SEAWAY 



47 



ESE ESE SSW W NW WNW WNW NW W 



ESE SE SSW NW NW WNW WMW NW HW 



CCC CC\Ay CW MW\Af urklWI \A[kJUI kMW MIW 



E 

 ^ 4 



-E 



Fig. 48 Wave height and period at Port of Sakata from 

 September 13-14, 1955 (from Ijima, 1957) 



Figs. 48 and 49 were chosen from this set primarily be- 

 cause they permitted the best reproduction. 



Observations were carried out using an underwater 

 pressure meter installed at about 900 ft from the end of a 

 breakwater, where the depth was ol.8 ft. Recordings 

 were made continuously for 20 min every 2 hr. The 

 solid curves in Fig. 49 show the recorded pressure-wave 

 heights and the dotted cur\'es show the wave components 

 corrected for gage-depth attenuation. The time histor- 

 ies of spectrum growth and deca^' \'aried somewhat for 

 different recorded cases. Average conditions are how- 

 ever well represented by Fig. oO. Confirming the con- 

 clusions of Neumann and Gelci, et al, the spectrum de- 

 velops gradually from the high-fre(|ucncy end. How- 

 ever, contraiy to Neumann's assumption, considerable 

 energy is contained in the spectrum at periods longer 

 than 7'cax. No detailed analysis was made to show the 

 degree of compliance of the curves in Fig. 50 with those 

 expected from the various spectriun definitions. By 

 visual examination alone (remembering the opposite 

 directions of the abscissas T and l/T), Ijima's spectra 

 are seen to bear a close resemblance to those of Gelci, et 

 al, Fig. 45. 



The decay process appears to be the reverse of the 

 growth one. The longer waves disappear first, and the 

 Tmax is gradually shortened. In addition there is a cer- 

 tain attenuation of short-wave amplitudes. 



6.5 Comparison of Various Spectra and Methods of 

 Wave Forecasting. In this section the results of the var- 

 ious investigations will be compared under: 



oj Spectra of fullj' arisen sea. 



0.01 O.OZO-05 001 005 006 0-08 010 0.12 0.14 0.16 0.16 020 

 I I I I I I I I I I I I I 1 — 



50 



2S 20 15 



10 9 

 (sec) 



Fig. 49 Wave-energy spectrum of waves shown in Fig. 48 

 (from Ijima, 1957) 



b) Wa\'e development with time and fetch. 



c) 0\'erall results of wave forecasts. 



In all cases the form of the spectral curve and the pre- 

 dicted mean wave characteristics will be distinguished. 

 Since the practical importance of the three subjects 

 grows from a) to c), it is gratifjing to find (as will be 

 shown in the following) that the discrepancies between 

 the various methods diminish from a) to c), although 

 thej' ne\Tr vanish. 



6.51 Spectra of fully arisen sea. A comparison of 

 various spectra formulations are shown in Figs. 46 and 

 51. The demonstrated differences among these spectra 

 formulations are appalling, and indicate to this author 

 the inadequacy of an empirical approach and the press- 

 ing need to pursue development of a rational theory of 

 wave generation l)y wind. 



The observational data ior direct verification of the 

 spectra are extremely meager. The ideal conditions of a 

 well-defined single fetch and wind duration are \'ery sel- 

 dom found in nature. In strong winds incompletely 

 developed seas are usually found, and in lighter winds 

 (of minor practical significance) waves are most often 



