The Andes and ENSO Statistics Correction

Abstract:

The height of the Andes has been commonly underrepresented with the smoothing in CMIP-class global climate models. This underrepresentation results in insufficient blocking for the mid-latitude westerly wind. This blocking is an important factor for the simulation of the Southeast Pacific cold tongue and will affect the ENSO variation. We have submitted a paper about how the height of the Andes affects the climate simulation accuracy in the Pacific Ocean with a slab-ocean model. However, the result from a fully-coupled model might be different because of the ocean circulation and air-sea coupling. In this study, we used the CESM 1.2.2 atmosphere-ocean coupled model and modified the height of the Andes according to the high-resolution data to see how the model result changes. In the Higher Andes case, the annual mean SST is lowered and precipitation is inhibited in the Southeast Pacific cold tongue, which eliminates model biases present in the simulation with default topography setting. The Higher Andes setting also changes the ENSO statistics. Compared with the control run model result, the Higher Andes case shows: (1) Less frequent extreme El Nino and La Nina events. (2) Less periodic signal for the ENSO cycle. (3) Higher skewness towards the positive phase. All these three changes are statistically significant and are more in line with the satellite observation. These results suggest elevating the height of the Andes can not only eliminate the long-standing model biases in the Southeast Pacific, but also improve the ENSO simulation accuracy.