The precise geometry and extent of Atlantic circulation changes that accompanied rapid climate changes of the last glacial period are still unknown. Here, we combine carbon isotopic records from 18 Atlantic sediment cores with numerical simulations and decompose the carbon isotopic change across a cold-to-warm transition into remineralization and circulation components. Our results show that the replacement of southern-sourced by northern-sourced water plays a dominant role below ~ 3000 m depth.

Greenland ice core records feature Dansgaard–Oeschger (D–O) events, abrupt warming episodes followed by a gradual-cooling phase during mid-glacial periods. There is uncertainty whether current climate models can effectively represent the processes that cause D–O events. Here, we propose a Marine Isotopic Stage 3 (MIS3) baseline protocol which is intended to provide modelling groups investigating D–O oscillations with a common framework.

We adopt an existing data assimilation technique to constrain a model simulation to follow three important modes of variability, the North Atlantic Oscillation, El Niño–Southern Oscillation and the Southern Annular Mode. How it compares to the observed climate is evaluated, with improvements over simulations without data assimilation found over many regions, particularly the tropics, the North Atlantic and Europe, and discrepancies with global cooling following volcanic eruptions are reconciled.

Evidence from tree rings and ice cores provides incomplete information about past volcanic eruptions and the Sun's activity. We model past climate with varying solar and volcanic scenarios and compare it to reconstructed temperature. We confirm that the Sun's influence was small and that uncertain volcanic activity can strongly influence temperature shortly after the eruption. On long timescales, independent data sources closely agree, increasing our confidence in understanding of past climate.

Palaeoclimatological reconstructions aid our understanding of current and future climate change. In the Pliensbachian (Early Jurassic) a climatic cooling event occurred globally. We show that this cooling event has a significant impact on the depositional environment of the Cardigan Bay basin but that the 405 kyr eccentricity cycle remained the dominant control on terrestrial and marine depositional processes.

Our ice core measurements show that methane, ethane, and propane concentrations are significantly elevated above their past atmospheric background for Greenland ice samples containing mineral dust. The underlying co-production process happens during the classical discrete wet extraction of air from the ice sample and affects previous reconstructions of the inter-polar difference of methane as well as methane stable isotope records derived from dust-rich Greenland ice.

The last deglaciation is a period of large warming from 21 000 to 9000 years ago, concomitant with ice sheet melting. Here, we evaluate the impact of different ice sheet reconstructions and different processes linked to their changes. Changes in bathymetry and coastlines, although not often accounted for, cannot be neglected. Ice sheet melt results in freshwater into the ocean with large effects on ocean circulation, but the timing cannot explain the observed abrupt climate changes.

Data–data and data–model vegetation comparisons are commonly based on comparing single vegetation estimates. While this approach generates good results on average, reducing pollen assemblages to single single plant functional type (PFT) or biome estimates can oversimplify the vegetation signal. We propose using a multivariate metric, the Earth mover's distance (EMD), to include more details about the vegetation structure when performing such comparisons.

Sea ice is important to stabilize the ice sheet in Antarctica. To understand how the global climate and sea ice were related in the past we looked at ancient molecules (IPSO25) from sea-ice algae and other species whose dead cells accumulated on the ocean floor over time. With chemical analyses we could reconstruct the history of sea ice and ocean temperatures of the past 14 000 years. We found out that sea ice became less as the ocean warmed, and more phytoplankton grew towards today's level.

We modelled atmospheric methane continuously from the last glacial maximum to the present using a state-of-the-art Earth system model. Our model results compare well with reconstructions from ice cores and improve our understanding of a very intriguing period of Earth system history, the deglaciation, when atmospheric methane changed quickly and strongly. Deglacial methane changes are driven by emissions from tropical wetlands, with wetlands in high northern latitudes being secondary.

In the observational period, spatially coherent sea surface temperatures characterize the northern North Atlantic at multidecadal timescales. We show that spatially non-coherent temperature patterns are seen both in further projections and a past warm climate period with a CO₂ level comparable to the future low-emission scenario. Buoyancy forcing is shown to be important for northern North Atlantic temperature patterns.