"This approach is very different from traditional ecosystem models, which are built upon a lot of complex, underlying ecosystem processes," Xiao explains. "Our method is a data-driven approach, meaning it is an empirical, statistical model using vast amounts of data being gathered year-round on an hourly basis by the network of flux towers. This approach can lead to model parameters that are more representative of the full spectrum of vegetation and climate conditions, and to more robust estimates of ecosystem carbon fluxes over broad regions."
The network of more than 100 towers dotted around the North American continent measures the covariance (how much two variables change together) of vertical wind velocity and fluctuating carbon dioxide (CO2) concentration around the tower's one-square-kilometer footprint. From this data, the exchange of CO2 between the ecosystem and atmosphere can be accurately calculated. Xiao's modeling method essentially allows the numbers from the individual tower sites to be crunched together into the bigger picture.
The data-driven methodology pioneered by Xiao, and now being funded by NSF, was novel enough that it took several years to convince the ecosystem modeling community at large that the technique could accurately be applied continent-wide and eventually, Xiao anticipates, globally as well. Xiao and Ollinger also propose to examine the impacts of disturbances on carbon cycling and to quantify the uncertainty of carbon fluxes using a forest ecosystem model developed at UNH and the relatively new technique of model-data fusion or data assimilation.
The creation of NSF's Macrosystems Biology program follows the agency's ambitious National Ecological Observatory Network (NEON), which is in the process of being put together and when co
|Contact: David Sims|
University of New Hampshire