Response Analysis: Spatial Sensitivity

 

Response analyses assigned values equal in magnitude, but opposite in sign, to those areas in the heart of each dipole region. Figures 6a and 6b depict the spatial extent of these high response regions and show the associated response to a ±2s in the NAO is in the range of ±4šC and ±200 mm/winter. Northern Europe and Scandinavia show cooling and drying during a negative NAO year while, western Greenland, the Mediterranean, and Turkey, show warming and increased precipitation. Even the smallest variations in Turkish precipitation are associated with large changes in water supply to the entire Near East. Average total winter precipitation in Turkey is 330 mm/winter (annual average is 650 mm/year). NAO-related changes are on the order of ±100 mm/winter, in which case NAO-related precipitation changes could conceivably account for a 30% variation in winter precipitation. Gleick (1994) estimates that variations on the order of 10% can push many Near East countries below the defined 500 m3 of water/person/year defined by hydrologists as a basic minimum for drinking, sanitation, and industrial needs.

Criteria for selecting the centers was based upon similar level of correlation among the 5 and a physically grounded connection linking each center to the NAO. The North Sea, Labrador Sea, and Sargasso regions were identified by Kushnir (1994) as dominant "centers of action" and explain 70% of the total variance, the Eastern Mediterranean and equatorial region were selected based on significant correlation. This selection of "centers of action" based on sensitivity to the NAOSLP index, results in an NAOSST index that is highly correlated with the NAOSLP index. The correlation between NAOSLP index and the generated NAOSST index is 0.67. The NAOSLP-SST correlation coefficient is roughly equivalent to the correlation between the SLP-based Southern Oscillation Index and the SST-based NINO3 index which is 0.65 (Chiang, 1997) .