Eddy covariance (EC) measurements of CO 2 are widely used to assess the CO 2 budget of ecosystems. However, important uncertainties exist about how to obtain flux values that are representative for the regional scale from local-scale EC measurements. This paper addresses the issues of nocturnal CO 2 advection in non-flat topography, and the representativeness of the local clover-rich grass surface for the larger region, at the end of the growing season. Local EC measurements were compared with the fluxes obtained from a boundary-layer budget method (BLB) which employed tethered balloon measured vertical profiles of CO 2 concentration in the lowest few hundred meters of the atmosphere. It was found that the diurnal change in CO 2 concentration was restricted to the lowest ≈450 m of the atmosphere.