Search Results
Abstract
In previous work the derivation of turbulence parameters from single-Doppler radar observations was performed with data acquired along a horizontal circle. Here the technique is extended to all the radar data within a horizontal cylindrical slice of finite depth using the same basic assumptions of linearity of the mean wind field and horizontal homogeneity of the turbulence. The method allows the extraction of the six Reynolds stress components, together with their vertical derivatives, and the turbulent fluxes of a scalar quantity deduced from the reflectivity data.
Experimental data were used for the performance evaluation of the methodology. A simple testing procedure was carried out to remove erroneous results. The statistical uncertainty in the measured Reynolds stress terms was found to be about 0.05 m2 s−2, except for the variance of the vertical component, which was poorly retrieved because of an absence of data at high elevation angles. Calculations showed that contamination of the vertical momentum flux measurements by the scatterer fall speed was negligible. An analysis of the response function of the technique to the atmospheric scales tended to show that the diameter of the processing slices corresponded to the largest turbulent scale dimension involved in the measured turbulence quantities.
Abstract
In previous work the derivation of turbulence parameters from single-Doppler radar observations was performed with data acquired along a horizontal circle. Here the technique is extended to all the radar data within a horizontal cylindrical slice of finite depth using the same basic assumptions of linearity of the mean wind field and horizontal homogeneity of the turbulence. The method allows the extraction of the six Reynolds stress components, together with their vertical derivatives, and the turbulent fluxes of a scalar quantity deduced from the reflectivity data.
Experimental data were used for the performance evaluation of the methodology. A simple testing procedure was carried out to remove erroneous results. The statistical uncertainty in the measured Reynolds stress terms was found to be about 0.05 m2 s−2, except for the variance of the vertical component, which was poorly retrieved because of an absence of data at high elevation angles. Calculations showed that contamination of the vertical momentum flux measurements by the scatterer fall speed was negligible. An analysis of the response function of the technique to the atmospheric scales tended to show that the diameter of the processing slices corresponded to the largest turbulent scale dimension involved in the measured turbulence quantities.
