Is there an impact of resolving the stratosphere on ENSO? A first approach from EC-EARTH

The European Consortium EC-EARTH climate model version 3.1 is used to assess the effects of a well-resolved stratosphere on the representation of El Niño-Southern Oscillation, and in particular on the simulation of extreme El Niño events, known as super El Niños. Three 100-year long experiments with fixed radiative forcing representative of present climate are compared: one with the top at 0.01hPa and 91 vertical levels (HIGH-TOP or HT), another with the top at 5hPa and 62 vertical levels (LOW-TOP or LT), and another high-top experiment but with the stratosphere nudged to the climatology of HT from 10hPa upwards (NUDG). The differences in vertical resolution between HT and LT start at around 100hPa. By comparing HT with LT we explore the influence of increased vertical resolution above the tropopause on ENSO, while by comparing HT with NUDG we evaluate the influence of stratospheric variability, with special emphasis on the Quasi-Biennial Oscillation (QBO). No extreme ENSO events occur in the two simulations without QBO (LT and NUDG), while HT is able to simulate such extreme events. These super El Niños coincide with a positive Indian Ocean Dipole (IOD) and the westerly phase of the QBO in the lower stratosphere during boreal summer and fall. Previous studies have proposed an interaction between El Niño and IOD-related sea surface temperature anomalies to explain the existence of super El Niños. Our work suggests that this interaction alone is not enough in our climate model to simulate super El Niños. We postulate that changes in the upper tropospheric circulation over the Indian Ocean-Maritime Continent during boreal summer and fall, related to the westerly phase of the QBO, establish favourable conditions for the development of El Niños, increasing the probability of having super El Niños.