Terrestrial biosphere models need better representation of vegetation phenology: results from the North American Carbon Program Site Synthesis

• DOI: 10.1111/j.1365-2486.2011.02562.x
• Publication Date: 2011-09-28T02:43:52Z
• Authors: Andrew D. Richardson; Ryan S. Anderson; M. Altaf Arain; Alan G. Barr; Gil Bohrer; Guangsheng Chen; Jing M. Chen; Philippe Ciais; Kenneth J. Davis; Ankur R. Desai; Michael C. Dietze; Danilo Dragoni; Steven R. Garrity; Christopher M. Gough; Robert Grant; David Y. Hollinger; Hank A. Margolis; Harry McCaughey; Mirco Migliavacca; Russell K. Monson; J. William Munger; Benjamin Poulter; Brett M. Raczka; Daniel M. Ricciuto; Alok K. Sahoo; Kevin Schaefer; Hanqin Tian; Rodrigo Vargas; Hans Verbeeck; Jingfeng Xiao; Yongkang Xue
• Keywords: Seasonality; Growing season

Abstract Phenology, by controlling the seasonal activity of vegetation on land surface, plays a fundamental role in regulating photosynthesis and other ecosystem processes, as well competitive interactions feedbacks to climate system. We conducted an analysis evaluate representation phenology, associated seasonality ecosystem‐scale CO 2 exchange, 14 models participating N orth A merican C arbon P rogram S ite ynthesis. Model predictions were evaluated using long‐term measurements (emphasizing period 2000–2006) from 10 forested sites within meri F lux luxnet‐ anada networks. In deciduous forests, almost all consistently predicted that growing season started earlier, ended later, than was actually observed; biases weeks or more typical. For these sites, most also unable explain small fraction observed interannual variability phenological transition dates. Finally, for misrepresentation cycle resulted over‐prediction gross +160 ± 145 g m −2 yr −1 during spring +75 130 autumn (13% 8% annual productivity, respectively) compensating tendency under‐predict magnitude peak summertime photosynthetic rates. Models did better job predicting exchange evergreen forests. These results highlight need improved understanding environmental controls phenology incorporation this knowledge into models. Existing are unlikely predict future responses change accurately therefore will misrepresent key biosphere–atmosphere coupled global