|
Description:
|
The flux of momentum between the atmosphere and ocean , described in terms of an aerodynamic drag coefficient (CD ) , is required to accurately forecast hurricane track and intensity ; to predict hurricane storm surge , ocean waves , and currents ; to define wind load standards ; and to generate hurricane risk models . Recently , a significant effort has been made to advance our understanding of air -sea momentum exchange in hurricane winds via the development , refinement , and implementation of novel instrumentation such as the Global Positioning System (GPS ) dropwindsonde and the ‘Best’ Aircraft Turbulence (BAT ) probe . However , these platforms are best suited for measurements in an open ocean (deep water ) environment . As deep water datasets continue to grow , it is apparent that data collected near the coast remain inadequate . Ironically , the coastal region is where the accuracy of hurricane models and building code provisions are most severely tested . This vacuity is exacerbated by the continued increase in wealth , infrastructure , and population along the hurricane -prone coast . Whether the nearshore drag coefficient differs from deep water observations or from historic linear formulations with wind speed has yet to be determined and is the focus of this dissertation .
To help fulfill the need for a better understanding of nearshore air -sea momentum flux , a combination of laboratory and full scale studies are presented . Since it is exceptionally difficult to conduct in -situ measurements over shoaling waves (waves that are in shallow -enough water to interact with the sea floor and undergo transformation processes ) , an innovative wind tunnel study was designed . The atmospheric boundary layer wind tunnel facility at Texas Tech University (TTU ) was used to characterize the wind flow and determine the drag coefficient over a statistically valid train of fixed shoaling wave models . Methodologies employed in this analysis were tested in a pilot experiment utilizing a train of fixed sinusoid waves (this work also provided general limits for CD ) . It was determined that fixed wind tunnel waves can be compared to propagating ocean waves , provided that the latter are in strong winds and limited by fetch (e .g . , hurricane waves ) . Results indicated that the drag coefficient over shoaling waves is approximately 50 % higher than those observed in deep water hurricane conditions . Wind tunnel values are in relative agreement with preliminary estimates of CD over shoaling waves in hurricane conditions . This study also found that the air flow does not separate over shoaling waves (the flow remains attached ) and that a pronounced speed -up region is present over the wave crest . Once validated by shoaling wave datasets , this methodology can be effectively used to estimate the drag coefficient and examine the flow over other critical wave shapes .
To complement the laboratory results , a joint field campaign during the 2008 Atlantic Hurricane Season collected valuable nearshore wind and wave data as Hurricane Ike made landfall near Galveston , TX . Coastal drag coefficient behavior was similar to that found in deep water , where CD increased with wind speed , reached a limiting value , and decreased thereafter . Crucially , at wind speeds below the limiting value , drag coefficients were significantly higher than those previously measured in deep water , in lakes , or in wind /wave laboratory studies . Based on this analysis , storm surge models using a deep water wind speed dependent drag coefficient are likely to underestimate hurricane storm surge , and additional parameterizations are needed . Coastal roughness lengths computed from these data provide evidence that the American Society of Civil Engineers (ASCE ) wind load code should prescribe Exposure D (smoother ) rather than Exposure C (rougher ) along hurricane prone coastlines . |