Improving ENSO in a Climate Model: Tuning vs. Flux correction

Despite improvements in simulating El Nino/Southern Oscillation (ENSO) in the last decades, current climate model still suffering in simulating important ENSO properties of ENSO, like the amplitude, the frequency, the phase locking to the annual cycle or the asymmetry between El Nino and La Nina (e.g. Bellenger et al. 2014). In a recent study of Bayr et al. (2018) the equatorial cold sea surface temperature (SST) bias in the tropical Pacific could be highlighted as one large contributor to biased ENSO dynamics in climate models. The cold equatorial SST bias, a common problem in climate models, causes a La Nina-like mean state with a too westward position of the rising branch of the Walker Circulation (by up to 30°), resulting in an erroneous convective response during ENSO events. This in turn biases the two most important atmospheric feedbacks, the positive (amplifying) wind-SST and negative (damping) heat flux-SST feedback, with error compensation between these two. Bayr et al. (2018) also give some indications how ENSO is tunable, as the study shows that it is possible to reproduce the same spread in cold SST bias and atmospheric feedback strength as seen in a CMIP5 multi-model ensemble with a perturbed physics ensemble of the Kiel Climate Model (KCM). This was achieved by changes in the cloud and convection parameters which are usually used to tune the climate models. Here we present a more detailed and systematic analyses of the influence of the tuning parameters on ENSO simulation in the KCM which employs ECHAM5 as the atmospheric component. We show the impact of the tuning parameters on the cold SST bias, the ENSO atmospheric feedbacks and important ENSO properties and demonstrate the possibility to tune ENSO and important global climate properties at the same time.