Improved prediction of farm nitrous oxide emission through an understanding of the interaction among climate extremes, soil nitrogen dynamics and irrigation water.

Reducing nitrous oxide (NO) emissions from agriculture soils is crucial, as it accounts for 5.6-6.8% of global anthropogenic emissions. This study aims to understand the interaction among climate, soil nitrogen (N) and applied N on NO emissions from the irrigated cotton farming system and its implications on farm economics. We conducted simulations for 116 years (1900-2015) and assessed the effect of different N-fertiliser application rates, initial soil nitrate (NO) N levels and rainfall conditions on NO emissions, NO emission factors (EFs) and financial returns (with and without NO costs). Results showed the following. 1) The proportional impact of higher N fertiliser rates on soil NO emissions was greater when initial soil N level was lower (5 mg NO kg) than higher (35 mg NO kg). However, the volume of impact was greater under higher initial soil N levels. 2) The relationship between N fertiliser rates and the EFs (range 0.03-7.2%) was not linear but bell-shaped. 3) Fertiliser N requirements increased with rainfall and decreased with initial soil N. Accordingly, the cotton returns for the driest rainfall condition (<10th percentile) were maximum at 300, 250 and 150 kg N ha for initial soil N of 5, 20 and 35 mg NO kg. For the wettest rainfall condition (>90th percentile), these rates were 50 kg ha higher across the initial soil N conditions. Any additional application of N-fertiliser above these rates was counterproductive. 4) Inclusion of NO cost into farm economics reduced the annual returns by up to $39 ha, but the optimal fertiliser application rates remain the same. 5) Optimising N fertiliser rates to soil N and rainfall conditions increased the annual returns by up to $303 ha, with a further increase of $15 ha from fertiliser use efficiency when the Australian Government incentives under the $2.55 billion dollar Emission Reduction Fund program was considered. These findings suggest that N-fertiliser application rates and NO emission mitigation strategies need further refinements specific to prevailing soil and climate variabilities.