THE SIGNIFICANCE OF TEMPERATURE STRATIFICATION 

 IN THE ARCTIC 



by 



Stephen Neshyba, Victor T. Neal 

 Dept. of Oceanography, Oregon State University 

 Corvallis, Oregon 



Warren Denner 



U, S, Naval Postgraduate School 



Monterey, California 



Introduction 



A number of measurements and studies of the fine structure of 

 thermal and haline properties of the ocean have been published (Stommel 

 and Fedarov, 1967; Gaul, 1968; Cooper and Stommel, 1968; Tait and 

 Howe, 1968). The emerging picture of the ocean structure shows the 

 vertical column separated into numerous layers, each more or less 

 homogeneous, and each separated from adjacent layers by thin sheets 

 characterized by large temperature and salinity gradients. Increasing 

 use of continuous profile sensors has resulted in several observations 

 of layering in the ocean. However, because of the difficulties involved 

 in obtaining field measurements, laboratory studies are also important 

 in providing an understanding of the processes involved. 



Results of laboratory experiments on vertical thermal convection 

 processes were published by Ostrach (1957) and by Globe and Dropkin 

 (I959). The latter workers used dimensional analysis to arrive at a 

 functional relationship between the Nusselt number, Nu, and the Rayleigh, 

 Ra, and Prandtl numbers, Pr: 



Nu = CRa Pr . 



C is a coefficient determined empirically^ and exponents m and n are also 

 determined empirically. The Nusselt number is an indicator of how 

 effectively heat is transferred by convection rather than conduction; when 

 this number is equal to 1, convection does not exist. Dropkin and Somer- 

 scales (I965) performed similar experiments on a slanted cell and reported 

 changes only in the value of the coefficient C. More recently, Prenger 



457 



