Search

Despite recent advances, the link between the evolution of atmospheric CO2 and climate during the Eocene greenhouse remains uncertain. In particular, modelling studies suggest that in order to achieve the global warmth that characterised the early Eocene, warmer climates must be more sensitive to CO2 forcing than colder climates. Here, we test this assertion in the geological record by combining a new high-resolution boron isotope-based CO2 record with novel estimates of Global Mean Temperature. We find that Equilibrium Climate Sensitivity (ECS) was indeed higher during the warmest intervals of the Eocene, agreeing well with recent model simulations, and declined through the Eocene as global climate cooled. These observations indicate that the canonical IPCC range of ECS (1.5 to 4.5 °C per doubling) is unlikely to be appropriate for high-CO2 warm climates of the past, and the state dependency of ECS may play an increasingly important role in determining the state of future climate as the Earth continues to warm.

The ultimate fate of (fossil fuel) CO2 emitted to the atmosphere is governed by a range of sedimentological and geological processes operating on timescales of up to the ca. hundred thousand year response of the silicate weathering feedback. However, how the various geological CO2 sinks might saturate and feedbacks weaken in response to increasing total emissions is poorly known. Here we explore the relative importance and timescales of these processes using a 3-D ocean-based Earth system model. We first generate an ensemble of 1 Myr duration CO2 decay curves spanning cumulative emissions of up to 20,000 Pg C. To aid characterization and understanding of the model response to increasing emission size, we then generate an impulse response function description for the long-term fate of CO2 in the model. In terms of the process of carbonate weathering and burial, our analysis is consistent with a progressively increasing fraction of total emissions that are removed from the atmosphere as emissions increase, due to the ocean carbon sink becoming saturated, together with a lengthening of the timescale of removal from the atmosphere. However, we find that in our model the ultimate CO2 sink - silicate weathering feedback - is approximately invariant with respect to cumulative emissions, both in terms of its importance (it removes the remaining excess ~7% of total emissions from the atmosphere) and timescale (~270 kyr). Because a simple pulse-response description leads to initially large predictive errors for a realistic time-varying carbon release, we also develop a convolution-based description of atmospheric CO2 decay which can be used as a simple and efficient means of making long-term carbon cycle perturbation projections. Key Points An ensemble of CO2 pulse emissions are modeled using an Earth system model Our impulse response function projects the atmospheric lifetime of emitted CO2 We characterize how the marine CO2 sinks tend to saturate at very high emissions.

This Discussion Meeting Issue of the Philosophical Transactions A had its genesis in a Discussion Meeting of the Royal Society which took place on 10-11 October 2011. The Discussion Meeting, entitled 'Warm climates of the past: a lesson for the future?', brought together 16 eminent international speakers from the field of palaeoclimate, and was attended by over 280 scientists and members of the public. Many of the speakers have contributed to the papers compiled in this Discussion Meeting Issue. The papers summarize the talks at the meeting, and present further or related work. This Discussion Meeting Issue asks to what extent information gleaned from the study of past climates can aid our understanding of future climate change. Climate change is currently an issue at the forefront of environmental science, and also has important sociological and political implications. Most future predictions are carried out by complex numerical models; however, these models cannot be rigorously tested for scenarios outside of the modern, without making use of past climate data. Furthermore, past climate data can inform our understanding of how the Earth system operates, and can provide important contextual information related to environmental change. All past time periods can be useful in this context; here, we focus on past climates that were warmer than the modern climate, as these are likely to be the most similar to the future. This introductory paper is not meant as a comprehensive overview of all work in this field. Instead, it gives an introduction to the important issues therein, using the papers in this Discussion Meeting Issue, and other works from all the Discussion Meeting speakers, as exemplars of the various ways in which past climates can inform projections of future climate. Furthermore, we present new work that uses a palaeo constraint to quantitatively inform projections of future equilibrium ice sheet change.