Analysis of CASES-99 atmospheric surface layer data under a weakly-stable and a strongly-stable atmospheric boundary layer, respectively.

This work focuses on analysis of measurements of the atmospheric surface layer (ASL) from the CASES-99 experiment. Two situations were included: a weakly-stable (WSBL) and a strongly-stable boundary layer (SSBL), respectively. The objectives were to help understand the key atmospheric boundary layer processes involved, and to estimate profiles of the quantities needed to support computational fluid dynamics (CFD) modelling of the ASL. Additional objectives were to help identify a suitable Reynolds averaging period, and to gain an idea of the variability of the averaged quantities.

The sonic anemometer 20 Hz velocity data were rotated to stream-wise coordinates. A multi-resolution decomposition (MRD) analysis was then applied to the data, in order to estimate a suitable averaging period for the data analysis. Turbulence generation rates were estimated numerically, while spectral techniques were applied to estimate the turbulence dissipation rates.

The MRD results indicated that 5-min was an appropriate Reynold's averaging period for both the velocity variances and the covariances, respectively. However, significant variability in the 5-min variances over a one-hour window was observed, which was attributed to the effects of both random sampling errors and non-stationarity. One-hour averages of the 5-min values were used in an attempt to reduce the influence of this variability. Steam-wise velocity variance for the SSBL was found to be only about 3% of that of the WSBL; while vertical velocity variance was negligible for the SSBL and turbulence was 2-dimensional. Key boundary layer processes involved were considered. The 10-m dissipation rate for stream-wise velocity variance was found to compare well with estimates of the variance source generation rate. This result is encouraging as both terms should be closely linked.

The objectives of this initial work were met. Further analysis will focus on better understanding of the variability of the 5-min variances; and on the estimation of the dissipation rates.