Diagnosing and Assessing Uncertainties of the Carbon Cycle in Terrestrial Ecosystem Models from a Multi-Model Ensemble Experiment
Weile
Wang, CSUMB & NASA Ames Research Center, weile.wang@gmail.com
(Presenting)
Jennifer
L.
Dungan, NASA Ames Research Center, jennifer.l.dungan@nasa.gov
Hirofumi
Hashimoto, CSUMB & NASA Ames Research Center, hirofumi.hashimoto@gmail.com
Andrew
Michaelis, CSUMB & NASA Ames Research Center, amac@hyperplane.org
Cristina
Milesi, CSUMB & NASA Ames Research Center, cristina.milesi@gmail.com
Kazuhito
Ichii, Fukushima University (Japan), kazuhito.ichii@gmail.com
Ramakrishna
R.
Nemani, NASA Ames Research Center, rama.nemani@nasa.gov
We conducted an ensemble modeling exercise using the Terrestrial Observation and Prediction System (TOPS) to characterize structural uncertainty in estimated carbon fluxes and stocks from different ecosystem models. For this purpose, we developed the Hierarchical Framework for Diagnosing Ecosystem Models (HFDEM) that decomposes the simulated biogeochemistry in three cascaded functional tiers and sequentially examine their characteristics into the climate (temperature-precipitation) space and other diagrams. The key findings indicate that annual GPP/NPP among the tested models is largely comparable, and relatively optimal in climate regions where the relationship between annual temperature (T, oC) and precipitation (P, mm) is defined by P = 50*T+500. However, substantial differences are found in the simulated biomass, induced mainly by differences in model parameters and algorithms that regulate carbon allocation, tissue turnover, and plant mortality. Large diversity is also found in soil carbon, of which the mean residence time ranges from decades to centuries under similar climate conditions. Finally, non-respiratory disturbances like fires are the main driver for NEP, yet its magnitudes vary between different models. Overall, these findings indicate that although the structures of the tested models are similar, the uncertainties among them can be large, highlighting the problem inherent in relying on only one modeling approach to map surface carbon fluxes or to assess vegetation-climate interactions.
Presentation Type: Poster
Poster Session: Carbon Cycle Science
NASA TE Funded Awards Represented:
NONE: Related Activity or Previously Funded TE Award