The wind-stress factor accounts for the nonlinear relationship between wind 

 stress and windspeed. 



The approximations and adjustments used are made to reduce biases in wind 

 data and to provide a reasonable means of providing information where adequate 

 measurements are not available. The collection of overwater wind data at a 

 site is preferable. Even if data can only be collected for a short period, 

 say 1 year, it may be of value in relating overland wind data to overwater 

 values. 



2. Procedure for Adjusting Winds Observed Over Water . 



Wind data gathered over the water are normally the most desirable for wave 

 prediction. Most overwater wind data are gathered by observers on ships as 

 visual observations of unknown quality. Cardone et al. (1969) reviewed bias 

 in ship-observed windspeeds and suggested that a correction of 



(I) 



W = 2.16 W^' (3-29) 



s 



where W is the ship-reported windspeed in knots and W is the corrected 

 windspeed in knots. In most cases, the elevation above the water surface 

 where ship windpseeds are measured is variable and unknown. Other wind 

 measurements may be taken on lightships or with automatic buoys. The 

 following procedures should be used in correcting winds observed over water 

 for use in the wave prediction formulas: 



(a) If the winds are from ships, they should be corrected for 

 bias by equation (3-29). 



(b) If the winds are measured at an elevation other than 10 

 meters, equation (3-26) should be used to correct the windspeed. 



(c) The windspeed should be adjusted for the stability effect 

 from Figure 3-14. 



(d) The duration-averaged windspeed is estimated by Chapter 3, 

 Section IV,l,b. 



(e) The windspeed is converted to the wind stress factor 

 (Chapter 3, Section IV,l,e). 



3. Procedure for Estimating Overwater Winds from Nearby Land Winds . 



The following procedure should be used to obtain the overwater windspeeds 

 from observations nearby on land. In the Great Lakes this procedure was 

 successfully used to obtain estimates up to 113 kilometers (70 miles) away 

 from wind stations normally within 16 kilometers (10 miles) of the lake; this 

 was possible because of the size of the lakes and storm systems and the 

 flatness of the topography. Also, multiple stations were used to obtain some 

 spatial variability. 



3-32 



