Conference Abstracts

The Atlantic meridional overturning circulation is important for global and regional climate due to the role it plays in the uptake and redistribution of climatically relevant tracers such as heat and carbon. Coupled climate models project that it will weaken during the 21st century due to increased heat and freshwater fluxes into the critical high-latitude North Atlantic sinking regions. Observations at 26N suggest that it has indeed declined since measurements began in 2004, although we do not know whether this trend reflects natural multi-decadal variability or a secular anthropogenically forced weakening. Sea surface temperature proxies suggest a longer-term decline. In this session we will discuss a wide range of studies focused on constraining and understanding the past, present and future of the AMOC. We will look at the evidence from palaeoceanographic proxies for changes in deep ocean circulation associated with deglacial climate shifts. We will consider the role of continental geometry in determining the AMOC’s strength and structure. We will assess the role of wind forcing in watermass transformation east of Greenland. And we will explore the close two-way links between the AMOC and winter Arctic sea-ice variability and change. A decade on from the inaugural ACDC summer school which focused on this topic, there remain many open questions about the fundamental drivers of the AMOC and its variability on long timescales.

J. McManus*, K. Costa, Y. Zhou, H. C. Ng, S. Hoffmann, L. Robinson, D. Oppo, L. Keigwin

The large-scale subsurface circulation of the ocean is an important component of the Earth’s climate system, and contributes to the global and regional transport of heat and mass, including carbon, nutrients and fresh water. Assessing how this system has changed in the past is thus a priority for understanding natural climate variability. Here we combine published and new time-series of U, Th, and 231Pa/230Th, along with δ13C, εNd and other proxy data from multiple North Atlantic deep-sea sediment cores to explore the robustness and timing of inferred changes in deep ocean circulation associated with deglacial climate shifts. These data reveal a series of deep ocean changes associated with cooling and warming during the last deglaciation. We find a coherent pattern of dramatic circulation changes in the deep western basin of the North Atlantic at times of millennial climate change, and contrasting evidence from the low-latitude eastern basin. A transect of sites from east to west across the mid-latitude subtropics provides evidence suggesting that episodes of weakening of the deep circulation proceeded from the abyss upward, while the subsequent strengthening was initiated higher in the water column and propagated to depth, consistent with the previously inferred role of changes in the northern sources of deep water.

Jerry McManus [#13]

S. Ragen*

The shape of ocean basins and the pattern of heat transport in the oceans plays a large role in the global climate. The width of an ocean basin and the latitudinal extent of the adjacent continents influence the overturning circulation. How much does the shape of a coastline influence meridional overturning? Interestingly, while the Atlantic Meridional Overturning Circulation (AMOC) is studied in detail, debate remains about the nature of the fundamental drivers of AMOC. I will use an ocean only model (MOM6) to run aquaplanet simulations with idealized continents to determine the most important factors driving overturning circulation in the Atlantic. I will spin up an Aquaplanet, a configuration with two basins (one narrow and one wide) separated by grid cell-wide ridges, and two closed-basin experiments, one with two basins separated by continents with straight coastlines and one with a more realistic coastline along the narrow basin. The closed-basin experiments do not allow for exchange flow between the two ocean basins. This will test whether the shape of the coastline around the Atlantic shapes overturning circulation and heat transport in the ocean. If AMOC persists under zonal forcing in the realistic coastline closed-basin simulation, then the route of inflow of water from the Indo-Pacific and the salt-advection feedback matter less than or equal to the shape of the boundaries to zonal flow.

Sarah Ragen [#18]

M. Årthun*, T. Eldevik, L. H. Smedsrud

During recent decades Arctic sea ice variability and retreat during winter have largely been a result of variable ocean heat transport. The relationship between ocean heat and sea ice anomalies has allowed for skillful predictions of winter sea ice extent, especially in the Barents Sea. Here we use the Community Earth System Model (CESM) large ensemble simulation to disentangle internally and externally forced winter Arctic sea ice variability, and to assess to what extent future winter sea ice variability and trends are driven by Atlantic heat transport. We find that in a warming world (RCP8.5), interannual to decadal winter sea ice variability is predominately driven by internal variability, whereas external variability is more important for multi-decadal sea ice trends. Ocean heat transport into the Barents Sea is a major source of internal Arctic winter sea ice variability today and in the future, and, as a consequence, ocean heat transport remains a good predictor of winter sea ice variability within this century. The warm Atlantic water gradually spreads downstream from the Barents Sea and further into the Arctic Ocean, leading to a reduced sea ice cover and substantial changes in sea ice thickness. The future long-term increase in Atlantic heat transport is carried by warmer water as the current itself is found to weaken.

