Premium
Effect of Sampling Frequency on Estimates of Cumulative Nitrous Oxide Emissions
Author(s) -
Parkin Timothy B.
Publication year - 2008
Publication title -
journal of environmental quality
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.888
H-Index - 171
eISSN - 1537-2537
pISSN - 0047-2425
DOI - 10.2134/jeq2007.0333
Subject(s) - flux (metallurgy) , sampling (signal processing) , confidence interval , environmental science , nitrous oxide , standard deviation , atmospheric sciences , statistics , jackknife resampling , soil science , mathematics , zoology , hydrology (agriculture) , chemistry , physics , geology , organic chemistry , estimator , detector , optics , biology , geotechnical engineering
It is generally recognized that soil N 2 O emissions can exhibit pronounced day‐to‐day variations; however, measurements of soil N 2 O flux with soil chambers typically are done only at discrete points in time. This study evaluated the impact of sampling frequency on the precision of cumulative N 2 O flux estimates calculated from field measurements. Automated chambers were deployed in a corn/soybean field and used to measure soil N 2 O fluxes every 6 h from 25 Feb. 2006 through 11 Oct. 2006. The chambers were located in two positions relative to the fertilizer bands—directly over a band or between fertilizer bands. Sampling frequency effects on cumulative N 2 O–N flux estimation were assessed using a jackknife technique where populations of N 2 O fluxes were constructed from the average daily fluxes measured in each chamber. These test populations were generated by selecting measured flux values at regular time intervals ranging from 1 to 21 d. It was observed that as sampling interval increased from 7 to 21 d, variances associated with cumulative flux estimates increased. At relatively frequent sampling intensities (i.e., once every 3 d) N 2 O–N flux estimates were within ±10% of the expected value at both sampling positions. As the time interval between sampling was increased, the deviation in estimated cumulative N 2 O flux increased, such that sampling once every 21 d yielded estimates within +60% and −40% of the actual cumulative N 2 O flux. The variance of potential fluxes associated with the between‐band positions was less than the over‐band position, indicating that the underlying temporal variability impacts the efficacy of a given sampling protocol.