Comparison of temperature and wind statistics in contrasting environments among different sonic anemometer-thermometers

Sonic anemometers-thermometers (SATs) are robust instruments used in numerous research and analytical micrometerological studies. The accuracy and precision of the measured mean and turbulent fluctuations of wind speed and temperature are unknown across a range of ambient conditions and among different model SATs. Here, we compared quantities from eight different model SATs: (i) mean temperature (T-s) in an acoustically isolated chamber, (ii) mean vertical wind speed (w(s)) in a low-speed wind tunnel, and (iii) wind statistics with data collected over a research field. Potential differences in buoyancy flux (w'T') due to different responses among SATs to changes in air temperature were also examined. The T-s response from each model SAT to air temperature departed from a 1:1 relationship across all, or part, of the range in tested temperatures. w(s) from all SATs did not behave 1:1 to an independent measure of vertical wind speed using a hot-film anemometer, and there were consistent patterns based on the physical T among SATs and their potential to affect scalar fluxes are discussed. Large design of the SAT. The observed differences in sigma(w)(2) and sigma(T)(2) among SATs and their potential to affect scalar fluxes are discussed. Large variability was observed in wind statistics among SATs in field conditions. Uncertainty in w'T' among sensors due to their different responses to T-a for each 15-min averaging period ranged -23.1 to +16.1%, and range from -1 to +8% when averaged over similar to 940 15-min periods. Use of SAT derived data are discussed for: (i) estimating fluxes, advection, and the WPL term, (ii) comparison of data from multiple SATs in an individual study, and (iii) temporal and spatial scaling or comparisons of flux estimates that were derived from different model SATs. (c) 2005 Elsevier B.V. All rights reserved.