Marius Årthun [#10]

K. Våge*, L. Papritz, L. Håvik, M. Spall, K. Moore

Warm subtropical-origin Atlantic water flows, as an extension of the Gulf Stream, northward across the Greenland-Scotland Ridge into the Nordic Seas where it relinquishes heat to the atmosphere and gradually transforms into dense Atlantic-origin water while circulating around the basin. Before returning southward as part of the East Greenland Current, a major contributor of dense water to the Denmark Strait overflow, the Atlantic-origin water subducts beneath a layer of cold, fresh surface water and sea ice. Here we show, using measurements from autonomous ocean gliders deployed from fall 2015 to spring 2016, that the Atlantic-origin water was re-ventilated while transiting the western Iceland Sea in winter. This re-ventilation is a recent phenomenon made possible by the retreat of the ice edge toward Greenland, which had previously insulated the waters of the East Greenland Current from the atmosphere. The fresh surface layer that characterizes this region in summer is diverted toward the Greenland shelf by enhanced onshore Ekman transport induced by stronger northerly winds in fall and winter. Severe heat loss from the ocean to the atmosphere offshore of the ice edge subsequently triggers convection which further transforms the Atlantic-origin water and may impact the properties of the Denmark Strait overflow. This re-ventilation is a counterintuitive occurrence in a warming climate, and highlights the difficulties inherent in predicting the behaviour of the complex coupled climate system.

Kjetil Våge [#09]

H. L. Johnson*, Y. Kostov, D. P. Marshall

Recent observations of the meridional overturning circulation in density space, made by the international Overturning in the Subpolar North Atlantic Project (OSNAP, www.o-snap.org), show that both the mean and the variability of the overturning are dominated by the component to the east of Greenland. This suggests that watermass transformation in the Nordic Seas dominates over that in the Labrador Sea. We use the ECCO configuration of the MIT general circulation model and its adjoint to explore the sensitivity of the overturning east of Greenland to surface forcing and properties. Our results show that, in contrast to the overturning across the RAPID array at 26.5N, variability in subpolar overturning is not dominated by wind forcing, which explains less than half of the variance in the ECCO 20 year simulation. Buoyancy forcing, particularly to the south of the array over the separated North Atlantic current, also plays a role. We attempt to reconstruct the variability observed by the OSNAP array, using our adjoint-based linear sensitivities and surface properties, and hence gain some insight into the forcing mechanisms responsible.

Helen Johnson [#09]

J. Rheinlænder*, D. Ferreira, K. H. Nisanciouglu

Changes in the geometry of ocean basins have been influential in driving climate change throughout Earth’s history. Here we focus on the appearance of the Greenland-Scotland Ridge (GSR) and try to understand its impact on the ocean state, including global circulation, heat transport, T and S properties and ventilation timescales, which will be useful for interpreting paleoproxies.To this end, we use a coupled atmosphere-ocean-sea ice model with idealized geometry and consider two geometrical configurations. The reference configuration (noridge) comprises two wide strips of land set 90° apart extending from the North Pole to 40°S, separating the Northern Hemisphere ocean into a small «Atlantic-like» and a large «Pacific-like» basin. In the ridge configuration a zonally symmetric oceanic ridge, that extends across the Atlantic-like basin at 60°N, mimicking the GSR, is added. In addition, we consider two climatic limits of noridge: a warm case where the northern high latitudes are ice-free and a cold case where a seasonal sea ice cover is present. In both cases of noridge deep-water formation occurs at the North Pole in the Atlantic-like basin. When the ridge is introduced, the flow of warm Atlantic water to the high latitudes is hampered and the ocean heat transport across 70°N decreases by ~36% which causes cooling and freshening of the water column north of the ridge. Downwelling shifts south of the ridge, thereby altering the structure of the upper overturning cell dramatically. Despite these changes, the Northern Hemisphere surface climate response is surprisingly small for both cold and warm cases. Our results highlight the possible disconnect between changes in the localization of deep-water formation, the structure of the AMOC and the properties of water masses and changes in Northern Hemisphere surface climate.

