Overview of Canada DNDC model development and studies pertaining to crop water use and nutrient loss
Citation
Smith, W.N., Grant, B., VanderZaag, A., Qi, Z., Drury, C., Desjardins, R., Pattey, E., Dutta, B., Guest, G. 2016. Overview of Canada DNDC model development and studies pertaining to crop water use and nutrient loss. China-Canada Water Resource Workshop, Ottawa, ON, Canada, November 15th, 2016.
Résumé en langage clair
Process-based agricultural models are useful tools for simulating the interrelationships between plants, soil, and the atmosphere by deterministic means and are excellent tools for estimating crop water use efficiency and the overall system water balance. A Canadian version of the Denitrification Decomposition model (DNDC) has been under development since 2011 and it is now being used by several research groups across Canada to estimate crop yields and GHG emissions. However, it has not yet to be tested for estimating drainage and nutrient loss for agroecosystems in Canada and the model contains a simple cascade water flow routine. The Root Zone Water Quality Model (RZWQM2) includes very detailed algorithms for estimating water and nutrient drainage and runoff from cropping systems. The objectives of this study were to i) improve the Canada DNDC model by including a heterogeneous soil profile and algorithms for root density functions ii) test the Canada DNDC model for simulating water flow and nitrogen loss to tile drains at two research sites (Harrow, Ontario and Iowa) which both have a corn-soybean rotation, with and without cover crops, iii) compare to the more detailed RZWQM2 model, and iv) recommend improvements for future research. The inclusion of a heterogeneous soil profile and root density functions greatly improved the estimation of soil water contents. It was interesting to find that the estimation of nitrogen and water loss to tiles drains was very well simulated with and without the improvements to soil water content. We compared the Canada DNDC model performance to the detailed process-based model RZWQM2 and found that Canada DNDC performed in a very similar manner. Canada DNDC, however, is not capable of simulating water table depth, and flow to drains at different drainage spacing, nor controlled drainage and thus we recommend the inclusion of a drainage sub-model.
Résumé
Process-based agricultural models are useful tools for simulating the interrelationships between plants, soil, and the atmosphere by deterministic means and are excellent tools for estimating crop water use efficiency and the overall system water balance. A Canadian version of the Denitrification Decomposition model (DNDC) has been under development since 2011 and it is now being used by several research groups across Canada to estimate crop yields and GHG emissions. However, it has not yet to be tested for estimating drainage and nutrient loss for agroecosystems in Canada and the model contains a simple cascade water flow routine. The Root Zone Water Quality Model (RZWQM2) includes very detailed algorithms for estimating water and nutrient drainage and runoff from cropping systems. The objectives of this study were to i) improve the Canada DNDC model by including a heterogeneous soil profile and algorithms for root density functions ii) test the Canada DNDC model for simulating water flow and nitrogen loss to tile drains at two research sites (Harrow, Ontario and Iowa) which both have a corn-soybean rotation, with and without cover crops, iii) compare to the more detailed RZWQM2 model, and iv) recommend improvements for future research. The inclusion of a heterogeneous soil profile and root density functions greatly improved the estimation of soil water contents. It was interesting to find that the estimation of nitrogen and water loss to tiles drains was very well simulated with and without the improvements to soil water content. We compared the Canada DNDC model performance to the detailed process-based model RZWQM2 and found that Canada DNDC performed in a very similar manner. Canada DNDC, however, is not capable of simulating water table depth, and flow to drains at different drainage spacing, nor controlled drainage and thus we recommend the inclusion of a drainage sub-model.