摘要：Abstract There are close relationships between the sea surface temperature (SST) and the surface wind over the tropical Pacific. To study the past climate variability over the tropical Pacific, the long-term monthly wind stress anomalies over the tropical Pacific for the period of 1856-2008 are reconstructed with an SVD (singular value decomposition)-based statistical atmospheric model, where the wind stress anomalies are slave and directly correspond to the SST anomalies. The verification results show that the reconstructed wind stress data have high correlations and a small root mean square (RMS) error with the three reanalysis/simulated surface wind datasets from the last 50 years. In addition, the simulated SST anomalies from an intermediate oceanic model (IOM), which is forced by the reconstructed wind stress, can simulate the realistic interannual and decadal variability of the ENSO (El Niño-Southern Oscillation); this indicates that this new long-term wind stress dataset is useful for various climate studies, especially for the large-scale interannual and decadal variability.
关键词：reconstructed wind stress ; statistical atmospheric model ; ENSO