Jonathan Rheinlænder [#16]

The link between Earth’s land ice and the ocean is a complex two-way connection, including short-term and long-term feedbacks and local to global impacts. In this session, we will begin with a peek into the significant scientific progress since 2010 and then look at current research on ice-ocean processes from pole to pole, and from the deep past to present day. Beginning in the present day, instruments on Antarctic ice shelves are helping to identify distant wave sources, measure sea ice thickness, understand how sea ice dampens wave action, and quantify wave forcing on ice shelves. Looking at processes off of the ice, increasing fresh glacial meltwater in Antarctica is reducing the formation of Antarctic Bottom Water, with warmer deep waters also remaining available to increase basal melt on Antarctic ice shelves, potentially increasing loss of grounded ice. In Greenland, icebergs act as important sources of freshwater, and the location and timing of iceberg melt is unique among freshwater sources, with much melt produced – and likely remaining – at depth. While the freshwater from ice sheets changes ocean temperature and salinity, changing ocean volumes also influence ice sheets. With sea level rise being produced from both Northern and Southern Hemispheres, the fingerprint of regional sea level rise is complex. Because sea level rise resulting from ice loss is of highest magnitude in the far-field, changing ocean levels create an interhemispheric teleconnection among large ice sheets. This process likely led to higher Antarctic ice volume peak during the Last Glacial Maximum and enhanced retreat through the Last Deglaciation. The sensitivity of the Antarctic Ice Sheet to sea level rise and evolution of the last glaciation is also being explored in a new intermediate complexity Earth System Model. Finally, a coupled glaciological and visco-elastic glacial isostatic adjustment model will provide additional detail about how the Greenland Ice Sheet formed and its sensitivity to Earth’s viscosity.

P. Heimbach*

If time permits, and if none of the former students is taking on this task, I will give my personal, subjective perspective on what has been learned since the 2010 ACDC session on ice sheet-ocean interactions.

Patrick Heimbach [#all]

M. C. Hell*, S. T. Gille, B. Cornelle, A. Miller, P. Bromirski, A. D. Crawford

Surface winds from Southern Ocean cyclones generate large waves that travel over long distances. Their spectra contain characteristic information about the wind speed, fetch size and storm intensity. Two years of seismic observations from the Ross Ice shelf, combined with modern optimization (machine learning) techniques are used to trace the origins of wave events in the Southern Ocean with an accuracy of 110km and 2h from a hypothetical point source. The observed spectra attenuate within sea ice and in the ice shelf, but retain characteristics that can be compared to parametric wave models.Comparing the MERRA2 and ERA5 reanalysis suggest that about 60% of the observed wave events cannot be matched with Southern Ocean high wind events. The reanalysis cyclones and winds are often miss placed by about 450 km/7 hours in MERRA2 or 360km/5hours in ERA5 compared to the most likely position inferred from the seismic data.In a second step these observations are used to measure critical sea-ice properties, such as thickness and wave attenuation, as well as forces on the ice shelf due ocean swell. The seismic observations and open ocean swell spectra inform a 1-dimensional sea ice-wave model that is optimized for a simplified sea-ice thickness distribution along the wave path. This "wave topomography” method can potentially improve sea ice thickness models around Antarctica, improve wave attenuation models under thin and thick sea ice, and may even set wave forcing at the ice shelf front into context with other forces.

