Potential seasonal predictability of the observed Euro‐Atlantic atmospheric variability using SST forced ECHAM4‐T42 simulations

摘要: To estimate the potential seasonal predictability of Euro-Atlantic atmospheric variability, canonical correlation analysis is used, and a comparison made between NCEP reanalysis with ensemble simulations ECHAM4-T42 forced observed sea surface temperature (SST) sea-ice boundaries for 1951–94 changing CO2 concentration. The method identifies those modes variability that have similar temporal evolution in observations mean simulations. Signals due to long-term changes forcing were first reduced by removing third-order polynomial from all data. Significant covariability 500 hPa geopotential height over region found autumn spring. best agreement patterns seen late winter, where mixed Pacific–North America (PNA) tropical/northern-hemisphere teleconnection pattern dominant signal. Although it tends modulate North Atlantic Oscillation (NAO), direct influence Europe limited. In cases, seems be remote tropical Pacific SST. However, model response El Nino Southern (ENSO) always shows PNA pattern, while related signal undergoes large changes. The level pressure (SLP) much less sensitive ENSO forcing, although traces its can detected winter. On other hand, highly significant modelled SLP anomalies autumn, reflecting SST on NAO. does not reproduce structure. It also shown reproducibility NAO undetrended ECHAM4 winter was Latif et al. same trends Indian Ocean variability. origin cannot determined as concentration specified Copyright © 2003 Royal Meteorological Society