The general geometry consists of a grid of 519 space points equally 

 spaced on a Lambert Conformal map. The directional spectra at each of 

 these points consist^, of the values for 12 directions and 15 frequencies. 

 The development and verification phases of this study required certain basic 

 wave and atmospheric data. 



Wave, Pressure, and Wind Data. 



The basic data on waves are described in reports by Moskowitz, 

 Pierson and Mehr (1962, 1963), Bretschneider, Crutcher et al (1952) and 

 Pickett (1962). These reports reproduce the spectra obtained by an analy- 

 sis of wave records obtained by weather ships of the United Kingdom and 

 by the U. S. Naval Oceanographic Office at Argus Island. 



These spectra as computed have provided a basic source of data for 

 many problems concerning waves. The supply of the reports by Moskowitz, 

 Pierson and Mehr (1962, 1963) is currently exhausted, but it is planned to 

 reissue it. These data are also available on magnetic tape. 



In order to compile wave climatology for the year 1959 (which was 

 the one chosen) it was necessary to have surface wind fields. Such data 

 were not available. The atmospheric dat^ which were available were 

 pressure fields every 12 hours over the JNWP grid. A subsidiary problem, 

 but one which turned out to require the most effort, was to generate correct 

 wind fields. This was not possible only in terms of the available pressure 

 observations. In addition to these pressure observations some 300,000 

 ship reports had to be obtained. 



It was a difficult task to use these data to obtain wind fields at six hour 

 intervals for the desired elevation above sea level. Wave hindcasts(and 

 forecasts) are extremely sensitive to the wind field, and it is our belief 

 that a major part of the error of our present procedure is due to errors in 

 describing the wind field. 



The computer program for determining the wind field over the North 

 Atlantic is quite interesting. A regression analysis of the pressure field 

 and ship reports as described by Thomasell and Welsh (1963) yields a first 

 estimate. The key ships are the weather ships and the U. S. Navy ships. 

 For these, anemometer heights are known, and the reported speeds are 

 corrected by the logarithmic wind profile to a constant reference level of 

 19-5 meters above the sea surface. These corrected winds were used to 

 refine the first estimate of the wind field computed from the pressure field. 

 The anemometer heights on these ships vary over a wide range and bias the 

 wind field in an uncontrollable way unless this correction is made. 



The pressure fields were available on magnetic tape only every 12 hours 

 at 00 Z and 12Z. To get the winds at 06Z and 18Z, it was necessary to 

 devise an interpolation scheme that may have degraded the hindcasts to 

 some extent. The low centers on the OOZ and 12 Z charts were located 



74 



