LAUSR

A robust estimate of the ENSO-Indian summer monsoon rainfall (ISMR) relationship in a changing climate is critically important for projection of the ISMR predictability. In this study, we show that the increasing nonlinear trend of ISMR and global sea surface temperatures (SST) associated with the trend of the global warming mode leads to a ‘long-term’ component to the ENSO-ISMR relationship together with a ‘short-range’ component associated with the ENSO mode. Using historical and projected simulations by some state-of-the-art CMIP5 models, we separate the global warming mode from other oscillatory modes including the ENSO mode through a combined empirical orthogonal function (CEOF) analysis, supported by a signal-to-noise maximizing EOF analysis. We find that the ‘short-term’ ENSO-ISMR relationship, represented by significant negative correlation between JJAS NINO3.4 SST and ISMR does not change appreciably with increase in green-house gas (GHG) forcing while the ‘long-term’ relationship changes from weak to significantly positive. We demonstrate that in a warming environment, the global warming mode must be taken into account together with the ‘natural’ ENSO mode in order to predict ISMR with confidence and emphasize that predictability of ISMR at any given time must be reassessed taking into account contributions from both short and long-term ENSO-ISMR relationships.