Near-Term Tropical Cyclone Risk and Coupled Earth System Model Biases
Most Earth system models predict that the equatorial Pacific exhibits an El Niño-like mean-state trend due to greenhouse gas effects. However, observations show a trend towards a La Niña-like pattern in recent decades. Compelling evidence emerging in recent studies shows that the equatorial Pacific response to greenhouse gas warming in these Earth system models may be incorrect. Using modeled tropical cyclones downscaled from CMIP6 via the Columbia (tropical cyclone) HAZard model as well as those directly simulated in HiResMIP model, scientists discussed how the errors in models’ projections of equatorial Pacific could lead to inaccurate projections of tropical cyclone activity in this Perspective article.
There is an ongoing debate regarding the incorrect responses of the equatorial Pacific to greenhouse gases in ESMs. However, less attention has been given to the impacts of these errors on projecting trends in extreme weather events, like tropical cyclones. While these errors may be temporary, it is important to understand the climate risks associated with these extreme events in order to adapt to climate change. Therefore, this article aims to raise awareness within the scientific community and advocates for the development of projections that consider a wider range of possible future warming scenarios for the equatorial Pacific. This includes scenarios that reflect recent historical trends, even if current coupled earth system models are unable to generate such projections at present.
Using modeled tropical cyclones downscaled from CMIP6 via the CHAZ model as well as simulated in HiResMIP model, scientists showed that the projected tropical cyclone activity is influenced by the projected sea-surface temperatures underlying storms. When atmospheric-only ESMs are forced with observed sea-surface temperatures, the modeled trend in tropical cyclone activity closely aligns with, similar to the impact of El Niño Southern Oscillation on global tropical cyclone activity at interannual time-scale, observed La Niña tropical cyclone anomalies. Following a similar analogy, erroneous El Niño-like mean-state projection in the equatorial Pacific could lead to an overestimation of cyclones in the North Pacific and an underestimation in the North Atlantic. The implications for other extreme weather events like severe storms and droughts are also discussed here. The authors and thus our project proposed potential solutions, including accurately characterizing scientific uncertainty for applications and employing methods like flux adjustment to address biases and develop alternative projections when models' historical simulations align more closely with observations.