Another major source of difference is, of course, the analysis 

 of the meteorological situation from the data shown on -the weather 

 maps. In this case the fetch lengths, decay distances, wind velocities 

 and durations were chosen essentially identical with those used by 

 Bretschneider; these values have been checked and used by other authors 

 (l,U,5) and would appear to be the best available data. However, the 

 selection of these values is largely subjective, and other analysts 

 might well select different values. The use of different values (as 

 say, lower wind velocities and positioning of the front edge of the 

 fetch closer to the indicated front - see Figure l) while not necessarily 

 reducing the differences between the results obtained by the two methods, 

 would result in values obtained by the Pierson-Neumann method being 

 much closer to the observed. 



In the use of the Pierson-Neumann method an additional subjective 

 factor enters in the determination of the fetch width. Fetch width 

 in relation to decay distance has important significance in the 

 Pierson-Neumann theory; that is, the greater the fetch width, the 

 greater the wave height, all other things being equal. There is 

 possibility of error here, in that exact determination of the fetch 

 width is dependent on the accuracy of judgment of the user, which is 

 to be obtained only by a great deal of experience in the use of the 

 method and comparison with recorded values. Such experience is lacking 

 here, and it is quite possible that a more applicable value could have 

 been chosen. 



The angular spread of the waves as they are propagated from the 

 fetch front to a point of observation is also of importance in this 

 method. An energy correction factor for this angular spread is a function 

 of the angles formed by drawing lines from the point of observation to 

 the two extremes of the fetch width and extending lines from the fetch 

 sides (see Figure l). Assuming a true initial angle of 7°, a difference 

 of 15° in fetch direction, in a negative (upward) direction, would result 

 in an increase of US percent in wave height. Conversely, a difference 

 of 15° , in a positive (downward) direction, would result in a decrease 

 of 68 percent in wave height (see Figure 3). A graph depicting per- 

 centage increases and decreases in wave height for true initial angles 

 of 7°, 12°, 22° and 32° and angular differences of 1° through l£° is 

 shown on Figure U. 



A combination of differences in fetch width and angular spreading 

 values may give rise to possible wave height increases of large magnitude. 

 For example, assuming an angular negative difference (upward) of 10° 

 in combination with a fetch width difference of 2 (twice the fetch 

 width being used) the resultant wave height would be 1.7 times that 

 otherwise obtained. This is based on an initial angle of 7°j for 

 initial angles of greater degree, the wave height increase would be 

 relatively smaller. Figure 5 illustrates the relative height increases 

 due to an angular negative difference of 1!?° for initial angles of 

 7°j 120,220 an d 320. For a positive (downward) angular difference of 

 150 f the resultant effect upon the wave height is negligible, ranging 

 from 0.93 to 1.12 for the same angular range as above. 



