Mtecroscopte Structures of Wind Waves 
flow regime, 
1 
spt 070: 590 Na" 
vn( 2 - R % . a (6) 
n 
wherein the Reynolds number (RL = ah . L/v) is defined in terms 
of the free-stream air velocity, U_, the distance between the fan 
and the test section, L, and the kinematic viscosity of air, V 
Judging from the trends of the data shown in figure 5, we 
see that 
a) the air-flow boundary layer seems to be in the pretransi- 
tion region for U_< 1.9 m/sec. Because the air was sucked into 
the wind-wave tank by an axial fan and through guiding vanes, which 
were arranged to straighten the flow but not to diminish the high tur- 
bulence level, the latter arrangement was helpful to increase the effec- 
tive lenght of the wind fetch. 
b) the effective transition region of the boundary layer 
from laminar to turbulent is very narrow, at wind velocities between 
1.9 and 2.4 m/sec. 
c) once the boundary layer becomes turbulent, U_ >2.4 
m/sec, the transition from smooth to rough boundary layer takes 
place and this process is completed at Uo = 3.5 m/sec, the aerody- 
namically smooth flow regime is rather narrow. 
d) for the aerodynamically rough flow regime, the two 
groups of data are separated physically by the transfer of the gover- 
ning mechanism of wind-wave interaction from surface tension to 
gravity (Wu 1968, 1969); this separation occurs at iy. = 9m/sec. 
It should be emphasized here that because of the difference 
of scales (such as wind fetch) between the ocean and laboratory con- 
ditions and of differences of the wind structures (such as turbulence 
levels) between various wind-wave tanks, the shear velocity rather 
than the wind velocity should be adopted to characterize the wind 
conditions (Wu 1968). By use of the shear velocity, data obtained in 
the present experiment may readily be correlated with results of 
other investigations. 
III.2 Wave conditions 
From the continuous wave-profile recording, the periods of 
more than 100 basic waves for each wind velocity are obtained (Wu 
1441 
