond to abrupt climate change. This study utilizes remotely sensed land surface phenology (LSP) and cloned lilac-based phenology models (extended spring indices, or SI-x) to investigate the national patterns of 2012 spring phenology anomalies. Results suggest that regional variations of phenological shift not only correspond to the spring temperature anomalies, but are also related to the varying climatic requirements of greenup at different geographic locations. Particularly, in most years with the spring temperature within the normal range, observed LSP tends to show more gradual change than that of modeled phenology along a cold-to-warm climate gradient. However, in 2012 (and 2010 with a minor anomaly in parts of the country), LSP demonstrated a bimodal departure pattern relative to the SI-x prediction. This implies that during an abruptly warmer spring, northern region plant phenology may be delayed rather than advanced due to a dampened chilling fulfillment. Consequently, the lower Midwest and Mid-south of the United States experienced the greatest amount of phenological change in 2012 due to the combined effect of sufficient chilling and maximum warming/forcing. The fundamental principle that can be used to explain such phenomena may be embodied in the growth efficiency hypothesis, which states that plants adapted to colder climates tend to require less heat accumulation but more chilling buildup for growing season onset, and therefore are more energy efficient (and vice versa for plants in the warmer climates). This study focuses on temperature-driven phenology and has not investigated the effect of drought or phenology in semiarid parts of the country. A better understanding of the geographical variations in phenology-climate interaction mechanisms facilitates the spatially explicit forecasting of climate change impacts on the biosphere.