ture as a function of day-of-year for each of the 10-minute-by- 

 10-minute shaded areas. The dots are measured temperatures 

 within the area. To the right of the strip is a histogram showing 

 the distribution of the observed data sample by months; to the 

 right of the histogram the variation of temperature with longitude 

 for the first day of the month is shown; and to the far right is 

 shown the distribution of the differences between observed and 

 estimated temperature. Pertinent statistics are given at the top 

 of each figure . 



A qualitative study of these f^ures does not reveal any 

 contradictions to generally accepted characteristics of the varia- 

 tion of sea-surface temperature in the areas covered by the 

 analyses. 



It is concluded that the regression model, as exjDressed in 

 equation (4), is a physically acceptable estimator of seasonal and 

 spatial variations in sea-surface temperature in the areas covered 

 by these analyses. 



59 



