Influences of FPAR source, number of reference sites, ecoregion configuration, and meteorological data source on parameter optimization of a diagnostic carbon cycle model
David
P
Turner, Oregon State University, david.turner@oregonstate.edu
(Presenting)
David
King, Oregon State University, david.king@oregonstate.edu
David
Ritts, Oregon State University, david.ritts@oregonstate.edu
Diagnostic carbon cycle models integrate information from meteorology, soil properties, and satellite observations. Spatial-mode application of the models thus permits evaluation of spatial patterns and interannual variation in carbon transfer between the land and the atmosphere. Parameter optimization is a critical step in the application of diagnostic models and observations of net ecosystem exchange (NEE) and gross primary productivity (GPP) at eddy covariance tower sites provide useful reference data. Here we examined the sensitivity of optimized parameters in the CFLUX model to 1) the source of the satellite FPAR data (fraction of incident PAR absorbed by the vegetation canopy), 2) the distribution and number of tower sites used in the optimization, and 3) the source of the meteorological data used to drive the model. Results support the scheme of multiple reference sites per Plant Functional Type (PFT), with further breakout by ecoregion in the case of the evergreen needle leaf PFT. EVI generally performed best among the potential sources of FPAR, more so with respect to seasonality than to interannual variation. Differences between tower and distributed meteorological data were large in some cases, but the differences generally did not strongly affect the goodness of fit between simulated and observed NEE. After further optimization exercises, the model will be applied over North America, with special attention to issues of spatial resolution and characterization of disturbance.
Presentation Type: Poster
Poster Session: Carbon Cycle Science
NASA TE Funded Awards Represented:
Turner, David
Assessing the Sensitivity of Net Ecosystem Exchange over North America to Climate and Disturbance with Prognostic and Diagnostic Models