The Carbon-Land Model Intercomparison Project (C-LAMP): A Protocol and Evaluation Metrics for Global Terrestrial Biogeochemistry Models

2008 
Described here is a protocol and accompanying metrics for evaluation of scientific model performance of global terrestrial biogeochemistry models. Developed under the auspices of the NCAR Community Climate System Model (CCSM) Biogeochemistry Working Group, the Carbon-Land Model Intercomparison Project (C-LAMP) experimental protocol improves and ex- pands upon the Coupled Carbon Cycle-Climate Model Intercomparison Project (C 4 MIP) Phase 1 protocol. However, unlike traditional model intercomparisons, C-LAMP has established sci- entific model performance metrics based upon comparison against best-available satellite- and ground-based measurements. Moreover, C-LAMP has partnered with the U.S. Department of En- ergy's Program for Climate Model Diagnosis and Intercomparison (PCMDI) to collect, archive, and distribute—via the Earth System Grid (ESG)—model results from C-LAMP experiments per- formed by international modeling groups in the same fashion as was done for the model results used in the Intergovernmental Panel on Climate Change (IPCC) Fourth Assessment Report (AR4). In addition, because future IPCC Assessment Reports are expected to be based on results from integrated Earth System Models (ESMs), C-LAMP is helping to establish the metadata standards for model output from terrestrial biogeochemistry components of ESMs. Proposed as an extension to the netCDF Climate and Forecast (CF) 1.1 Convention, these metadata standards will facilitate future model-model and model-measurement intercomparisons. A prototype diagnostics tool has been developed for C-LAMP that summarizes model results, produces graphical representations of these results as compared with observational data sets, and scores models on their scientific performance.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    11
    References
    8
    Citations
    NaN
    KQI
    []