Momme Hell [#17]

A. Silvano*, S. Rintoul, B. Peña-Molino, W. Hobbs, E. van Wijk, S. Aoki, T. Tamura, G. Williams

Strong heat loss and brine release during sea ice formation in coastal polynyas act to cool and salinify waters on the Antarctic continental shelf. Polynya activity thus both limits the ocean heat flux to the Antarctic Ice Sheet and promotes formation of Dense Shelf Water (DSW), the precursor to Antarctic Bottom Water. However, despite the presence of strong polynyas, DSW is not formed on the Sabrina Coast in East Antarctica and in the Amundsen Sea in West Antarctica. Using a simple ocean model driven by observed forcing, we show that freshwater input from basal melt of ice shelves partially offsets the salt flux by sea ice formation in polynyas found in both regions, preventing full-depth convection and formation of DSW. In the absence of deep convection, warm water that reaches the continental shelf in the bottom layer does not lose much heat to the atmosphere and is thus available to drive the rapid basal melt observed at the Totten Ice Shelf on the Sabrina Coast and at the Dotson and Getz ice shelves in the Amundsen Sea. Our results suggest that increased glacial meltwater input in a warming climate will both reduce Antarctic Bottom Water formation and trigger increased mass loss from the Antarctic Ice Sheet, with consequences for the global overturning circulation and sea level rise.

Alessandro Silvano [#18]

T. Moon*, D. Sutherland, E. Enderlin, D. Carroll, D. Felikson, L. Kehrl, F. Straneo

Freshwater flux from Greenland Ice Sheet mass loss raises global sea level, influences large-scale ocean circulation and stratification, and affects biological systems. Freshwater flux in glacial fjords comes from several sources: ice sheet surface melt discharged subglacially at the glacial termini, terrestrial runoff, submarine terminus melt, and melt from icebergs throughout the fjord (here, including icebergs, bergy bits, and melánge). Melt from icebergs has proven a particularly difficult metric for assessing and incorporating into ocean models. Here, we report on a suite of research that is illuminating the role of icebergs and iceberg melt in the Greenland coastal region. Results include clear evidence that iceberg melt can dominate the annual freshwater flux budget in fjords with active iceberg-producing glaciers. Since iceberg melt is most responsive to ocean conditions, spikes in iceberg melt are also offset from other freshwater sources that are more closely link with atmospheric conditions, creating a distinct temporal footprint of iceberg melt. The footprint of iceberg melt is also spatially distinct, with production occurring in fjord and open waters, and across the full iceberg depth. Along with improved identification of iceberg distributions, new methods are pinpointing the location and magnitude of iceberg melt at depth. These results show that substantial iceberg melt production occurs at depth and that fjord properties commonly prevent that freshwater from moving into the surface layers. This is in strong juxtaposition to standard ocean modeling practices, which commonly ignore iceberg melt or input all freshwater into the ocean surface layer. Together, this research provides a more detailed understanding of the strong role of icebergs as freshwater producers and highlights pathways to improve freshwater budgets and correctly incorporate Greenland freshwater production into ocean models.

Twila Moon [#10]

N. Gomez*, M. Weber

Evolution of the Antarctic ice sheet is influenced by sea level at its edge. A sea level fall in the vicinity of a marine ice sheet decreases the ice loss across the grounding line, which in turn can slow grounding line retreat or initiate grounding line advance; a sea level rise, on the other hand, can slow grounding line advance or enhance retreat. Over the Last Deglaciation, global sea level changes were controlled by variations of the global ice sheets. In the vicinity of a retreating ice sheet, local gravity-driven draw down of the sea surface and uplift of the solid Earth lead to a sea level fall that is more than an order of magnitude larger than the sea level rise that occurs farther away. The possibility of inter-hemispheric teleconnections through sea level may arise when the ice loss/gain in one hemisphere is substantially larger than in the other. This was the case during the last glacial cycle, when variations of the Northern Hemisphere (NH) ice sheets contributed more to global mean sea level change than the Antarctic ice sheet. We apply a global, coupled ice sheet – sea level model to explore the relative contributions of Northern and Southern Hemisphere ice to sea level in Antarctica and their influence on ice dynamics leading up to and during the Last Deglaciation. We compare simulations in which NH ice cover is fixed throughout the simulation to simulations that include the contribution to sea level in Antarctica from the NH. Our results show that sea level changes across Antarctica associated with NH ice variations lead to a higher peak Antarctic ice volume at the Last Glacial Maximum and more and earlier retreat through the Last Deglaciation, with timing consistent with Antarctic iceberg rafted debris records.

