Open access publications in the SLU publication database

Abstract

A simple model was developed in order to investigate nitrogen retention by denitrification in pnds under the influence of different pond dimensions. Model input data, concerning water flow rate and nitrification concentration in the water, were taken from the runoff records from two small basins on arable land in southern Sweden. A water voume balance and a nitrate-nitrogen balace were determine in discrete time increments. The denitrification was assumed to take place only in the bottom sediments, depending on the nitrate concentration of the pond water. Four different calculations of denitrification was preformed, representing two diffrent sediment types with two seasonsl variations respectively, in addition, the calculations were preformed in three different ways, in three versions. In version I the nitrate concentration was assumed to be homogeneous in the whole of the pond, while in version II and III, the pond was assumed to be homogeneous in the whole of the pond, while in version II and III, the ond was divided into eight sections, in order to get more diversified concetrations in the pnd. In version III, a bottom outlet was added with the intention of getting a smoother outflow. Calculations with the version II and III were mainly performed with the purpose to compare them and combinations of them with each other, both in large single ponds and in pond systems/series with ponds of different sizes. The difference between the nitrate concentration at the inlet and the outlet depended in the current flow rate. At low flow periods, there was a conciderable difference between inlet- and outlet concentrations, while during high flow cnditions this difference decreased and, at times, was negligible. The yearly amount of denitrified nitrogen increased with increasing nitrogen input, corresponding to a high average nitrate concentration in the pond. At the same time, the relative reduction of transported nitrogen decreased, due to the connection between nitrogen transport and water flow. At high flow situations, the time during which the water was exposed to the sediment was reduced, leadeing to small reduction in nitrate concentrations in the pond. When compareing ponds with the same depth, a good agreement was found between the early relative rdeuction of the nitrogen transport and the average water retention time. The amount of denitrified nitrogen increased with increasing pond size. It was also found that the pond area has greater inpact than the pond depth. In comparisons btween version II and III, the total differences in the amout of denitrified nitrogen between different pond system were small. The denitrification was lower in the first pond with bottom outlet. As its water vlume decreased faster during low flow periods than in the pond without bottom outlet, its nitrate supply for denitrication also decreased faster. However, the smoother outflow from the pond with bottom outlet gave better conditions for denitrification to the subsequent ponds. The good agreement between average water retention time and relative reduction seems to be a reasonable starting point in using this type of model for pond dimensioning. However, special care must be taken when wirking with areas where the runoff pattern and nitrate concentrations differ considerably from those used as imput data. Comparing simulations were made, in which real measurements in a pond in southern Sweden were compared with calculated nitrogen retentions. The measured retention wasthen about 50 % higher than estimated. As the model calculations only deal with denitrification, it is also very likely that the total retention in sunderestimated. The denitrification itself is propably underestimated as well, as diffusion of nitrate into the sediment should increase in situations with high water flow. The strong dependence between flow rates and relative reduction of the nitrate transport seems, however, to agree with some published measurements.