The array used during the fall of 1959 consisted of two buoys 

 500 feet apart forming an equilateral triangle with the vertical guide 

 cables below the hydrographic room of the tower (Figure 3). The 

 two buoys on the east-west base of the 500-foot triangle supported 

 resistance thermometers at nominal depths of 65 and 100 feet below 

 mean water level (MWL). Thermometers at the tower were placed 

 at 65, 80, 100, and 130 feet below MWL. During the spring of I960, 

 the array consisted of three buoys positioned as shown in Figure 

 4 with resistance thermometers placed at approximately 50, 70, 90, 

 and 110 feet below MWL. Note that the equilateral triangle was 400 

 feet on a side instead of 500 feet and the vertex was 225 feet south of 

 the tower to minimize influences of the structure. 



The buoy arrays were manually installed from a 26-foot surfboat. 

 Sextant angles taken from the boat, supplemented by sight lines relative 

 to tower caissons, were used in the fall to position the two buoys. 

 Considering drift rate, sea conditions, sextant and sighting errors, 

 and the 40- to 50-foot freedrop of the buoy when cut loose over station, 

 the actual positions of the buoys are estimated to be within 40 feet of 

 the predetermined positions. 



Two azimuth instruments accurately located at the end points of 

 an east-west 100-foot baseline of the tower's flight deck were used 

 for horizontal control in positioning the three buoys in the spring. 

 Azimuth angles were computed to the nearest 15 minutes of arc 

 — equivalent to the precision with which the sighting instruments could 

 be preset. The relative locations of these buoys are estimated to be 

 within 10 feet of the calculated stations. 



Two 4-channel "Brown" balancing potentiometer recorders were 

 used during the fall to record resistance thermometer output on a 

 0° to 30°C full scale. The sampling rate for a given channel depended 

 on the temperature differences between the four channels and normally 

 averaged 6 to 10 seconds. For the spring survey, the outputs from 

 the four thermometer levels of all three buoy positions were recorded 

 on a single Brown recorder modified to handle 12 different information 

 channels. The sampling interval for an individual information channel 

 was usually from. 18 to 30 seconds. GMT marks were manually added 

 to the records at one- to eight-hour intervals to serve as time control 

 between recorders and supplementary observations. Precision of 

 these marks is estimated to be ±3 seconds for the majority of chart 

 speeds. 



