An evaluation of the flux-gradient and the eddy covariance method to measure CH4, CO2, and H2O fluxes from small ponds

Despite their small overall area, small ponds play a large role in the greenhouse gas budgets of inland water bodies. This study aims to evaluate the performance of the flux-gradient (FG) and the eddy covariance (EC) method for measuring the fluxes of CO2, CH4, and H2O at two small fish ponds (fetch < 120 m) in subtropical climate conditions. The EC fluxes were subject to two sources of error: high frequency flux loss and footprint contamination. Of the three gaseous fluxes, the CH4 flux suffered the largest high frequency loss (18%) due to a combination of low EC instrument height and long optical path of the CH4 analyzer. Despite the low measurement height, the EC fluxes were influenced by sources outside the boundary of the target fish ponds, with the footprint contamination most severe on the CO2 flux and least severe on the CH4 flux. With regards to the FG method, one major uncertainty lies in the eddy diffusivity calculation. Of the three eddy diffusivity models evaluated [the aerodynamic (AE) model deploying the full Obukhov stability correction, the modified Bowen ratio model using H2O as a tracer, and the wind profile model for neutral stability], the AE model yielded the best results for the CO2 and CH4 fluxes. Our results support Horst's (1999, Boundary-Layer Meteorology 90, 171) theoretical prediction that the footprint of the AE flux based on a two-level concentration profile measurement should be much smaller than that of the gradient flux footprint and the EC flux footprint at the geometric mean of the two heights. We conclude that the most appropriate micrometeorological method for measuring fluxes from small water bodies is a hybrid scheme, whereby an EC system is deployed to measure the eddy diffusivity and a precision gas analyzer is used to obtain the concentration gradient of the target gas.