Participation in the AmeriFlux Real-Time Modeling Effort
In response to the need for improved regional assessment of biospheric responses to increasing atmospheric CO2 concentrations worldwide, eddy covariance flux tower researchers (AmeriFlux) and ecological modelers (Biome-BGC, LoTEC, and PnET-DAY) began a collaborative effort to provide a structure for the continuous monitoring of the terrestrial biosphere [Running et al., 1999]. This activity, known as the real time modeling effort, was initiated In October 2000. As part of the real time modeling effort, participating eddy covariance tower researchers voluntarily provide standardized micrometeorological data to the Carbon Dioxide Information Analysis Center (CDIAC) for dissemination to the participating modelers. Model results of daily evapotranspiration (ET), net ecosystem exchange (NEE), gross primary productivity (GPP) and net primary productivity (NPP) are simulated and then submitted to CDIAC to be posted on the internet site on a weekly basis for intercomparison of measured and modeled results as well as for use by the AmeriFlux community. Ecosystem process models such as Biome-BGC are used to create data gap filling strategies; to provide the component fluxes of both ET and NEE; to identify the relationship between measured NEE and plant biomass; and to resolve scaling issues, allowing point measurements from eddy flux towers to be extrapolated to regional scales. Weekly estimates from the Biome-BGC ecosystem process model have three benefits for assessing the impacts of environmental change by:
- estimating NEE, model results will be used to fill missing tower data;
- providing estimates of NPP and GPP, the model will be used in association with tower measurements to find relationships between NEE, plant biomass and underlying physical and environmental controls and component processes; and
- providing a framework for scaling tower measurements to the regional level and over longer (decades to centuries) time periods, where the effects of increasing CO2 and associated climate change may be most noticeable.
Our current aim is to improve upon the present method for validating new and existing AmeriFlux eddy covariance tower sites and integrating these data with ecosystem process models in real time for improved understanding of vegetative productivity, carbon and hydrological cycle dynamics, one of the goals of the North American Carbon Plan. Eddy covariance towers provide measurements of net ecosystem CO2 exchange (NEE), but it is difficult to relate these measurements directly to vegetative productivity. Site measurements of net primary productivity (NPP) are both time-consuming and expensive. In addition, site measurements can provide only annual estimates of NPP. Estimates of gross primary productivity (GPP) require knowledge of site-specific ecosystem respiration and its inherent interaction with site vegetation. Similarly, site eddy covariance measurements of canopy H20 fluxes and evapotranspiration (ET) are unable to distinguish component fluxes (i.e., transpiration, canopy and soil evaporation) and underlying processes. Models, including the ecosystem process model Biome-BGC, use tower meteorological data and general site characteristics to provide daily estimates of ET, NEE, NPP, GPP and their component fluxes, as well as quantitative linkages to underlying processes and environmental controls for biomes ranging from forests to grasslands. There are nearly twenty participating flux towers within the AmeriFlux network to validate and improve the Biome-BGC model for these specific sites. The Biome-BGC ecosystem process model has been created to simulate the flux and storage of ecosystem carbon, nitrogen, and water in the soil-vegetation-atmosphere continuum over time, and the model logic is summarized in Figure 1. Current model simulations rely upon generalized parameterizations developed by White et al. . Validation for each site will be done by using eddy covariance and meteorological measurements made at a site for the years 2000-2002 or, for new sites, prior to its joining the modeling effort to ensure the parameterization adequately estimates site measurements.
Results of the current simulation protocol from 2001 indicate that the Biome-BGC model captures the seasonality of both NEE and GPP data when compared to site estimates, although the model tends to overestimate both NEE and GPP at most contributing tower sites. These estimates can be improved by providing the model with the site-specific parameterizations mentioned previously. In addition, the Park Falls site is a mixed forest, containing both deciduous broadleaf and evergreen needleleaf species, but is currently modeled as a deciduous broadleaf forest. Using the spatial ensembling technique of Law et al.  and combining non-interactive simulations of both forest types, it will be possible to improve the estimates of vegetative productivity at this site.
Law, B.E., P.E. Thornton, J. Irvine, P.M. Anthoni, and S. Van Tuyl, Carbon storage and fluxes in ponderosa pine forests at different developmental stages, Global Change Biology, 7, 755-777, 2001.
Running, S.W., D.D. Baldocchi, D.P. Turner, S.T. Gower, P.S. Bakwin, and K.A. Hibbard, A global terrestrial monitoring network integrating tower fluxes, flask sampling, ecosystem modeling and EOS satellite data, Remote Sensing of Environment, 70, 108-127, 1999.
Thornton, P.E., B.E. Law, H.L. Gholz, K.L. Clark, E. Falge, D.S. Ellsworth, A.H. Goldstein, R.K. Monson, D. Hollinger, J.C. Paw U, and J.P. Sparks, Modeling and measuring the effects of disturbance history and climate on carbon and water budgets in evergreen needleleaf forests, Agricultural and Forest Meteorology, 113, 185-222, 2002.
White, M.A., P.E. Thornton, S.W. Running, and R.R. Nemani, Parameterization and sensitivity analysis of the BIOME-BGC terrestrial ecosystem model: net primary production controls, Earth Interactions, 4 (3), 1-84, 2000.