Introduction

We now turn to the past for what can be learned about those aspects of Earth-system function and environmental change that bear most directly on the present and future. This involves critical choices about the geological timespan and the types of palaeo-evidence of greatest relevance. The geological record contains evidence from periods of even higher atmospheric CO2 concentrations and higher global temperatures than are anticipated for the coming century. At first sight, it may seem logical to turn to these periods as possible analogues for the future, but there are good reasons for rejecting any simple search for past analogues of this kind. Many Earth-system components with major functional implications were crucially different – the disposition of the continents, landmass topography, ocean circulation and plant cover at the very least. These changed characteristics imply completely different boundary conditions, forcings and feedbacks. Rather than turn to the past for analogues of state with respect to a limited number of Earth-system characteristics, we need to use the evidence from the past to address questions about amplitudes and modes of variability, process interactions and rates of change. The emphasis here is therefore on the Quaternary period, though one aspect of the pre-Quaternary geological record may prove to be of some relevance. Recent evidence (Svensen et al., 2004; Dickens, 2004) provides strong support for the hypothesis that the initial Eocene thermal maximum, some 55 million years ago, was triggered by a massive release of methane from below the Norwegian Sea.