Monthly temperature summaries to the near- 

 est 0.1° F. by 2-degree squares for the region 

 bounded by the west coast of the Americas to 

 longitude 120° W. and latitude 30° N. to 20° S. 

 were extracted from the historical records to 

 form 12-year (1947-58) monthly averages, 

 which are presented in Part I. Two or more 

 observations were required in any month before 

 a 2-degree summary was utilized from the 

 records. In the monthly average charts, cover- 

 ing the 12-year period, temperatures are listed 

 only when 6 or more years were represented; 

 the number in the upper right-hand corner of 

 each 2-degree square denotes the number of 

 years that were included in the average. Each 

 year' s data received equal weight. 



Monthly charts presented in Part II show 

 anomalies to the nearest 0.1° F. by 2-degree 

 squares for each month of each year for 1947- 

 58. Shaded squares indicate regions colder 

 than average, and unshaded squares indicate 

 warmer than or no deviation from the average. 



Twelve-year (1947-58) monthly means to the 

 nearest 0.1° F. for the three coastal stations 

 are presented in Part I. Monthly anomalies for 

 the coastal stations, included in Part II, are 

 based on these 12-year monthly means. Plus 

 values indicate temperatures warmer than 

 mean and minus values colder than mean. 



Because the area bounded by latitude 20° N. 

 to 30° N. and longitude 110° W. to 120° w. is 

 the general region of overlap of the temperate 

 and tropical tuna fisheries of the eastern 

 Pacific Ocean, charts of this region are pre- 

 sented here as well as in the report by 

 Johnson (1961). Minor differences between the 

 two reports that appear in the average charts 

 for this region are due to additional superficial 

 editing which occurred since the publication of 

 the report for the temperate region. Differ- 

 ences should be resolved in favor of charts 

 presented here. 



Average temperatures and anomalies for the 

 region south of the equator should be viewed 

 with caution. For the most part, adequate tem- 

 peratures were available only for squares 

 contiguous to the coast of South America. 

 Because of coastal upwelling, this is a region 

 of sharp horizontal temperature gradient. 



Deviation of the position of a ship with respect 

 to the "normal" shipping lane could cause a 

 considerable bias in the averages and 

 anomalies. Depending on ships' positions within 

 a 2-degree square, reported injection tempera- 

 tures might vary as much as 10° F. In a 

 2-degree square represented by a minimum of 

 two observations, this type of bias becomes 

 increasingly significant. 



Because of the limited number of observa- 

 tions available for the region south of the 

 Equator and the attendant possible bias de- 

 scribed above, it is recommended that the 

 temperatures and anomalies shown for that 

 area be compared with data presented by other 

 authors. 



Wooster (1961) has published yearly iso- 

 grams of average sea surface temperatures by 

 1 -degree squares from latitude 4° S. to 18° S. 

 for the years 1939-56, based on "Mapas 

 Mensuales" of the Peruvian Compania Admin- 

 istradora del Guano. 



Another report presenting temperature data 

 of the eastern tropical Pacific Ocean is that of 

 Bjerknes (1961) in a study of "El Nino" 

 conditions off the South American coast. This 

 report, however, does not provide monthly 

 temperature values but, rather, features long- 

 term seasonal averages and their corre- 

 sponding progressive changes for the years 

 1935-57. 



RELIABILITY OF THE DATA 



In order to present some indication of the 

 variability of unedited source data, confidence 

 intervals of the charted means and anomalies 

 were calculated for five selected 2-degree 

 squares. The circled areas shown in the 

 density chart (fig. 1) are the 2-degree squares 

 chosen to represent areas featuring the general 

 characteristics listed below. 



Square 1 . Cape San Lucas frontal zone - 

 sharp horizontal thermal gradient 

 and expected large temperature 

 change in time and space. 