Natalya Gomez [#10, 18]

L. Tarasov*, A. Long, G. Milne, D. Roberts, S. Woodroffe, S. Funder, K. Kjellerup Kjeldsen, B. Lecavalier

Do the details of the earth viscosity model matter for ice sheet evolution and there inferences thereof? To date, all published data-constrained deglacial reconstructions either assume an earth model viscosity profiled derived independently or sweep over earth models using the identical ice sheet history (usual with an history derived using a GIA model that lacked the detailed visco-elastic physics). The litterature has made clear that ice sheet evolution is sensitivity to the physics included in the GIA model (e.g. simple relaxation bedrock response versus full visco-elastic) but to date the sensitivity of ice sheet evolution to the viscosity structure in a coupled ice and visco-elastic GIA model has not been documented.I will present the first joint calibration of a coupled glaciological and visco-elastic GIA model for the last glacial cycle evolution of the Greenland ice sheet (including revised bounds on its contribution to the Eemian high-stand). I will also document the sensitivity of the ice sheet evolution in the coupled model to the earth viscosity structure.

Lev Tarasov [#13,16]

M. Geng*, L. Tarasov, T. Bahadory

The LCice v1.0 (coupled Atmosphere-Ocean-Vegetation model LOVECLIM to the Glacial System Model (GSM) for the northern hemisphere ice sheets) has been used in an ensemble simulation of the last glacial inception. The updated LCice (v1.1) includes the Antarctic and Patagonian ice sheets, and expanded Eurasian ice sheet to cover the Siberia and Tibet, besides the North American and Greenland ice sheets. In this study, we compare the climate evolution of the last glacial inception with all the major ice sheets included, with the LCice v1.0 results, and previous studies. The sensitivity of the Antarctic ice sheet, especially to sea level changes introduced by the variations of the northern hemisphere ice sheets are analyzed.

Marilena Geng [#18]

The transition from the Last Ice Age to the present warm Interglacial –the last deglaciation (~18,000-11,000 years ago)– was a critical period of climate shifts during which every component of the Earth’s system underwent numerous abrupt and rapid large-scale changes. During the last deglaciation, huge Northern Hemisphere (NH) ice sheets melted away, sea level rose about 120 m, atmospheric CO₂ increased by about 100 ppm, and the Earth experienced a mean global warming of about 3.5 ˚C. In particular, the waning of extensive NH ice sheets to about their present volume represents one of the largest and most rapid natural climatic changes in the planet’s history. Therefore, not only the last deglaciation provides an ideal natural laboratory to decipher the mechanisms behind rapid climate change, but it may also help characterise the impact of melting ice sheets on the climate system in response to global warming.

This session will explore different aspects of the interplay between cryosphere, ocean and atmosphere at regional scales during the last deglaciation, with a special focus on ice sheet dynamics, stability and melting histories.

B. Markle*, E. Steig, G. Roe, Q. Ding

The Antarctic is an important region for the Earth's energy balance and climate. Here, we investigate the spatial patterns of Antarctic temperature change across a range of timescales. We use multiple ice-core water isotope records and an improved method for inverting water-isotope ratios to moisture-source and precipitation-site temperatures. We find that between the Last Glacial Maximum and the Holocene, East Antarctic warming was less than previously thought, and less than concurrent warming in West Antarctica; the latter point is confirmed by independent temperature estimates. More broadly, we identify a pattern of temperature variability in which lower, warmer sites (such as in West Antarctica) show more temperature change than higher, colder sites (such as in East Antarctica). This spatial pattern of change is robust across across timescales, including glacial-interglacial, AIM events and millennial variability, is consistent with observed changes in the last century, and holds for both warming and cooling. Because this pattern persists across climate changes driven by diverse forcing over a range of timescales, it may reflect a fundamental aspect of Earth's energy balance. We investigate potential sources for this pattern with a simple energy balance model and idealized General Circulation Model experiments.

