OOS 87-3
Using conceptual theory and observations to evaluate mechanisms of N limitation in Earth system models

Friday, August 14, 2015: 8:40 AM
327, Baltimore Convention Center
R. Quinn Thomas, Forest Resources and Environmental Conservation, Virginia Tech, Blacksburg, VA

Nitrogen (N) limitation of the terrestrial carbon (C) cycle is widespread across the globe.  Consequently, most Earth system models now include representations of C-N interactions to improve predictions of carbon uptake and climate change.  These global models with C-N cycles have shown that both the terrestrial C sink and climate are sensitive to the availability of N, but differ markedly in their predicted sensitivity.  In order to evaluate and improve model predictions of N limitation, new datasets and conceptual approaches are needed. Here, I discuss how a range of data sources can be used to evaluate multiple aspects of the coupled C and N cycles in Earth system models.  I also discuss how, in light of a paucity of global data on C-N interactions, conceptual theory of N limitation can be valuable for evaluation. 


Using simulations from multiple Earth system models, I demonstrate that N addition experiments, 15N tracer studies, small watershed scale N input-output budgets, and forest responses to N deposition gradients can all be used to evaluate different processes in coupled C-N interactions. For example, comparing the output of two models to N addition experiments and forest C responses across N deposition gradients helped simultaneously evaluate N retention capacity and non-N limited productivity. Finally, I show that Earth system models have used a range of approaches to represent the conceptual theory of N limitation.  Models particularly differ in how they simulate N losses that are unable to be controlled by plants, a strong driver of N limitation, and in how they simulate N fixation.  In conclusion, the high spatial and temporal variability of the N cycle and the lack of globally integrative datasets for Earth system model evaluation necessitates the use of multiple data sources and conceptual theory to describe the quality of coupled C-N cycle predictions.