The Andes Affect ENSO Statistics

Current coupled global climate models have biases in their simulations of the tropical Pacific mean-state conditions as well as the El Nino-Southern Oscillation (ENSO) phenomenon. Specifically, in the Community Earth System Model (CESM version 1.2.2), the tropical Pacific mean state has overly weak sea surface temperature (SST) gradients in both the zonal and meridional directions, ENSO is too strong and too regular, and El Nino and La Nina events are too symmetrical. A previous study with a slab-ocean model showed that a higher elevation of the Andes can improve the tropical Pacific mean-state simulation by adjusting the atmospheric circulation and increasing the east-west and north-south SST gradients. Motivated by the link between the mean tropical Pacific climate and ENSO variations shown in previous studies, here we explored the influence of the Andes on the simulation of ENSO using the CESM 1.2.2 under full atmosphere-ocean coupling. In addition to improving the simulated tropical Pacific mean state by increasing the strength of the surface easterly and cross-equatorial southerly winds, the Higher Andes experiment decreases the amplitude of ENSO, increases the phase asymmetry, and makes ENSO events less regular, resulting in a simulated ENSO that is more consistent with observations. The weaker ENSO cycle is related to stronger damping in the Higher Andes experiment according to an analysis of the Bjerknes index. Our overall results suggest that increasing the height of the Andes reduces biases in the mean state and improves the representation of ENSO in the tropical Pacific.

Automated summary

Current global climate models have biases in simulating the mean state and ENSO phenomenon in the tropical Pacific. Previous studies have shown that increasing the elevation of the Andes can improve the simulation of the mean state. In this study, we explored the influence of the Andes on ENSO simulation and found that higher Andes elevations can decrease the amplitude of ENSO, making it more consistent with observations.