B024-04
Evaluating Terrestrial Biological Nitrogen Fixation in CMIP6 Earth System Models

Tuesday, 8 December 2020: 10:44
Virtual
Taraka Davies-Barnard, University of Exeter, Exeter, EX4, United Kingdom, Pierre Friedlingstein, University of Exeter, Exeter, United Kingdom, Sönke Zaehle, Max Planck Institute for Biogeochemistry, Jena, Germany, Victor Bovkin, Max Planck Institute for Meteorology, Hamburg, Germany, Yuanchao Fan, NORCE Norwegian Research Centre AS, Bergen, Norway, Rosie Fisher, National Center for Atmospheric Research, Boulder, CO, United States, Hanna Lee, Norwegian Research Centre (NORCE), Bjerknes Centre for Climate Research, Bergen, Norway, Daniele Peano, Euro-Mediterranean Center on Climate Change, Lecce, Italy, Benjamin Smith, Department of Physical Geography and Ecosystem Science, Lund University, Lund, Sweden, David Warlind, Lund University, Department of Physical Geography and Ecosystem Science, Lund, Sweden, Andy Wiltshire, Met Office Hadley Centre, Exeter, United Kingdom and Tilo Ziehn, CSIRO Oceans and Atmosphere, Aspendale, VIC, Australia
Abstract:
Biological nitrogen fixation (BNF) is a key contributor to sustaining the terrestrial carbon cycle, providing nitrogen input that plants require. This is relevant for projections of biogeochemical cycles as increased atmospheric carbon dioxide concentration may allow for ‘fertilisation’ if other plant requirements, such as nitrogen, do not prevent increases in productivity. The nitrogen cycle is a new feedback in most CMIP6 models and while studies show it improves aspects of the carbon cycle, many uncertainties remain. The mechanisms, global amount, and spatial distribution of BNF is highly disputed and consequently its process representation in land surface models varies. We evaluate the BNF representation in the CMIP6 earth system models against site-specific values and a new comprehensive meta-analysis of BNF field measurements that gives an upscaled global range. We find that models differ considerably in their estimates of present-day and recent past BNF, both spatially and temporally. The variation in productivity between earth system models means that productivity-based BNF functions result in differences in BNF between Earth System Models with the same BNF function almost as large as those between Earth System Models with contrasting BNF functions. Crucially, models disagree about both the signal and amplitude of the change in BNF over the twentieth century, reducing confidence in projections of the future. More observations and mechanistic understanding of BNF is required to help build understanding and reliable BNF modelling to contribute to robust earth system feedback projections.