Measurement of horizontal and vertical advection of CO 2 within a forest canopy

Abstract

Eddy covariance measurements often underestimate the net exchange of CO 2 between forest canopies and the atmosphere under stable atmospheric conditions, when horizontal and vertical advection are significant. A novel experimental design was used to measure all terms in the mass balance of CO 2 in a 50 m × 50 m wide, 6 m tall control volume (CV) located on the floor of a 40 m tall Eucalyptus forest to examine the contributions of the eddy flux, the change in storage and the horizontal and vertical advection terms. Horizontal flux divergences between the four vertical walls of the CV were determined using perforated tubing arranged parallel to the ground to measure CO 2 mixing ratios. The change in storage was calculated using CO 2 concentration profiles measured in the centre of the CV. Vertical advection was calculated using these profiles, combined with vertical velocities, w c , calculated using the mass continuity equation and horizontal velocities measured at the mid-point of each wall of the CV. Vertical and horizontal advection and the eddy flux terms all contributed significantly to the mass balance of the CV at night, while the eddy flux term was dominant and negative for a short period around noon when photosynthesis exceeded respiration. Large vertical gradients of CO 2 at night cause estimates of vertical advection to be extremely sensitive to small errors in w c with standard errors of the mean flux exceeding 3 μmol CO 2 m −2 s −1. Vertical velocities need be measured to an accuracy better than 1 mm s −1 to minimize errors in vertical advection when vertical gradients of CO 2 ratios are very large at night. Calculated horizontal advection is sensitive to errors in the wind vectors through the faces of the CV when horizontal concentrations gradients are large. Errors in eddy fluxes and change in storage are smaller than for the advection terms and errors for all components are smaller during the day than at night.