Development of a calibration approach using DNDC and PEST for improving estimates of management impacts on water and nutrient dynamics in an agricultural system

Authors: BHATTARAI, ABHA - The Ohio State University; STEINBECK, GARRETT - The Ohio State University; GRANT, BRIAN - Agriculture And Agri-Food Canada; KALCIC, MARGARET - The Ohio State University; King, Kevin; SMITH, WARD - Agriculture And Agri-Food Canada; XU, NUO - The Ohio State University; DENG, JIA - University Of New Hampshire; KHANAL, SAMI - The Ohio State University

Submitted to: Journal of Environmental Modeling and Software
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 8/5/2022
Publication Date: 8/11/2022
Citation: Bhattarai, A., Steinbeck, G., Grant, B.B., Kalcic, M., King, K.W., Smith, W., Xu, N., Deng, J., Khanal, S. 2022. Development of a calibration approach using DNDC and PEST for improving estimates of management impacts on water and nutrient dynamics in an agricultural system. Journal of Environmental Modeling and Software. 157 Article 105494. https://doi.org/10.1016/j.envsoft.2022.105494.
DOI: https://doi.org/10.1016/j.envsoft.2022.105494

Interpretive Summary: Environmental models are convenient and necessary tools for assessing the response of natural systems under various management scenarios. However, prior to implementing the models and due to the number of parameters required to parameterize the model, calibration is often required to ensure more accurate predictions. Simultaneous calibration of the DeNitrification DeComposition (DNDC) model was shown to improve water and nutrient predictions compared to other calibration approaches, but had no effect on nitrous oxide predictions. Use of the simultaneous calibration approach should provide more confidence in model predictions for researchers, academicians, and policy makers that use the model and results for environmental decisions.

Technical Abstract: Biogeochemical ecosystem models, when combined with measured data, are powerful tools for understanding nutrient dynamics in agroecosystems. We evaluate three approaches, including simultaneous, sequential, and asynchronous, for calibration of 43 parameters of the DeNitrification DeComposition (DNDC) model through inverse modeling using the PEST, an open-source parameter estimation and uncertainty analysis software. During the calibration period, statistical analysis of the calibration strategies indicated better model performance for water and nitrate leaching and crop yield in the simultaneous strategy compared to sequential and separate calibration strategies. Simultaneous calibration reduced the total sum of weighted square residuals by 37% with improvements in the timing of peak water leaching fluxes compared to the default simulation. The timing of major N2O flux events however was not impacted across calibration strategies. Sensitivity and uncertainty analyses provided insights into parameter-observation relationship as well as reliability of model-based predictions.