Simulating agroecosystem soil inorganic nitrogen dynamics under long-term management with an improved SWAT-C model

Despite the extensive application of the Soil and Water Assessment Tool (SWAT) for water quality modeling, its ability to simulate soil inorganic nitrogen (SIN) dynamics in agricultural landscapes has not been directly verified. Here, we improved and evaluated the SWAT-Carbon (SWAT-C) model for simulating long-term (1984-2020) dynamics of SIN for 40 cropping system treatments in the U.S. Midwest. We added one new nitrification and two new denitrification algorithms to the default SWAT version, resulting in six combinations of nitrification and denitrification options with varying performance in simulating SIN. The combination of the existing nitrification method in SWAT and the second newly added denitrification method performed the best, achieving R, NSE, PBIAS, and RMSE of 0.63, 0.29, -4.7 %, and 16.0 kg N ha-1, respectively. This represents a significant improvement compared to the existing methods. In general, the revised SWAT-C model's performance was comparable to or better than other agroecosystem models tested in previous studies for assessing the availability of SIN for plant growth in different cropping systems. Sensitivity analysis showed that parameters controlling soil organic matter decomposition, nitrification, and denitrification were most sensitive for SIN simulation. Using SWAT-C for improved prediction of plant-available SIN is expected to better inform agroecosystem management decisions to ensure crop productivity while minimizing the negative environmental impacts caused by fertilizer application.

Automated summary

This study improved and evaluated the SWAT-C model for simulating long-term dynamics of soil inorganic nitrogen (SIN) in different cropping systems. By adding new nitrification and denitrification algorithms, the model achieved better performance in SIN simulation. The revised SWAT-C model's performance was comparable to or better than other agroecosystem models tested in previous studies. Sensitivity analysis showed that parameters controlling soil organic matter decomposition, nitrification, and denitrification were most important for SIN simulation. The improved prediction of plant-available SIN using SWAT-C can inform agroecosystem management decisions to enhance crop productivity and minimize negative environmental impacts.