Modeling Nitrogen Transport and Transformations in Soils

This is the authors' abstract. We don't add key points for this paper.

We evaluated the nitrogen model described in Part 1 (Selim and Iskandar 1981) using experimental data from a greenhouse lysimeter study for two soils, Windsor sandy loam and Charlton silt loam. Secondary treated waste water was applied to each soil at the rate of 3.8 centimeters twice weekly for 25 weeks. Furthermore, 15N-enriched NH4+-N was applied, at the beginning of the experiment, in one waste water application. A mixture of grasses was grown on each lysimeter and was harvested every 2 to 4 weeks. Solution samples were collected and analyzed for N, and the soil-water pressure head was monitored frequently at different soil depths. Model predictions agreed well with pressure head data with depth and time, as well as gravimetrically determined soilwater content with depth for the two soils. Model predictions of NO3-–N concentration versus time were adequate except for intermediate soil depths (30 and 45 cm). In contrast, NH4+-N predictions were consistently higher than measured data. Reasonable agreement was obtained between calculated and experimental N uptake results, but after 60 days the model overestimated the N uptake by as much as 15 percent. Results from the 15N data were utilized to provide estimates for the nitrification rate coefficients. The rate coefficients obtained were within two and four times the values used to obtain best model predictions for Charlton and Windsor soil, respectively.