HATS: Field Observations to Obtain Spatially-Filtered Turbulence Fields from Crosswind Arrays of Sonic Anemometers in the Atmospheric Surface Layer

T.W. Horst1, J.Kleissl2, D.H. Lenschow1, C. Meneveau2, C.-H. Moeng1, M.B. Parlange2, P.P. Sullivan1, and J.C. Weil3
1.National Center for Atmospheric Research, Boulder, CO, 2. The Johns Hopkins University, Baltimore, MD, 3.CIRES, University of Colorado, Boulder, CO

ABSTRACT: The Horizontal Array Turbulence Study (HATS) field program utilized horizontal, crosswind arrays of sonic anemometers to calculate estimates of spatially- filtered and subfilter--scale (SFS) turbulence, corresponding to its partitioning in Large Eddy Simulations (LES) of atmospheric flows. Measurements were made over a wide range of atmospheric stability and for z/\Delta_f nominally equal to 0.25, 0.5, 1.0 and 2.0, where z is height, \Delta_f = 4S is the width of the spatial filter, and S is the crosswind sonic spacing. An analysis is made of uncertainties in the derived turbulence fields. Aliasing in the crosswind direction, caused by the finite spacing of the sonic anemometers, is found to be minimal except for the spatially--filtered vertical velocity and for SFS second moments. In those cases, aliasing is significant for k_{yN}\Lambda_w^y < 3, where k_{yN} = \pi/S is the Nyquist wavenumber of the crosswind sampling and \Lambda_w^y is the wavelength at the peak in the crosswind spectrum of vertical velocity. Aliasing errors are estimated to be of a similar magnitude for the crosswind gradients of filtered variables. Taylor's hypothesis and the mean wind speed U are used to interpret sonic time series as spatial data for the purpose of surrogate streamwise filtering. HATS estimates of the turbulence advection velocity U_c based on the spatial autocorrelation function suggest that, for near--neutral stratification, the ratio U_c/U depends on the turbulence variable of interest and is typically between 1.0 and 1.2. Analysis of LES turbulence data finds that the correlation between the true 2D-spatially-filtered SFS stress component \tau_{13}and the same variable obtained with surrogate streamwise filtering exceeds 0.98 for z/\Delta_f > 0.25. Within the limits already noted, it is concluded that the horizontal array technique is sufficient for the estimation of resolved and subfilter-scale turbulence variables.

(2004) J. Atmos. Sci. 61, 1655-1581.

full pdf article -- ( © AMS, see http://www.ametsoc.org/AMS/)

§ Archival Journal Publications: Articles may be downloaded for personal use only!
Any other use requires prior permission of the Author and the Publishers.