/AnMtgsAbsts2009.52670 Accuracy and Precision Analysis of Chamber-Based Nitrous Oxide Gas Flux Estimates.

Wednesday, November 4, 2009: 3:30 PM
Convention Center, Room 411, Fourth Floor

Rodney Venterea, Kurt Spokas and John Baker, Soil and Water Mgmt Research Unit, USDA-ARS, St. Paul, MN
Abstract:
Chamber-based estimates of soil-to-atmosphere nitrous oxide (N2O) gas flux tend to underestimate actual emission rates due to inherently non-linear time series data.  In theory, this limitation can be minimized by adjusting measurement conditions to reduce non-linearity and/or by using flux-calculation (FC) schemes that account for the so-called “chamber effect”.  The current study utilizes gas transport theory and stochastic analysis to evaluate accuracy and precision of N2O flux determinations under specific soil and chamber conditions.  The analysis demonstrates that measures taken to increase the absolute accuracy of flux estimates, including shorter deployment times, larger chamber heights, and non-linear FC schemes, will also increase the variance in flux estimates to an extent that depends on errors associated with sampling techniques and analytical instrument performance.  These effects, in the absence of any actual variation in fluxes, can generate coefficients of variation ranging from 3% to 70% depending on measurement conditions.  It is also shown that non-linear FC schemes are prone to generating positively skewed distributions. These effects decrease confidence in N2O flux estimates and inhibit the detection of differences arising from experimental factors.  In general, a linear FC scheme will be more likely to detect relative differences in fluxes, although less accurate in absolute terms than non-linear schemes.  The techniques described here have been codified into an accessible spreadsheet-based tool for evaluating accuracy and precision trade-offs under specific measurement conditions.