The disjunct eddy covariance (DEC) method is a recent development for trace gas flux measurements. In this method flux is calculated using only a subset of the continuous time-series used in the conventional, or continuous, eddy covariance (CEC) method. Each sample is taken very quickly (e.g. in 0.1 s.) but there is longer interval (e.g. 20 s.) between samples. This makes eddy covariance flux measurements of trace gases possible using relatively slow chemical analysers.

This far the DEC has been applied only to surface based measurements. We are developing an airplane based disjunct eddy sampling (DES) system trace gas flux measurements. As a part of the development we need to specify the measurement strategy best suited for special conditions of airborne measurements.

We have studied various sources of uncertainty by data simulations in order to find specifications for an airborne DEC system. The sources of uncertainty and underestimation of fluxes measured by DEC include sub-sampling of the time-series, sample carry-over, sampler displacement from the anemometer and non-instantaneous samples. We used high frequency data from surface-, ship- and airborne turbulence experiments to simulate measurements by disjunct eddy covariance method and other methods. For example, we compared the DEC fluxes with CEC fluxes to specify the number of samples needed for a specified accuracy of the calculated flux.

The consequences of the simulation results on the airborne DEC system will be discussed. Various possible designs for an airborne DES will be presented with their respective advantages.