The complexity of clear air turbulence and its measurement and modeling for aviation safety,turbulent transport, and propagation studies

At the 9th Conference on Aviation, Range, and Aerospace Meteorology in September 2000 we reported on aircraft turbulence measurements in the winter sub-tropical jet stream, which used the Airborne Research Australia (ARA) GROPB 520T EGRETT. Subsequently reported budget and spectral analyses of the data have shown the need to extract more complete information about turbulence dynamics to support development of a clear air turbulence forecast method. A critical step toward that goal is more complete information about correlations of fluctuating static pressure gradients with fluctuating velocity components or temperature. These correlations, through sign and amplitude, scale the inter-component energy transfers that lead to isotropy and the important destruction rates of Reynolds stress and heat flux. The complete atmospheric turbulence dynamics and budgets can also be investigated by a direct numerical simulation (DNS) model calculation. The DNS problem is how much of the large scale is ignored because it cannot fit into the computational volume and how much of the small scale is ignored because it falls in the unresolved sub-grid volume. Aircraft measurements have a similar problem: how long a sample to take and how fast can or should the instrument be sampled. Furthermore aircraft measurements in the atmosphere occur at turbulence Reynolds numbers about 1000 to 10,000 larger than can be presently simulated in a DNS calculation. The impact of this Reynolds number gap on turbulence model development using both aircraft and model data needs to be determined. Turbulence structure function models derivable from turbulence energy balances involve turbulent length scales that cannot be derived from model output of meso-scale numerical weather prediction models such as MM5 or WRF. Our current goal is to relate the important length scales and length scale ratios with fine resolution (less than 100 meters?) bulk gradient Richardson numbers using both aircraft and DNS data and use these relationships in MM5 or WRF models. More complete turbulence budget calculations are expected from a two aircraft turbulence measurement campaign planned for July/August 2004 over the Flinders Range of mountains in Australia.