and September and again in December and 

 January. 



A paper entitled "On the heat and salt 

 balance of the California Current region" 

 was published in Journal of Marine Research, 

 vol. 18, no. 1, p. 36-61, 1959. 



The work was supported in part (about 50 

 percent) by the California Marine Research 

 Committee. 



METHODS IN PHYSICAL OCEANOGRAPHY: 

 EVALUATION AND DEVELOPMENT 



In the course of the STOR program, ad- 

 vantage was taken of some opportunities that 

 presented themselves to improve or evaluate 

 certain methods in physical oceanography. 



Reliability of Ocean Measurements of 

 Temperature and Salinity 



(W. S. Wooster and B. A. Taft) 



Although the validity of descriptive studies 

 of the ocean depends on the quality of data 

 employed, there have been few studies of the 

 reliability of field measurements. Therefore 

 a statistical study of such oceanographic data 

 was carried out. 



By following various pairs of thermometers 

 through many reversals, and by exannining the 

 scatter of data on temperature-depth curves 

 where observation points were closely spaced, 

 estimates were made of the measurement er- 

 rors in temperature data ( °C.) obtained in the 

 field. These estimates, expressed as standard 

 deviations for averages of two thermometer 

 readings, are as follows: 



( upper limit 0.014° - from temperature-depth curves 

 total error ( 



( lower limit 0.011° ) 



random error 

 systematic error 



0.009° ) 



) 



0.007° ) 



. from paired thermometer- 



analysis 



An upper limit of measurement error in 

 salinity data from the field, obtained from 

 an analysis of salinity -depth curves, is 0.018 /oo 

 (standard deviation). Both temperature and 

 salinity measurement errors as thus esti- 

 mated are not significantly different from 

 comparable estimates by earlier investiga- 

 tors. 



In the upper layers of the ocean other 

 errors may be significantly greater than those 

 of measurement. This is particularly true in 

 the upper few hundred meters where vertical 

 gradients of temperature and salinity are 

 usually large. Here sannpling errors due to 

 vertical oscillations, and errors resulting 

 from uncertainty in depth determination, may 



o ° 



be as large as 1 to 2 and QA/oo. At greater 



depths, where vertical gradients are small, 



such errors are comparable to measurement 



errors. 



The effect of salinity measurement errors 

 on the determination of specific volume anom- 

 aly is much greater than that of temperature 

 errors and amounts to a standard deviation 

 of about 1.5 cl./ton. The resulting standard 

 deviation of geopotential anomaly (0 over 

 1,000 db.) for a station on which standard 

 depth intervals are used is 0.004 dyn. m. 

 The measurement error of the difference 

 between two such stations is +0«011 dyn. m. 

 (two standard deviations), and at latitude 30 

 and 100 km. separation, the corresponding 

 error in geostrophic current speed is +1.5 

 cm. /sec. 



A paper entitled "On the reliability of field 

 nneasurements of temperature and salinity in 

 the ocean," was published in Journal of Ma- 

 rine Research, vol. 17, p. 552-566, 1960. 



The work was supported in part (about 75 

 percent) by the Office of Naval Research. 



Modified Roberts Current Meter (J. A. Knauas) 



The propeller-type current meter of Roberts 

 (1952) was modified by: 



1. The substitution of enlarged fiberglass 

 propellers and tail fins. 



2. The addition of a Vibrotron pressure 

 element to measure depth. 



3. The substitution of positive action wire 

 wiping contacts for the original platinum 

 contacts. Although this last modification in- 

 creased the frictional resistance of the meter 

 it was felt to be necessary in order to in- 

 crease the reliability of the contacts. 



4. The substitution of nylasint bearings for 

 the original lignum vitae. 



5. The use of a 24-volt a.c. power supply 

 which allows one to distinguish between the 

 velocity and direction contact. Power is fed 

 down a single conductor wire through the elec- 

 trical swivel into the nneter. Two silicon 

 diodes provide opposite phased currents to 

 the direction and velocity contacts. Sea water 

 is the return circuit and the electrical contact 

 is recorded on one channel of a Brush re- 

 corder. 



This instrument was designed to make sub- 

 surface nneasurements of current direction and 

 velocity, and proved very satisfactory for 

 this purpose on the Scripps Institution's ICY 

 Expeditions Dolphin and Doldrums in 1958. 



24 



