Future high-resolution El Nino/Southern Oscillation dynamics

High-resolution climate models exhibit reduced tropical Pacific mean-state biases due to better representation of ocean mesoscale processes, like tropical instability waves. With climate warming, these improved dynamics project weaker El Nino/Southern Oscillation sea surface temperature variability. The current generation of climate models does not properly resolve oceanic mesoscale processes in tropical oceans, such as tropical instability waves. The associated deficit in explicit vertical and lateral heat exchange can further contribute to large-scale equatorial temperature biases, which in turn impact the representation of the El Nino/Southern Oscillation (ENSO) and its sensitivity to greenhouse warming. Here, using a mesoscale-resolving global climate model with an improved representation of tropical climate, we show that a quadrupling of atmospheric CO2 causes a robust weakening of future simulated ENSO sea surface temperature variability. This sensitivity is caused mainly by stronger latent heat flux damping and weaker advective feedbacks. Stratification-induced weakening of tropical instability wave activity and the corresponding growth of ENSO instability partly offset the effect of other negative dynamical feedbacks. Our results demonstrate that previous lower-resolution greenhouse warming projections did not adequately simulate important ENSO-relevant ocean mesoscale processes.