Aeolian boundary-layer characteristics across a sand beach

Wind velocities were measured across a beach in the Guadalupe-Nipomo Dune Preserve, Santa Barbara County, California using six instrument towers aligned over 125 meters in the cross-shore direction. The towers contained 54 anemometers with heights above the sand surface ranging between 0.1m to 1.5m. Measurements were conducted during on-shore wind conditions to examine the character of internal boundary-layer evolution across this sandy beach.; The first part of this thesis addresses the long-term (ca. 15 minutes), time-averaged characteristics of the turbulent boundary layer. Parameter estimates from a constant stress model (the law of the wall), the linear stress model (a pressure gradient modified version the law of the wall), and a more general non-linear stress model (the law of the wake) are compared and contrasted. The least-squares fit of these theoretical velocity profiles to the time-averaged velocity data does not indicate whether any one of these profile is clearly better. The maximum relative differences among the estimates of shear velocity from these known velocity profiles is less than 32%, even with similar R2. Because of uncertainty in estimates of shear velocity, predicting sediment transport by various empirical formulae that incorporate shear velocity as the sole flow-field variable remains problematic.; The second part of this thesis examines the character of unsteady flow structures embedded within the wind field. The time series of horizontal wind speed has a gusty character with varying scales. These detailed temporal structures are not revealed by ordinary moving-average or block-average methods. A variable-interval time-averaging technique (VITA) was therefore applied to the time series. VITA is an objective means of identifying particularly energetic "events" embedded within a time series (e.g., sweep-ejection events in turbulent boundary layers). This analysis shows that the VITA-identified events are of similar duration (about 3--12 seconds), but aperiodic and of variable magnitude. Characteristic shear velocities were then estimated from the burst-based velocity profiles, and these are 5%--10% greater than the overall shear velocities obtained from the mean wind-speed profiles. The implications for aeolian sediment transport may be significant. First, it is unlikely that process-response interactions in these systems will reach an equilibrium state (except, perhaps, in the very short term). Second, the conventional approach to sediment-flux prediction using long-term averages is unlikely to yield accurate results given the non-linear relation between sediment flux and shear velocity and given the unsteady (gusty) nature of the wind field. Nevertheless, we can presume that the empirical constants in these empirical model must account, in part, for some of the unsteadiness.