FISHERY BULLETIN: VOL^ 76, NO. 3 



and literature, are included in Table 1; and, the 

 typical index and index component trends as- 

 sociated with them can be noted in Figure 3a-h. 



Eguiguren's (1894) data were evaluated to ex- 

 tend the record back even further. He classified 

 rainfall at Piura, Peru (lat. 5°5'S, long. 80°38'W) 

 into five categories: dry (0), light ( 1), moderate (2l, 

 good (3l, and extra ordinary (4). Considering the 

 distribution of events over the period 1891-1976 in 

 relation to Eguiguren's rainfall category distribu- 

 tion for 1791-1890, it appeared that we could re- 

 late his category 4 to a strong El Nino, category 3 

 to a moderate El Nino, and categories 2 and 1 to 

 weak and very weak events respectively. One 

 must realize that for this Peruvian desert area 

 long-term average rainfall values have little 

 meaning, since averages combine data from the 

 more-frequent drought years with data from the 

 smaller number of event years when significant 

 rainfall may occur. A year when an average 

 amount of rain fell is likely to have been a year 

 when an event occurred. Whereas the categories 3 

 and 4 rainfall situations were likely to have been 

 associated with El Nino, there is no assurance that 

 the categories 1 and 2 rainfalls were associated 

 with the oceanic events. 



A study of presumed event occurrences (based 

 on Eguiguren's information) in relatiion to trends 

 of the Southern Oscillation indices and index 

 components for a period when overlapping data 

 records were available (1841-90) showed a high 



degree of compatibility. Table 1 lists years in ac- 

 cordance with Eguiguren's rainfall classification 

 as well as our interpretation of event intensity 

 after considering his indications, the index and 

 index component trends, and the various data and 

 information sources listed early in this section. 



After 1790 and prior to 1841, when no pressure 

 records were available for a cross-check, we avoid- 

 ed the weak event category, but accepted Eguigu- 

 ren's stronger categories 3 and 4 events. Events 

 occurring prior to 1791, as reported by Frijlinck 

 ( 1925), were considered to be of the strong variety. 

 Sources for each event are listed in Table 1. 



STATISTICAL STUDY OF 

 EL NINO TYPE EVENTS 



From a practical standpoint, we were concerned 

 with the question of when the next event is likely 

 to occur and what its intensity might be. There- 

 fore, this study referred to the onset times for 

 separate events and the interval between onset 

 times. When the year immediately following the 

 year of onset reflected an event of equal or lower 

 intensity, it was assumed that the initial event 

 extended into this next year or that the effects of 

 the initial event held over into the early part of the 

 next year; and, the whole situation was treated as 

 a single event. However, when a weaker initial 

 event preceded a stronger event in a following 

 year they were treated as two separate events, 



Table l. — Year of onset of El Nino type events, 1726-1976, as classified according to event intensity by Eguiguren ( 1894), left, and the 

 present authors, right (events below intensity 3 were not accepted prior to 1841 when pressure data became available). Numbers refer 

 to event intensity: 1, very weak; 2, weak; 3, moderate; and 4, strong. Asterisks indicate onset of events considered separate (see text). 



672 



