Realism of local and remote feedbacks on tropical sea surface temperatures in climate models

[1] An important emerging issue in climate research is the degree to which a sea surface temperature (SST) change in one tropical ocean basin affects the SST in other basins. In this study, the SST interactions among eight broadly defined regions of coherent SST variability in the tropical Pacific, Indian, and Atlantic oceans are estimated using 3 observational and 76 climate model simulation data sets of the 20th century. The eight-dimensional SST feedback matrix is estimated separately using each data set by constructing a Linear Inverse Model based on the lag-covariance statistics of the 100 year monthly SST time series. The simulated feedback matrices are found to differ in several key respects from the observed matrices and also from one another. In particular, the influence of the eastern Pacific ENSO region on other regions and of the other regions on the ENSO region is found to vary considerably from model to model. The representation of remote interactions with the Indo-Pacific Warm Pool region is also found to be highly variable. It is argued that these large errors/differences arise mainly from differences in the representation of the remote atmospheric teleconnective feedbacks, and to a lesser extent the local radiative-thermodynamic feedbacks, on the SSTs in the models, whereas differences in the representation of the tropical oceanic wave dynamics are likely less important.