Validation of the GLEAMS simulation model for estimating net nitrogen mineralisation and nitrate leaching under cropping in Canterbury, New Zealand

Abstract

Simulation models require testing and calibration prior to their application to regions beyond those involved in their development. This paper reports on the calibration and testing of the groundwater loading effects of agricultural management systems (GLEAMS) model for the simulation of nitrate leaching under cropping in Canterbury. The GLEAMS model was first calibrated using crop and nitrogen leaching data collected from 4 consecutive years (1991–94) of spring-sown cereals following the ploughing of a temporary grass/clover pasture. Nitrate leaching losses were calculated from a combination of measured soil-solution nitrate concentration at 0.6 m depth, estimated drainage, and mineral N from soil cores. These calculated leached-N values were then used to calibrate the GLEAMS model. Parameters controlling denitrification and mineralisation rate in the model needed modification to provide sufficient mineral N for plant growth and nitrate leaching. The calibrated model was then tested against 3 independent validation data sets that were collected over 3 years from an adjacent experimental site, under the same management practices. Predictions from the calibrated GLEAMS model provided close agreement with measured values of mineralisation and leached N for the validation data sets. The amount of leached N averaged 43 kg N/ha.year and varied from 14 to 104 kg N/ha.year. The annual amount of drainage accounted for 97% of the variance in leached N, but the period in arable cropping was poorly correlated with leached N.