Progress 04/01/24 to 03/31/25
Outputs
Target Audience:This research is aimed at the broad community involved in farmland management across the U.S. Corn Belt, including farmers, agricultural cooperatives, private companies, conservation officers, landowners, and non-governmental organizations (e.g., Practical Farmers of Iowa). The goal is to support the identification and adoption of best management practices that inform conservation program implementation. Our work will benefit farming communities by evaluating the profitability of alternative management strategies, while also supporting surrounding communities that rely on water resources from agricultural river basins and watersheds. Changes/Problems:The development of the new SWAT-DNDC model has been delayed due to personnel changes and limited funding within our research group (Avay Risal left our group early this year). We are actively seeking additional resources to support continued model development and to maintain the necessary personnel capacity within the team. What opportunities for training and professional development has the project provided?Postdoctoral researcher Muhammad Rizwan Shahid at Iowa State University co-authored two manuscripts during this project year. He received training in analyzing SWAT model outputs, which enhanced his understanding of soil nutrient cycling and the effects of management practices on crop yield, soil health, and water quality. Postdoctoral researcher Avay Risal at the University of Maryland received training in developing a comprehensive SWAT model for large-scale basins, including the Upper Mississippi River Basin. He also gained experience in processing remote sensing data, such as MODIS and SMAP, for use in model calibration. He served as the first author of a research paper published in 2025. How have the results been disseminated to communities of interest?We have shared our findings with relevant communities through participation in international conferences, including the 2024 AGU Fall Meeting and the 2025 EGU Meeting. In addition, we regularly presented our results at meetings of the USDA LTAR Drainage Working Group - Modeling Subgroup. What do you plan to do during the next reporting period to accomplish the goals?We plan to conduct additional scenario analyses at both the Kelley site and the SFW to evaluate the impacts of adopting cover crops, transitioning from conventional to conservation or no-tillage practices, and shifting between manure and mineral fertilization. Our analyses will also incorporate the "4R" nutrient management principles. The objective is to identify optimal combinations of management practices that deliver the greatest environmental benefits--such as reductions in nitrate leaching and NO emissions--while maintaining or enhancing crop yields to support economic viability. We will continue to develop and test the proposed SWAT-DNDC model at both site and watershed scales. Additionally, we aim to compare the performance of the SWAT-DNDC model with the DayCent-based SWAT model to assess their respective strengths and limitations, informing future model development and refinement.
Impacts
What was accomplished under these goals?
We employed a comparative modeling approach to evaluate the performance of SWAT in simulating watershed-scale tile drainage and N losses (including NO3leaching and N gas fluxes) in a typical Midwestern tile-drained watershed. We compared eight model configurations, combining original and new tile drainage modules, calibrated and non-calibrated tile parameters, and original versus new N modules. Daily streamflow and NO3load data (2001-2018) from three monitoring stations were used for calibration, validation, sensitivity, and uncertainty analyses. Watershed-scale water and N balances were assessed and compared to previous studies within the watershed and at the nearby Kelley experimental site. Our results indicated that the comparative modeling approach helped identify key processes contributing to prediction uncertainty, as well as the additional observational and/or soft data needed to enhance model calibration and accuracy. Model evaluation indicated that all eight model configurations effectively simulated daily and monthly streamflow and NO3loads, though they tended to underestimate peak streamflow and NO3loads. The watershed-scale water balance terms mostly fell within the range reported in previous studies. We found that the new tile module generally enhanced daily streamflow predictions compared to the original module under both conditions of calibrated and non-calibrated tile parameters. For both the new and original tile modules, further calibration of additional tile parameters did not lead to noticeable improvements in daily streamflow. Meanwhile, the new N module consistently outperformed the original N module in simulating daily NO3loads, regardless of the tile module configuration used. Overall, we recommend using the new tile and N modules for simulating streamflow and NO3fluxes in tile-drained watersheds. Parameter sensitivity and uncertainty analyses revealed high uncertainty, particularly for NO3load predictions at all three monitoring stations due to the complexity of N cycling processes. Specifically, we found that denitrification/nitrification and associated N gas fluxes (inducing N2O, NO, and N2) varied significantly across configurations, reflecting fundamental differences between the original and new N modules. The new N module tended to underestimate