Comparing measured and Expert-N predicted N2O emissions from conventional till and no till corn treatments

Agricultural soils are a major source of the greenhouse gas nitrous oxide (N2O). Nitrous oxide emission models can be used to predict the effectiveness of N2O mitigation strategies; however, these models require rigorous testing before they can be used with confidence. Expert-N, a modular process based N2O emission model, was tested to determine its ability at predicting nitrogen (N) cycling in the soil–plant–atmosphere system under Canadian agroclimatic conditions. Ancillary data and N2O emissions were collected/measured from a corn cultivated clay-loam soil that was under different tillage and red clover treatments. The treatments were conventional till (CT) with and without red clover (rc) underseeded in the previous year's wheat crop (CT-Crc and CT-C, respectively), and no till (NT) with and without red clover underseeded in the previous year's wheat crop (NT-Crc and NT-C, respectively). Expert-N provided good estimates of N2O emissions, and predictions correlated well (positive) with the measured emissions (r 2∼ 0.55–0.83). There was no statistically significant difference between measured and predicted daily emissions. The predicted emissions, integrated over the growing season (25 May–4 October, 1995), were 0.56, 0.57, 0.62, and 0.62 kg N2O-N ha−1 for CT-C, CT-Crc, NT-C, and NT-Crc, respectively. The measured emissions over the same period were 1.29, 1.07, 0.96, and 1.04 kg N2O-N ha−1 for CT-C, CT-Crc, NT-C, and NT-Crc, respectively. The modelled emissions underestimated the integrated measured emissions by ∼ 35–55%; however, the integrated measured emissions had an estimated uncertainty of ±35%. The model provided good predictions of the soil temperatures, moisture contents, and soil nitrate levels with no significant difference from the measured data. Correlations between modelled and measured values for these soil properties in the first 30 cm soil layer were positive and high with r 2 0.71–0.93.