Toward "optimal" integration of terrestrial biosphere models. Issue 11 (10th June 2015)

• Record Type: Journal Article
• Authors: Schwalm, Christopher R.; Huntzinger, Deborah N.; Fisher, Joshua B.; Michalak, Anna M.; Bowman, Kevin; Ciais, Philippe; Cook, Robert; El‐Masri, Bassil; Hayes, Daniel; Huang, Maoyi; Ito, Akihiko; Jain, Atul; King, Anthony W.; Lei, Huimin; Liu, Junjie; Lu, Chaoqun; Mao, Jiafu; Peng, Shushi; Poulter, Benjamin; Ricciuto, Daniel; Schaefer, Kevin; Shi, Xiaoying; Tao, Bo; Tian, Hanqin; Wang, Weile; Wei, Yaxing; Yang, Jia; Zeng, Ning
• Is Part Of: Geophysical research letters. Volume 42:Issue 11(2015:Jun.)
• Page Start: 4418
• Page End: 4428
• Publication Date: 2015-06-10
• Subjects: modeling -- carbon cycle -- model integration; Geophysics -- Periodicals; Planets -- Periodicals; Lunar geology -- Periodicals; 550
• DOI: 10.1002/2015GL064002
• Languages: English
• ISSNs: 0094-8276

Abstract

Abstract: Multimodel ensembles (MME) are commonplace in Earth system modeling. Here we perform MME integration using a 10‐member ensemble of terrestrial biosphere models (TBMs) from the Multiscale synthesis and Terrestrial Model Intercomparison Project (MsTMIP). We contrast optimal (skill based for present‐day carbon cycling) versus naïve ("one model‐one vote") integration. MsTMIP optimal and naïve mean land sink strength estimates (−1.16 versus −1.15 Pg C per annum respectively) are statistically indistinguishable. This holds also for grid cell values and extends to gross uptake, biomass, and net ecosystem productivity. TBM skill is similarly indistinguishable. The added complexity of skill‐based integration does not materially change MME values. This suggests that carbon metabolism has predictability limits and/or that all models and references are misspecified. Resolving this issue requires addressing specific uncertainty types (initial conditions, structure, and references) and a change in model development paradigms currently dominant in the TBM community. Key Points: Naive (simple mean) and skill‐based multi‐model ensemble integration are indistinguishable This holds at the grid cell level and globally for land sink strength and reference variables Carbon metabolism has predictability limits and/or models/references are misspecified