P aper 8 



logical investigation of the southeast part of the North Pacifier 

 time series plots of SST were studied for several Marsden Square 

 (MS) quadrants in order to find particularly suitable areas for 

 diagnosing changes associated with the El Nino type developments. 

 Figure 8.4 sho«s two of the quadrants selected; Ms 10(1) 

 reflects changes taking place along the equator^ and MS 10(3) 

 reflects changes in an area affected by the north equatorial 

 countercurrent . Figure 8.5 shows how the Easter-Darwin index 

 anomaly trend relates to SST anomaly trends for MS 10(1) and 

 MS 10(3) when the data are subjected to the triple 6-mo running 

 mean filter. MS 10(1) is next to the equator^ and when the 

 circulation is relatively strong (S.O. index is high) it shows 

 lower SST's due to the advection of cooler Peru current water 

 into the area and/or equatorial upwelling; when the circulation 

 is weak (index is low)/' SST's rise. MS 10(3) SST's reflect 

 changes in the north equatorial countercurrent transport 

 discussed earLy in the section. Table 8.1 shows lag correlation 

 coefficients between index and SST anomalies. Changes show up 

 about a month earlier in the index than in the MS 10(1) SST trend 

 and about three months earlier than in the MS 10(3) SST trend. 

 Events of significant magnitude are reflected in both the index 

 and SST trends of Fig. 8.5. The value of the corroborative SST 

 evidence is shown in 1961/' when the indication of a fairly deep 

 Easter -Darwi n index trough was not substantiated in the SST 

 trend. Since this trough was of much less significance in the 

 other index trends of Fig. 8.1ar an unrep resent at ive ly low Easter 

 input was the misleading factor in this case. As an interesting 

 sidelights Berla^e in 1960 put out a widely publicized forecast 

 for an El Nino in 1961 (Schweigger 1961). However/^ there was no 

 significant El Nino development in 1961. A comparison of the 

 East er -Da rwi n index anomalies with anomalies of the other indices 

 and the MS 10(1) SST anomalies would have precluded trie forecast. 



Three-month running mean plots of the index anomalies and the SST 

 anomalies for MS 10(1) (Fig. 8.6) become particularly useful to 

 the forecast process between 18 and 3 months prior to the onset 

 of El Nino. However* they should also be followed closely during 

 the course of an event to determine whether a subsequent 

 secondary trough in the index is likely or a persistent recovery 

 from the initial event is probable. 



THE 1976 EL NINO EVENT 



In summer 1975 an outlook for El Nino type activity in 1976 was 

 prepared. The outlook was given at the October 1975 Eastern 

 Pacific Oceanic Conference and at several subsequent meetings* 

 workshops* and seminars. The 12-mo running mean of the Easter- 

 Darwin index was predicted to rise from the shallow early 1975 

 trough to a small peak by the middle to latter part of 1975 and 



95 



