An Examination of Vertical Sampling Methods and Their 

 Influence on Dynamic Height Calculations 



by 

 Ronald C. Kollmeyer, U.S. Coast Guard 



Introduction 



The primary objective of the Ice Patrol surveys 

 is to gather dynamic height information on the 

 Labrador Current and Gulf Stream off the Grand 

 Banks of Newfoundland for the purpose of pre- 

 dicting iceberg drift from geostrophic current 

 maps. Dynamic heights are calculated from 

 temperature and salinity measurements of the 

 water at various depths. 



For 35 years Ice Patrol surveys have sampled 

 temperature and salinity at standard depths of 

 0. 25, 50, 75, 100, 150, 200, 300, 400, 600, 800, 

 1,000, and 1,500 meters. This standard sampling 

 system is a compromise between the ideal of 

 continuous vertical water column values on the 

 one hand, and the practical necessity of rapid 

 survey with a minimum quantity of samples on 

 the other. 



A greater economy of effort and equipment, 

 and more reliable oceanographic inferences, could 

 be obtained if gradient depths (i.e., depths of 

 sharp water property change) were known prior 

 to attaching Nansen bottles to the wire. The 

 intervals between bottles would then be set to 

 assure that no important changes were missed, 

 and that depths of changes were more accurately 

 determined. Toward this end an electronic bathy- 

 thermograph with a depth limit of 410 meters 

 was used during Ice Patrol 1965 to provide 

 information critical to the spacing of Nansen 

 bottles. This gradient sampling method is com- 

 pared to the standard sampling method herein, 

 in an attempt to determine their merits and 

 failings. 



A continuous sampling device which gives a 

 very great or unlimited number of data points 

 such as to provide an unbroken record of tempera- 

 ture and salinity through the entire column is 

 ideal, eliminating problems arising from Nansen 

 bottle spacing altogether. The measurement ac- 

 curacies of such a device, as they apply to dynamic 



height calcidation, was examined and the results 

 comprise the latter part of this paper. 



Sampling Methods 



Analysis of the waters off the Grand Banks 

 discloses that this area has some of the most un- 

 usual variation in vertical distribution of water 

 properties in the world. With the Labrador Cur- 

 rent running parallel to the Gulf Stream in some 

 areas and sinking beneath it in other locations, 

 vertical changes in temperature and salinity are 

 quite severe over short distances. Figures iD, 

 2D and 3D show e.xamples of temperature and 

 salinity distribution of water columns down to 

 1,000 meters as determined from data obtained 

 during Ice Patrol 1965. 



Designing an adequate sampling program for 

 water masses such as shown presents a problem. 

 Since great changes in the vertical distribution of 

 temperature and salinity occur at various levels 

 for different stations, a fixed point sampling pro- 

 gram will frequently miss critical maximum and 

 minimum points in the property distribution. An 

 example of this is demonstrated in figure 2D 

 where the extreme values would have been missed 

 by the standard sampling points shown above. 

 Conversely, many Ice Patrol areas are very uni- 

 form, as shown in figure 4D, where using even a 

 small number of Nansen bottles would suffice. 



During Ice Patrol 1965, a gradient sampling 

 approach was used. This method allows the posi- 

 tioning of Nansen bottles to be based on the results 

 of a preliminary bathythennograph drop at each 

 station. A Hytech Electronic Bathythermograph, 

 Model 480 MOD 1 with a modified sensing fish 

 was used to detennine the temperature structure 

 down to 410 meters, the limiting depth of the 

 instnmient. Thus a more complete description 

 of the temperature distribution will be obtained 

 by reversing thermometers and if a stable water 

 column is assumed the variation of salinity with 



65 



