The equatorial tropical Pacific, characterized by the western Pacific warm pool and eastern Pacific cold tongue, is known to modulate global tropical cyclone activities, especially through its interannual variability, the El Niño-Southern Oscillation. In response to greenhouse gas forcing, this zonal SST gradient is projected to trend toward an "El Niño-like" state with weakened Walker circulation in recent generations of CMIP models. However, observations over the past few decades show a more "La Niña-like" trend. Recent research suggests that the differences between models and observations are unlikely to be entirely due to internal variability or observational issues, and probably result at least in part from systematic model mean state biases. The resulting erroneous "El Niño-like" projection is likely to lead to errors in these models’ projections of tropical cyclones.

In this presentation, I will discuss our group’s efforts in tackling this issue. I will first illustrate a possible role of mean state biases within the models in regulating their forced responses to radiative forcing in the tropical Pacific. I will show that using flux adjustment to reduce the mean-state bias in a coupled global climate model over the tropical regions results in a more "La Niña-like" trend pattern in the tropical Pacific in both historical and near-term future under high emission scenarios. I will then discuss how global tropical cyclones respond to the "El Niño-like" and "La Niña-like" trends.