INTRODUCTION 



The information in this report was obtained from BT(bathythermograph) 

 data collected at or near the assigned positions of North Atlantic Weather 

 Ships. The observations are on file at the Woods Hole Oceanographic Institu- 

 tion and were loaned to Texas A. & M. for this study. The methods of handling 

 these data were given in detail in Technical Report No. 1 of this project, Some 

 Methods Used in Representing Bathythermograph Data , but will be restated 

 briefly here. 



For each year in which there are sufficient data at a given station to 

 warrant this type of study, three time graphs have been drawn. These are: 



1) Depth of mixed layer and of water 2°F colder than the 

 mixed layer; 



2) Temperatures at selected depths; and 



3) Depths of selected differences from sea surface temperature. 



To obtain these time gi'aphs of BT data, a procedure involving four separate 

 steps was employed: 



1. Selection of cards. Only one card was used to represent a parti- 

 cular day. This card was one selected as being typical for that day, factors 

 considered being sea surface temperature, depth of mixed layer, depth and 

 nature of the thermocline, time of observation and general character of the 

 curve. When possible, the observation made nearest to 1200 hours (GCT) 

 was chosen. 



2. Reading of cards. A team of two people was used for all card read- 

 ing, each reading being checked to assure accuracy. Temperatures at ten 

 chosen depths (0,25, 50,75, 100, 150,200,250,350, and 450 feet) were read, 

 with an agreement between readers of - 0. 1°F being required for each depth. 

 The depth of the mixed layer (this depth being defined as the depth to a 



■t 0. 3 F change from the temperature of the sea surface) and the depth of 

 water 2°F colder than the mixed layer were then read, the required agree- 

 ment in these cases being - 3 feet. Finally the depths of selected differences 

 from sea surface temperature were read, agreement again being - 3 feet. 



3. Smoothing of data. A running five day average was determined to 

 make a picture showing the nature of gradual changes with time. The value 

 for a particular day was obtained by averaging the value for that day with the 

 two immediately preceeding and the two following days. Exceptions occurred 

 at the beginning or end of a series of days or at places where for one or two 

 days there were no data available. In these cases four day or, in a few cases, 



■1- 



