Comparison of 3-D sonic anemometers

Flux measurements were made with a Campbell Scientific CSAT3 3-D sonic anemometer and a vertically symmetric Gill R2 sonic anemometer at 6m above a 33-m tall Douglas-fir stand, Vancouver Island, BC, Canada. An open path hygrometer and a closed path infrared gas analyzer were used in the water vapour and carbon dioxide flux measurement. Analogue eddy covariance signals were measured with a data acquisition system sampling at 125 Hz and transferred to a PC for on line flux calculations. Sensible heat fluxes were calculated using the sonic temperature and temperature from a fine wire thermocouple. Half-hourly fluxes, friction velocity and wind speed from the two anemometers agreed well with linear regressions having R²>0.96, slopes of >0.95 and intercepts close to zero. For example, regression of sensible heat fluxes using sonic temperatures were CSAT3=0.967Gill + 14.1 W/m², R²=0.951, n=289, and CSAT3=0.979Gill mmol/mol/m²/s, R²=0.993, n=277 for carbon dioxide. A comparison of the Gill heat fluxes from sonic and thermocouple temperatures show a diurnal hysteresis. This was not present in the CSAT3 comparison. In a separate study, a comparison of half-hourly sensible heat fluxes from the CSAT3 and a Campbell Scientific 1-D sonic anemometer, 0.6 m above 1-m tall lodgepole pine and fireweed had CSAT3=1.014D1+1.7 W/m², R²=0.98, n=192.