give rise to an erroneous temperature wave, the period of which would be 

 dependent on the chart interval. If the plotted values in Figure 3 had 

 act\xally resulted froia a meandering current, they could just as well be 

 interpreted as having arisen from straight currents ^ 



II. DATA EVALUATION 



Quasi- sjmoptic sea surface teoperature charts for extensive ocean 

 areas are of necessity based on injection temperature observations re- 

 ported at 6-hour intervsLLs to various meteorological organizations 1^ 

 commercial ships. 



Injection intake depths normall;'- range between 20 and 25 feet, 

 whereas bucket readings are taken near the air-water boundary. Assum- 

 ing the thermometers are accurate, injection readings will be relatively 

 low when the vertical temperature gradient is negative, relatively high 

 when the gradient is positive, and equal to bucket readings when the 

 water is uniformly mixed. 



Choice of temperature recording instruments obviously must be com- 

 patible with planned use of the data. The depth of the heated layer in 

 cold currents varies fr<Mi zero to roughly 200 feet, depending on lati- 

 tude and season. !Bierefore, again assuming the thermometers are accu- 

 rate, bucket temperatures are less suitable for outlining ocean currents 

 during summer than injection temperatures; the latter fail completely 

 between 15 and 20 degrees of latitude either side of the Equator. !Item- 

 peratures recorded by bucket thermometers in the heated layers of cold 

 currents at low latitudes may slightly exceed those of warm currents. 

 At low latitudes the current systems are apparently well outlined by 

 isotach analyses - surface speeds of water masses being transmitted from 

 below the heated layer. 



About 20 percent of injection tessperature data is estimated to con- 

 tain gross and minor errors from all causes. Saople checks of ship 

 weather logs indicate that approximately 13 percent of these errors 

 arises through processing of water temperatures for transmission in terms 

 of surface air temperature and the air temperature minus sea temperature 

 difference factor, D. Gross errors caused by incorrect difference factor 

 signs occur near coastal areas and in the vicixiity of warm and cold cur- 

 rent systems where D values may be large. These errors result in water 

 temperatures 2 x D too high over cold waters or 2 x D too low over warm 

 watersi therefore, they are an additional cause of abrupt changes in iso- 

 therm patterns. Such data errors are difficult to detect with certainty, 

 because sharp horizontal temperatxire gradients (possibly as much as 8° F 

 per mile) can be associated with oceanic currents. 



III. SOME CONSIDERATIONS OF THE MATURE OF THE SEA SORFACE 



Analyses of weather charts are facilitated by employing wind vectors 

 to show isobaric spacing and orientation. Conversely, both salinity and 



