Investigation of Carbon Flux Modeling using Multi-criteria Calibration Techniques

Estimating sources and sinks of the global carbon budget has become one of the most challenging tasks that scientists currently face. The amount of CO2 produced by fossil fuel burning and deforestation appears to exceed the amount that is being taken up by the atmosphere and oceans. It is believed that global primary production or photosynthesis fixes as much as 20 times the carbon released by fossil fuel emissions. Several theories exist to account for this "missing sink", one of which is the accumulation of carbon in northern and mid-latitude forests. Improved estimates of carbon uptake by various vegetative surfaces are crucial to providing researchers with better quantification of this sink. This research focuses on the use of recently developed multi-criteria optimization techniques to investigate carbon flux modeling at the earth's surface. An advanced land-surface model, BATS2, is used to simulate carbon uptake and respiration, along with traditional latent and sensible heat fluxes, at various land surfaces. The BATS2 model has been revised from the previous version to include a foliage growth model, and to link stomatal conductance and carbon assimilation. Initial point studies reveal the model performs well in mature vegetated surfaces (evergreen broadleaf), but has trouble simulating carbon fluxes for perennial crop surfaces. Investigation and modification of the model parameterization is undertaken to improve performance. Incorporation of multi-criteria optimization in these processes also provides better estimates of model parameters and improves simulations of all surface fluxes.