Noble, Alasdair

A new approach to calculating flux measurements from chamber based soil greenhouse gas data.

Alasdair Noble¹, Rod Venterea², Asger Pedersen³, Soren Petersen⁴, Tim Parkin⁵, and Cecile de Klein⁶

1. Agresearch, Lincoln

2. USDA - ARS, St. Paul, USA

3. Department of Clinical Medicine , Aarhus University, Denmark

4. Department of Agroecology - Soil Fertility, Aarhus University, Denmark

5. USDA - ARS, Ames, USA

6. Agresearch, Invermay

Measuring the flux of a greenhouse gas being emitted from soil is important for assessing National totals for Climate Agreement protocols. In a “good” experiment a chamber is placed on the soil and four measurements are taken over time. Diffusion theory suggests that the concentration of the gas of interest (Nitrous Oxide in this case) will increase non-linearly over time. The variable of interest is the instantaneous gas emission at time 0, ie the slope of the line at that time. Various linear and non-linear models are fitted and the best fitting model is generally accepted to predict the slope. With so few data points R squared values of over 99% are common for all of the models so the decision process is not clear cut. We propose a model averaging approach which improves the stability of the estimates.