"science," similar to that existing in modern weather fore- 

 casting. In figure 2 the final forecasted sounding is shown 

 by the solid line. 



6. Example of forecast 

 a. Introduction 



The methods for estimating the effect of the meteor- 

 ological parameters on the thermal structure, as outlined in 

 Sees 2 to 5 above, were tested by attempting a prediction 

 utilizing actual data. For this purpose there was available 

 a collection of observations taken by the U. S. Navy Hydrographic 

 Office on cruise AMOS VIII in April 1951. On this cruise, some 

 of the latest instrumentation developed by the Oceanic Develop- 

 ment Branch, Division of Oceanography, were employed. It was 

 decided to utilize for the forecast that portion of the data 

 gathered in the Sargasso Sea area, since here the data were 

 more complete and the mean wind speeds were generally 12 knots 

 or below, thus allowing the assumption that the turbulent effect 

 vjas small. These data include: (a) hourly BT's; (b) hourly 

 observations of meteorological variables; (c) hourly observat- 

 ions of sea and swell; (d) transparency stations using photocell 

 equipment every few days; and (e) continuous pyrheliometer 

 recording of incoming and outgoing radiation. 



The type of radiation data available is illustrated 

 in figure 3, It will be noted that these data were taken 

 by three separate pyrheliometer s, each recording directly 

 the rate of radiation in g, cal, cmT'- minrl. Pyrheliometers 

 1 and 2 viere mouxi.ted on either side of the ship to measure 



19 