denitrification while overestimating nitrification. In addition, the N budget differed from values reported at the Kelley site. This highlights that relying solely on outlet NO3load data to capture these processes proved insufficient, emphasizing the need for additional observational data (for example, N2O fluxes) to reduce uncertainty in modeling denitrification/nitrification and associated N2O fluxes at the watershed scale. When additional measurements are impractical, incorporating soft data is a valuable alternative. Soft data can support proper process representation, significantly improving model calibration and validation, and helping to better constrain soil N cycling and refine N budget estimates. Sources may include peer-reviewed literature, technical reports, theses, and field surveys. Utilizing such data is essential for developing accurate nutrient budgets and ensuring realistic modeling of biogeochemical processes. Tile drainage systems are extensively implemented across the U.S. Midwest to enhance crop productivity in poorly drained soils; however, they also pose environmental challenges by significantly altering nitrogen fluxes within agricultural landscapes. In response, sustainable intensification strategies aim to increase agricultural output while reducing environmental impacts, often through improved management practices such as cover cropping. Effectively evaluating the trade-offs and synergies of agricultural management practices demands advanced modeling tools capable of representing coupled biogeochemical and hydrological processes across diverse spatial and temporal scales. This study presents the first application of an enhanced version of the Soil and Water Assessment Tool (SWAT), integrated with Century/DayCent-based biogeochemical modules, to simulate both nitrate (NO3) loss and nitrous oxide (N2O) fluxes in a tile-drained corn-soybean system. The model was applied to long-term field data (2004-2010) from an Iowa site with two treatments: with and without winter rye cover crops. With careful calibration, the model closely matched observed crop yields, tile drainage, NO3losses, and N2O fluxes under both winter rye cover crop and no-cover scenarios. It simulated a ~42% reduction in NO3leaching due to cover cropping, consistent with observed values, while the impact on N2O fluxes varied across years and conditions (percentage change ranging -24 to 22%). Overall, the enhanced SWAT model shows strong potential for evaluating nitrogen dynamics and conservation practices in tile-drained agroecosystems, though further refinement is needed to improve simulation of daily N2O fluxes and complex cover crop interactions. This study demonstrated that the enhanced SWAT-N2O model, which integrates Century/DayCent-based soil C and N cycling, is capable of simulating key N loss pathways--such as NO3leaching and N2O fluxes--in tile-drained agroecosystems of the U.S. Midwest, as well as the effects of cover crops on these losses. By integrating EPIC-derived humus pool partitioning, enhancing nutrient dynamics in the litter layer, and adjusting key parameters related to crop growth, residue quality, and drainage, the model performed well in reproducing observed tile drainage flow, crop yields, and NO3losses under both cover crop and no-cover crop scenarios. The model effectively captured the impact of winter rye cover crops, simulating an average 42% reduction in NO3losses, closely matching observed values. However, consistent with previous studies, the model showed limitations in reproducing daily N2O flux peaks, likely due to simplified representations of soil hydrology and limited vertical resolution. N2O responses to cover crops varied across years and crop types, highlighting the complex interplay between microbial N cycling, management practices, and environmental conditions. Overall, when properly parameterized, the enhanced SWAT-N2O model offers a valuable tool for assessing the environmental benefits of cover crops in tile-drained agricultural landscapes.
Publications
- Type:
Peer Reviewed Journal Articles
Status:
Published
Year Published:
2025
Citation:
Risal, A., Karki, R. and Qi, J., 2025. Integrating Remote Sensing with Watershed-Scale Ecohydrological Modeling in Diverse Forest and Agricultural Landscapes of the Upper Mississippi River Basin. Environmental Modelling & Software, p.106588.
- Type:
Peer Reviewed Journal Articles
Status:
Published
Year Published:
2025
Citation:
Karki, R., Qi, J., Zhang, X. and Srivastava, P., 2025. Evaluating SWAT-3PG simulation of hydrologic and water quality processes in a forested watershed: A case study in the St. Croix River Basin. Journal of Hydrology, 648, p.132393.
- Type:
Other Journal Articles
Status:
Under Review
Year Published:
2025
Citation:
Junyu Qi, Robert Malone, Kang Liang, Kevin Cole, Bryan Emmett, Daniel Moriasi, Muhammad Rizwan Shahid, Michael Castellano.Comparative Modeling of Nitrogen Losses in a Tile-Drained Watershed Using SWAT Model: Uncertainty and Calibration Considerations. Frontiers in Environmental Science (under review)
- Type:
Conference Papers and Presentations
Status:
Published
Year Published:
2024
Citation:
QI, J., Shahid, M.R., Malone, R.W., Liang, K., Emmett, B., Cole, K., Moriasi, D. and Castellano, M.J., 2024, December. Assessing the Impacts of Agricultural Management Practices on N2O Emissions, NO3 Leaching, and Crop Yield in a Typical Tile-Drained Watershed in the Midwest. In AGU Fall Meeting Abstracts (Vol. 2024, No. 82, pp. H51J-082).
