Table 2 is simplified by only listing values for stations where the correlation magnitude exceeds 0.5. Shading is employed in the table to draw attention to the strongest relationships. From this table one can identify three Andean subregions with different climatic characteristics. The correlation coefficients between the precipitation indices in the Andean region and the SST anomaly indices for El Niño region 1+2 are moderately positive (the precipitation tends to be greater than the average during El Niño) in the north end of the northern subregion during November-March. The correlation coefficients are neutral in the central subregion, but moderately positive in November, and are slightly negative (the precipitation tends to be smaller than average during El Niños) in the southern subregion, but moderately negative in February.
The correlation coefficients between the precipitation indices in the Andeas and the SST anomaly indices for El Niño region 3.4 are weakly positive in the northern subregion, neutral in the central subregion and moderately negative in the southern subregion, particularly in February. The correlation coefficients between the precipitation indices and the SST anomaly indices for El Niño 4 region are very small in the north and moderately negative in the central and south subregions, particularly in February.
The correlation analysis indicates that precipitation in the northern coast is strongly related to the SST anomaly indices for El Niño region 1+2, mainly in the period January-March. During El Niño events, as defined by SCOR, the relationship intensifies, confirming the results of previous studies (Woodman, 1999, [and numerous others]). The correlation coefficients between the precipitation indices in the Peruvian Amazon region and the SST anomaly indices for the four El Niño regions are small in magnitude.
Figure 3 depicts the correlation between the precipitation indices and SST anomales for the El Niño 1+2 region for the calendar months October to March. Figures 4 and 5 show the results of the correlation between precipitation indices and SST anomaly indices for El Niño 3.4 and El Niño 4 regions, respectively. In these figures the magnitudes of the correlation coefficients are indicated by the size of the open and shaded circles. Stronger positive correlations, represented as large shaded circles, are shown in Figure 3 for the northern coast from November to March. Similarly, in Figure 4 moderate positive correlations are observed in the northern coast from November to March, and in Figure 5 moderate positive correlations are observed along the Andean region in November. Moderate negative correlations with both El Niño 3.4 and El Niño 4 are observed along the Andean region in January, February and March (Figures 4 and 5).
[November is characterized by positive correlations with respect to all SST indices at all latitudes of the Andeas. It’s there for one month and then goes away. What is so special about November? Is this just due to one ENSO?]
[Do you have the possibility of using other colors in Figures 3-5? The tendency for precipitation has become to use green for wet (positive correlations) and brown for dry (negative correlations).
Table 3 shows the lagged correlation coefficients between SST anomaly indices in El Niño 3.4 and El Niño 1+2 regions for six calendar months for the years 1950-2006. Each row is the correlation between Nino 3.4 in a specific calendar month and Nino 1+2 for months preceding, simultaneous to, or following that month. The month of largest correlation magnitude in each row is shaded. Positive lags are months where Nino 1+2 follows Nino 3.4. The results indicate that SST anomalies in the El Niño 3.4 region in October and November could be used to predict December SST anomalies in the El Niño 1+2 region. [Why not try a table organization where the ordinate is the Nino 3.4 month and the abscissa is the Nino 1+2 month? That would make it much