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.
Original language | English (US) |
---|---|
Pages (from-to) | 4418-4428 |
Number of pages | 11 |
Journal | Geophysical Research Letters |
Volume | 42 |
Issue number | 11 |
DOIs | |
State | Published - Jun 16 2015 |
Fingerprint
Keywords
- carbon cycle
- model integration
- modeling
ASJC Scopus subject areas
- Earth and Planetary Sciences(all)
- Geophysics
Cite this
Toward "optimal" integration of terrestrial biosphere models. / 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.
In: Geophysical Research Letters, Vol. 42, No. 11, 16.06.2015, p. 4418-4428.Research output: Contribution to journal › Article
}
TY - JOUR
T1 - Toward "optimal" integration of terrestrial biosphere models
AU - Schwalm, Christopher R
AU - Huntzinger, Deborah N
AU - Fisher, Joshua B.
AU - Michalak, Anna M.
AU - Bowman, Kevin
AU - Ciais, Philippe
AU - Cook, Robert
AU - El-Masri, Bassil
AU - Hayes, Daniel
AU - Huang, Maoyi
AU - Ito, Akihiko
AU - Jain, Atul
AU - King, Anthony W.
AU - Lei, Huimin
AU - Liu, Junjie
AU - Lu, Chaoqun
AU - Mao, Jiafu
AU - Peng, Shushi
AU - Poulter, Benjamin
AU - Ricciuto, Daniel
AU - Schaefer, Kevin
AU - Shi, Xiaoying
AU - Tao, Bo
AU - Tian, Hanqin
AU - Wang, Weile
AU - Wei, Yaxing
AU - Yang, Jia
AU - Zeng, Ning
PY - 2015/6/16
Y1 - 2015/6/16
N2 - 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.
AB - 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.
KW - carbon cycle
KW - model integration
KW - modeling
UR - http://www.scopus.com/inward/record.url?scp=84934439686&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84934439686&partnerID=8YFLogxK
U2 - 10.1002/2015GL064002
DO - 10.1002/2015GL064002
M3 - Article
AN - SCOPUS:84934439686
VL - 42
SP - 4418
EP - 4428
JO - Geophysical Research Letters
JF - Geophysical Research Letters
SN - 0094-8276
IS - 11
ER -