Resumen
Models of dichotomic and polytomic logistic regression are applied to series of historical records spanning one and a half centuries (1840-1990) of the North Atlantic Seesaw (NAS) and El Niño (EN), looking for simultaneous occurrence of both phenomena during the northern winter months and classifying NAS in four modalities and EN in three intensities. NAS refers to the temperature anomalies in Greenland and Norway; its modalities are: both positive (BA), both negative (BB), Greenland positive with Norway negative (GA) and Greenland negative with Norway positive (GB). A bigger incidence of NAS is found as winter progresses, as well as simultaneity of occurrence of 61% between the NAS and EN. NAS heterogeneous modalities (GA and GB) coincide in 71% with EN events, while homogeneous modalities (BA and BB) do so in 29%. Heterogeneous modalities have a higher coincidence with EN greatest intensity (3). When the frequencies computed through the logistic polytomic model were compared to the observed frequencies, a very close goodness of fit was found, when the whole winter season was considered.