Computation of nitrate leaching to groundwater, using simple models
Drecht G van , Goossensen FR , Hack-ten Broeke MJD , Jansen EJ , Steenvoorden JHAM
RIVM Report 724901003
Abstract niet beschikbaar
Nine steady state nitrate leaching models have been tested. Predicted annual leaching rates have been compared with measured nitrate leaching of field experiments both statistically and visually (double mass). Five models give satisfactory results for grassland on sandy soil. No sufficient data are available to test the models for grassland on clayey soil. None of the teted models can predict the measured nitrate leaching from silage maize fields on sandy soil satisfactorily. All tested models underestimate the measured nitrate leaching rates from silge maize fields. This can be explained by the high mineralization rates of organic N in these fields that probably have contributed to the nitrate leaching. Scenario studies have been carried out for cut and grazed grassland, silage maize and arable land (crop rotation). In a situation of cut grassland on well drained sandy soil and a N-fertilization level of 400 kg.ha-1.a-1, five models, that gave satisfactory results in the test, show a wide range of computed nitrate leaching rates 953 to 172 kg.ha-1.a-1). However most models compute the same reduction of about 50% of nitrate leaching rates if the N-fertilization level is decreased to 300 kg. ha-1.a-1 and slurry is applied in spring by the method of injection instead of spreading.