Comparing the Performance of Three Land Models in Global C Cycle Simulations: A Detailed Structural Analysis: Structural Analysis of Land Models
- Department of Microbiology and Plant Biology, University of Oklahoma, Norman OK USA; Joint Global Change Research Institute, Pacific Northwest National Lab, College Park, MD USA
- School of Ecological and Environmental Science, East China Normal University, Shanghai China; Research Center for Global Change and Ecological Forecasting, East China Normal University
- Department of Microbiology and Plant Biology, University of Oklahoma, Norman OK USA
- Joint Global Change Research Institute, Pacific Northwest National Lab, College Park, MD USA
Land models are valuable tools to understand the dynamics of global carbon (C) cycle. Various models have been developed and used for predictions of future C dynamics but uncertainties still exist. Diagnosing the models’ behaviors in terms of structures can help to narrow down the uncertainties in prediction of C dynamics. In this study three widely used land surface models, namely CSIRO’s Atmosphere Biosphere Land Exchange (CABLE) with 9 C pools, Community Land Model (version 3.5) combined with Carnegie-Ames-Stanford Approach (CLM-CASA) with 12 C pools and Community Land Model (version 4) (CLM4) with 26 C pools were driven by the observed meteorological forcing. The simulated C storage and residence time were used for analysis. The C storage and residence time were computed globally for all individual soil and plant pools, as well as net primary productivity (NPP) and its allocation to different plant components’ based on these models. Remotely sensed NPP and statistically derived HWSD, and GLC2000 datasets were used as a reference to evaluate the performance of these models. Results showed that CABLE exhibited better agreement with referenced C storage and residence time for plant and soil pools, as compared with CLM-CASA and CLM4. CABLE had longer bulk residence time for soil C pools and stored more C in roots, whereas, CLM-CASA and CLM4 stored more C in woody pools due to differential NPP allocation. Overall, these results indicate that the differences in C storage and residence times in three models are largely due to the differences in their fundamental structures (number of C pools), NPP allocation and C transfer rates. Our results have implications in model development and provide a general framework to explain the bias/uncertainties in simulation of C storage and residence times from the perspectives of model structures.
- Research Organization:
- Pacific Northwest National Lab. (PNNL), Richland, WA (United States)
- Sponsoring Organization:
- USDOE
- DOE Contract Number:
- AC05-76RL01830
- OSTI ID:
- 1344042
- Report Number(s):
- PNNL-SA-116497
- Journal Information:
- Land Degradation and Development, Vol. 28, Issue 2; ISSN 1085-3278
- Publisher:
- Wiley
- Country of Publication:
- United States
- Language:
- English
Similar Records
Divergent predictions of carbon storage between two global land models: Attribution of the causes through traceability analysis
Evaluating the Community Land Model in a pine stand with shading manipulations and 13CO2 labeling