Temporal Variability of Nitrous Oxide from Fertilized Croplands: Hot Moment Analysis

Nitrous oxide (N2O) emissions were monitored using the micrometeorological eddy covariance technique from manure-fertilized cropland on a large dairy farm in New York State in 2006 to 2009. Nitrous oxide emissions demonstrated episodic behavior with intermittent short-duration peak fluxes up to 39.7 mg N2O-N m(-2) d(-1), whereas most of background fluxes during the annual agricultural cycle were below 6.5 mg N2O-N m(-2) d(-1). This paper discusses temporal variability of measured N2O emissions using a "hot moment" approach. To identify and quantify peak events as potential hot moments and to determine whether or not they could be treated statistically as outliers, N2O daily fluxes were analyzed by the box plot method using multiple thresholds. Peak events exceeding outlier thresholds contributed up to 51% of cumulative annual N2O emissions, although they represented <7% of the total observation time. Individual N2O peaks were also categorized by their duration, as single day spikes and multiday events. The highest contributing instances were multiday N2O peaks during summer precipitation and early spring thaw, largely enhanced by manure fertilization. These high-intensity emission events demonstrated repetitive seasonal responses to a combination of environmental factors and were therefore identified as hot moments. Abrupt rises in both temperature and soil moisture appeared to trigger major hot moments, whereas the availability of manure N controlled their magnitude. In the absence of strong correlations between time-series of individual environmental factors and N2O emissions, the hot moment approach can be a promising tool for the integrated analysis of most significant N2O events in cultivated fields receiving manure applications.