CO2 efflux from agricultural soils in Eastern Germany -: comparison of a closed chamber system with eddy covariance measurements

In order to quantify the effects of temperature and soil water content on soil respiration, during June and July 2002 CO2 soil efflux was measured with a closed chamber (non-steady state, flow through) system in the field. The amount of CO2 emission was highly dependent on the land-use in the observation area, which consisted of meadow soil and brownfield. The CO2 emission from the brownfield ranged from 0.9 to 5.5 mu mol CO2 m(2)s(-1), and that for meadow soil from 1.1 to 12.6 mu mol CO2 m(-2) s(-1). Soil respiration, as a function of soil temperature (T-soil), relative soil water content (RSWC), soil PH, and the soil carbon/nitrogen ratio (C/N), was analysed by a modified closed non-linear regression model. Between 63% and 81% of the variation of soil CO2 emission Could be explained with changes of T-soil RSWC, pH, and C/N for the individual chambers on the brownfield. Subsequent analysis involved a comparison of the soil chamber results with eddy covariance (EC) measurements of one week, and included a footptint analysis to account for the influence of the different land use types on the measurements. For this, EC data ( 143 measurements after quality check) were restricted to those originating from the brownfield area with more than 90% of the flux. For a second comparison, the net ecosystem exchange (NEE) was calculated for different parts of the meadow using the SVAT model PROXEL. Together with the respiration from the brownfield. a weighted average of model NEE was produced using the flux contribution determined by the footprint model. Acceptable agreement (r(2) = 0.69) was found between the modelled data and individual EC measurements, except during situations where the performance of the footprint model was disturbed by internal boundary layer effects.