40 N. 



30* 



20' 



160' 



150* 



140* 



13 0* 



I20°W. 



Figure 6.— Location of observations, SS CALIFORNIAN Voyage 105. Inbound from Honolulu to San Francisco. The 

 bar labeled 56-62 gives location of six closely spaced observations (fig. 9). 



PROBLEMS WITH INSTRUMENTATION 



During the trip a number of problems were 

 encountered with the Model T-3XBT system, 

 which gave rise to errors in some of the 

 traces and to a few outright failures. Subse- 

 quently, design and engineering changes have 

 been made in the circuitry, recorder, and 

 probes. The systems now in use are Model 

 T-4. In spite of the problems, many valid 

 data were obtained. The paragraphs which 

 follow in this section are devoted to a dis- 

 cussion and diagnosis of these instrumental 

 problems with regard to the interpretation 

 of the data presented. 



Temperature Errors 



When compared with the NSRT temperature 

 or a surface bucket temperature, a number of 

 XBT traces showed near-surface temperature 

 error greater than the specified design tol- 

 erances of -0.2° C. Comparison of tempera- 

 tures at 1,500 ft. (460 m.) indicated that these 

 same traces generally were in poor agreement 

 with adjacent observations. These data, along 



with tests that our Laboratory made in the 

 Hawaii area in 1965, indicated that the tem- 

 perature correction for traces having an 

 "offset" is not constant. The error appears 

 as if a constant resistance were added into one 

 side of the balancing bridge circuit or the 

 other, and because of the nonlinear relation of 

 temperature to resistance, the error (or cor- 

 rection) changes withtemperature. Such errors 

 occurred in eight of the observations published 

 in this report: 7, 12, 15, 17, 22, 27, 58, and 64. 

 The temperature corrections shown with the 

 curves were applied in preparing the tempera- 

 ture sections (figs. 7, 8, and 9). The individual 

 observations published in this report are direct 

 tracings, as recorded. If data are taken from 

 these traces, temperature corrections should 

 be applied. 



Failures 



Total failure to obtain any valid data with a 

 given probe was evidenced in two ways. Wild 

 oscillations of the pen (temperature spikes in 

 the trace) probably indicated electrical leakage 



