Search

The terrestrial biosphere is an important pool of carbon, with its size governed by the opposing processes of CO2 uptake through photosynthesis and release through respiration. It is therefore critically important to understand and reliably and accurately model these processes and predict changes in carbon exchange in response to key drivers. Pasture-based livestock production is particularly important for the New Zealand’s economy but it is also a main contributor to NZ’s greenhouse gas budget. My Ph.D. work used half-hourly eddy-covariance (EC) data, previously collected over 2 consecutive years from a grazed pasture in the Waikato region. The main aims of this study were to assess whether there was any bias in gap-filled eddy covariance measurements, to assess whether incomplete capture of cow respiration during grazing events could have led to biased observations, and to quantify the resulting difference on the net carbon budget of the farm. I approached the work by developing a new process-based model, CenW_HH, running at a half-hourly time step, to predict the energy and CO2 exchange of grazed pastures. I implemented and evaluated different photosynthesis models and upscaling schemes and modelled the energy budget separately for the canopy, litter layer, and the soil. CenW_HH was then parameterised and validated with the available EC measurements. The paddocks surrounding the EC tower were rotationally grazed, which caused heterogeneities in respiratory pulses when grazing events were in the flux footprint and subsequent vegetation cover on the different paddocks. To deal with that heterogeneity, the model was run independently for each individual paddock and a footprint model was used to estimate resultant net fluxes at the EC tower. Modelled fluxes agreed well with half-hourly observed fluxes as seen by model efficiencies of 0.81 for net ecosystem productivity, 0.75 for gross primary production, 0.70 for ecosystem respiration, 0.87 for latent heat flux, 0.76