| dc.description.abstract | The goal of this research was to evaluate the ability of two agricultural system models, the Root Zone Water Quality Model (RZWQM) and the Coupled Model (CoupModel), to simulate temporal nitrate mass loading below the root zone. To tackle this evaluation effort the following research objectives were developed: (1) determine the sensitivity of key model output responses for a selected agricultural system model to the variability of input parameters over different vertical-spatial and temporal domains; (2) compare and elucidate the ability of two agricultural system models to simulate water flux and nitrate loading at the plot scale; (3) explore the capability of an agricultural system model that is fully calibrated at one location to simulate the water flux and nitrate loading at another location with similar soil and cropping characteristics; and (4) use a calibrated and validated agricultural system model to predict groundwater recharge and nitrate loading resulting from the implementation of a best management practice (BMP) established on a parcel of land where groundwater impacts due to nitrogen application have been observed. Two study sites where elevated nitrate concentrations in groundwater have been observed were used in this research. The primary study site was the shallow unconfined Abbotsford Aquifer, located in Lower Fraser Valley, British Columbia, where elevated groundwater nitrate concentration is attributed to excess nitrogen inputs compared with the nitrogen demand of the red raspberry crop. The secondary study site was the Thornton Well Field in southwestern Ontario where a legacy of agricultural activities in the area has resulted in an increased groundwater nitrate concentration. In summary, global sensitivity analysis not only identified the most influential parameters of the model that required calibration but also provided a useful guide to define the timing and vertical-location of the observation data that is most effective to use as the calibration target, and design appropriate experiments for collecting such data. Both selected models, RZWQM and CoupModel, were reliable for prediction of nitrate loading time series below the raspberry root zone; however, the CoupModel performed better than the RZWQM due to its flexibility for modifying growth parameters when perennial crops are simulated. The calibrated CoupModel was applicable for simulating nitrate flux below the raspberry root zone in a nearby farm within the Abbotsford region except for the years when organic fertilizer was applied. While the results of this transportability effort are promising, additional validation at similar fields under different management practices is encouraged. Also, development of models that capture the effects of raspberry inter-rows cropping system on nitrate and water flux below the raspberry root zone is essential. At the Thornton Well Field, the BMPs were effective in reducing nitrate load into the groundwater; however, various time frames are needed to observe significant response to the BMPs at different farmlands. | en |
