We evaluated an idealized boundary layer (BL) model with simple parameterizations using vertical transport information from community model outputs (NCAR/NCEP Reanalysis and ECMWF Interim Analysis) to estimate regional-scale net CO2 fluxes from 2002 to 2007 at three forest and one grassland flux sites in the United States. The BL modeling approach builds on a mixed-layer model to infer monthly average net CO2 fluxes using high-precision mixing ratio measurements taken on flux towers. We compared BL model net ecosystem exchange (NEE) with estimates from two independent approaches. First, we compared modeled NEE with tower eddy covariance measurements. The second approach (EC-MOD) was a data-driven method that upscaled EC fluxes from towers to regions using MODIS data streams. Comparisons between modeled CO2 and tower NEE fluxes showed that modeled regional CO2 fluxes displayed interannual and intra-annual variations similar to the tower NEE fluxes at the Rannells Prairie and Wind River Forest sites, but model predictions were frequently different from NEE observations at the Harvard Forest and Howland Forest sites. At the Howland Forest site, modeled CO2 fluxes showed a lag in the onset of growing season uptake by 2 months behind that of tower measurements. At the Harvard Forest site, modeled CO2 fluxes agreed with the timing of growing season uptake but underestimated the magnitude of observed NEE seasonal fluctuation. This modeling inconsistency among sites can be partially attributed to the likely misrepresentation of atmospheric transport and/or CO2gradients between ABL and the free troposphere in the idealized BL model. EC-MOD fluxes showed that spatial heterogeneity in land use and cover very likely explained the majority of the data-model inconsistency. We show a site-dependent atmospheric rectifier effect that appears to have had the largest impact on ABL CO2 inversion in the North American Great Plains. We conclude that a systematic BL modeling approach provided new insights when employed in multiyear, cross-site synthesis studies. These results can be used to develop diagnostic upscaling tools, improving our understanding of the seasonal and interannual variability of surface CO2 fluxes.
Journal of Geophysical Research: Biogeosciences
American Geophysical Union (AGU)
Digital Object Identifier (DOI)
Dang, X.R., Lai, C.T., Hollinger, D., Schauer, A., Xiao, J.F., Munger, W., Owensby, C., Ehleringer, J.R. (2011). Combining tower mixing ratio and community model data to estimate regional-scale net ecosystem carbon exchange by boundary layer inversion over 4 flux towers in the U.S.A., Journal of Geophysical Research - Biogeosciences, 116, G03036, https://dx.doi.org/10.1029/2010JG001554.
©2011. American Geophysical Union. All Rights Reserved.