The pattern effect and its implications for climate sensitivity

Abstract

Radiative feedbacks depend on the spatial pattern of sea-surface temperatures (SSTs) and thus can change over time as SST patterns evolve – the so-called pattern effect. The radiative feedbacks within atmospheric general circulation models (AGCMs) forced by historical observed SST patterns are generally more negative than those within fully-coupled GCMs forced with increasing CO2, yielding to a lower value of equilibrium climate sensitivity (ECS) estimates from historical energy budget constraints. The anomalously negative radiative feedbacks are traced to the observed SST pattern over recent decades, featuring enhanced warming the tropical western Pacific and broad cooling in the tropical eastern Pacific and part of the Southern Ocean, which tends to strengthen lower tropospheric stability and enhance tropical low cloud cover. In Chapter 1, I first describe the global-mean energy budget framework underlying the modern estimates of ECS and radiative feedbacks, and common approaches to quantify the global energy budget in GCMs and in observations. Then I introduce the pattern effect and its implications for ECS estimates. Last, I summarize the overarching research questions that motivate this thesis. In Chapter 2, I examine inter-model differences in the magnitude of the pattern effect, the source of their differences, and how these differences contribute to the spread in ECS within CMIP5 and CMIP6 models. The results show that the inter-model spread in ECS is dominated by the spread in radiative response on fast timescales (mainly reflecting different atmospheric physics) rather than by the spread in feedback evolution over time (governed by evolving SST patterns). In Chapter 3, I assess ECS and radiative feedbacks derived from historical energy budget constraints within historical simulations of CMIP6 GCMs, and compare them with those from observed energy budget constraints and those from long-term GCM simulations forced by CO2 quadrupling. The results show that historical energy budget constraints generally underestimate values of ECS from CO2 quadrupling, owing to the pattern effect. The observation-based ECS estimates are even lower than those from GCM historical simulations, due to discrepancies between modeled and observed historical SST patterns. In Chapter 4, I explore the causes of the observed tropical Pacific SST pattern, and particularly focus on linking the tropical Pacific SSTs with the observed Southern Ocean climate change. I find that there is a two-way teleconnection between the tropical Pacific and the Southern Ocean via atmospheric pathways, associated with regional atmospheric circulations instead of zonal-mean energetic constraints as previously proposed. The results highlight potential impacts on the tropics from the extratropical climate changes that have occurred over the instrumental record. In Chapter 5, I discuss implications of this study and open research questions inferred from the results of this study.