A tunnel-shaped flow-through chamber for minimum disturbance net ecosystem flux measurements

Eddy covariance measurements are the method of choice to determine net land surface – atmosphere fluxes of energy and matter, but in some cases cannot be applied in a representative way. In particular, requirements on the measurement height and the principles underlying the fetch or footprint of the sensor inhibit straightforward application in small (<< 100 m upwind) ecosystems. Small chambers as used on the leaf and soil level, on the other hand, considerably disturb the ecosystem and require numerous repetitions in space, if robust area-average net fluxes are aimed at. Within the framework of a project also including the upscaling of chamber- and downscaling of micrometeorological measurements, we here present a third approach: A chamber with strengths and weaknesses that are expected to be in an intermediate range between traditional chamber and micrometeorological methods. In order to ensure comparability to micrometeorologically determined fluxes, the chamber was tested on three different fields of sufficient size against eddy covariance stations. The chamber is a flow-through type and covers a comparatively large (1.7 m²) ground surface. Measures to minimize ecosystem disturbance include a thin FEP foil ceiling, a wide in- and outlet, and ventilation at the order of magnitude of outside wind speed. Here, we focus on experiments with passive ventilation, where the in- and outlet are aligned into the mean wind and longitudinal matter advection is measured while suppressing any vertical and crosswind turbulent or advective flux. In order to comply with the small resulting concentration differences between in- and outlet, a differential closed-path CO₂ and H₂O analyzer is used. Comparisons to eddy covariance measurements show a good agreement and slight negative bias for evapotranspiration (latent heat flux), and a somewhat larger scatter and bias for CO₂ flux. The scatter of the CO₂ flux is hypothetically attributed to spatial variability between the footprint of both methods. The bias indicates a possible slight to intermediate underestimation of fluxes, depending on the hypothetical reasons of energy balance non-closure of the eddy covariance measurement. Possible reasons of such an underestimation are discussed, including remaining microclimate modifications as well as the longitudinal turbulent flux, and further tests of or modifications to the system are discussed.