- Type:
Conference Papers and Presentations
Status:
Published
Year Published:
2024
Citation:
Shahid, M.R., Emmett, B., Malone, R.W., Castellano, M.J. and Qi, J., 2024, December. Mitigating N2O Emissions and NO3 Leaching for Sustainable Corn-Soybean System in Midwest. In AGU Fall Meeting Abstracts (Vol. 2024, No. 818, pp. H51J-0818).
|
Progress 04/01/23 to 03/31/24
Outputs
Target Audience:
Nothing Reported
Changes/Problems:Iowa State University initially hired a graduate student for this project, but the student subsequently dropped out. We then re-advertised the position for both students and post-docs. We have now hired a highly qualified post-doc, Rizwan Shahid, who will begin in June 2024. As a result, progress on the ISU side experienced some delays. However, we anticipate significant progress in the next project year. What opportunities for training and professional development has the project provided?Iowa State University hired a graduate student for this project and the student dropped out. At that time, we re-advertised the position for students and post-docs. We hired a highly qualified post-doc, Rizwan Shahid, and he will begin June 2024. Dr. Shahid has extensive experience calibrating and applying ecosystem process models and he will be able to make up for lost time on modeling objectives given our original plan was to train a student. How have the results been disseminated to communities of interest?
Nothing Reported
What do you plan to do during the next reporting period to accomplish the goals?We will conduct further scenario analysis to assess the impacts of adding cover crops, transitioning from conventional tillage to conservation/no-tillage, and switching between manure and mineral fertilization. Additionally, we will consider "4R" practices in our scenario analysis. Our goal is to identify the best combination of management practices that provide the most significant environmental benefits (reducing nitrate leaching and N2O emissions) while maintaining or increasing crop yields for economic benefits. We will develop and test the proposed SWAT-DNDC model at both site and watershed scales. Furthermore, we plan to compare the SWAT-DNDC model with the DayCent-based SWAT model to evaluate their strengths and weaknesses, facilitating potential model development.
Impacts
What was accomplished under these goals?
The National Laboratory for Agriculture and the Environment collected data on nitrous oxide gas emission and nitrate loss in tile drainage from corn-soybean rotations under long-term conservation practices. In 2023 the field was in the soybean phase of the rotation and allowed evaluation of multiple winter cover strategies including rye cover crop, harvested rye cover crop and winter camelina-soybean double cropping systems. In 2024, field data collection continued in the corn phase of the rotation allowing evaluation of 4R nitrogen management on gas emissions. Data will contribute to the long-term dataset used to calibrate and validate the SWAT-DNDC model. Field data collection also included high temporal resolution nitrous oxide emission data from automated chamber systems that can be used to evaluate model capacity to estimate emissions during peak emission events. The University of Maryland developed an enhanced version of the SWAT model incorporating the latest DayCent-based N2O module. Additionally, we improved the nitrate loss simulation with tile drainage to better reflect the impacts of water table dynamics. The enhanced DayCent-based model was applied to two different SWAT projects. For the South Fork Watershed, we created a SWAT project using detailed data on crop rotation, manure application, and tillage practices collected from the watershed. We used the SSURGO soil map, a DEM with 10 m resolution, and a field farm-based land use map. The model was calibrated using data from five monitoring stations with daily streamflow and nitrate loading. We also developed a SWAT project with a single HRU for the Kelley site, incorporating detailed local weather data, crop rotation, fertilization, and the timing and amount of applications. The model was calibrated using monitored tile drainage water, nitrate loading, crop yield, and N2O emissions. The SWAT model includes two tile drainage modules: one simplified and one more physically based. After autocalibration with daily stream data, we determined that the simplified module better simulated drained water. Model calibration and validation indicated satisfactory performance for both drainage water discharge and nitrate loading at both site and watershed scales. N2O emissions and crop yields at the Kelley site were also simulated satisfactorily.
Publications
- Type:
Journal Articles
Status:
Published
Year Published:
2024
Citation:
Luo, X., Risal, A., Qi, J., Lee, S., Zhang, X., Alfieri, J.G. and McCarty, G.W., 2024. Modeling lateral carbon fluxes for
agroecosystems in the Mid-Atlantic region: Control factors and importance for carbon budget. Science of The Total
Environment, 912, p.169128.
- Type:
Journal Articles
Status:
Published
Year Published:
2023
Citation:
Tijjani, S.B., Qi, J., Giri, S. and Lathrop, R., 2023. Modeling Land Use and Management Practices Impacts on Soil Organic
Carbon Loss in an Agricultural Watershed in the Mid-Atlantic Region. Water, 15(20), p.3534
|