Bradley Markle [#14]

B. Lecavalier*, L. Tarasov

To better understand the role Antarctica played in the global climate, the observed contemporaneous change, and to make predictions of its future behaviour, reconstructions of past ice sheet evolution are required with uncertainty estimates. Glaciological modelling is an effective tool to generate continental-scale reconstructions over glacial cycles, but the models depend on parameterisations to account for the deficiencies (e.g., missing physics, unresolved sub-grid processes, uncertain boundary conditions) inherent in any numerical model. These parameters, considered together, form a parameter phase space from which sets of parameters can be sampled; each set corresponds to an ice sheet reconstruction. We explore an updated Glacial Systems Model (GSM) of the Antarctic ice sheet over the last glacial cycle by performing a Latin hypercube sampling of the parameter phase space. This yields a large ensemble of Antarctic reconstructions which can be compared against observational constraints. This provides the opportunity to assess the ability of the GSM to envelope the observational constraints given the parametric uncertainties and discuss the implications for the evolution of the Antarctic Ice Sheet.

Benoit Lecavalier [#13]

E. Gasson*, R. DeConto, D. Pollard, C. Clark

Recently obtained geophysical data show sets of parallel erosional features on the Lomonosov Ridge in the central Arctic Basin, indicative of ice grounding in water depths up to 1280 m. These features have been interpreted as being formed by an ice shelf—either restricted to the Amerasian Basin (the “minimum model”) or extending across the entire Arctic Basin. Here, we use a numerical ice sheet-shelf model to explore how such an ice shelf could form and what significance it would have to the Earth System. In particular we focus on how such a floating ice mass would affect sea level estimates and what impact it would have on ocean overturning during its break-up.

Edward Gasson [#11]

T. Lane*, V. Jomelli, V. Rinterknecht

The periphery of Greenland contains a large number of easily small ice caps and independent glaciers close to the present Greenland Ice Sheet margin. These independent ice masses respond sensitively to climate, in particular summer air temperatures. Reconstructing the fluctuations of these ice masses over timescales of 102 – 104 years allows them to be used as a climate proxy in a location close to the present ice sheet margin. Here we reconstruct the fluctuations of the A.P. Olsen ice cap in Zackenberg, East Greenland using geomorphological mapping and surface exposure ages (10Be). More extensive positions of the ice caps are marked by extensive moraine deposits and widespread erratic boulders. Results suggest ice cap extents were more extensive than previously thought, expanding over land formerly cited as covered by the Greenland Ice Sheet. It is likely that the ice cap became confluent with Greenland Ice Sheet ice. A new regional chronology has been produced using 10Be surface exposure ages from moraine boulders and glacially scoured bedrock samples from Store Sødal. These results suggest an earlier deglaciation than previously thought, with ice caps both on Clavering Ø island and the mainland displaying retreat through the Younger Dryas (12.9 – 11.7 ka BP). This demonstrates that ice cap recession not in phase with local air temperatures, which are thought to have been depressed by up to 15°C during Greenland Stadial 1.

Timothy Lane [#16]

R. C. A. Hindmarsh*, C. D. Clark, J. C. Ely, G. Hiess, G. R. Bigg, S. L. Bradley, R. C. Chiverell, D. Fabel, N. Gandy, E. G. Gasson, L. J. Gregoire, C. Ó. Cofaigh, D. Pollard, J. D Scourse, C. K. Ballantyne, S. Benetti, T. Bradwell, S. L. Callard, D. J. A. Evans, D. H. Roberts, M. Saher, D. L. Small, P. L. Whitehouse. 

The edge of the grounded part of an ice sheet is the grounding-line (GL). GLs on a reverse bed-slope, where the sea shallows in the direction of ice flow, could be unstable, since ice thickness and flux at the GL increase upon retreat, possibly leading to a marine-ice-sheet instability (MISI). It is a concern for the centurial future, since any MISI might cause sea-level rise of several metres. Observations of modern GLs are insufficiently longstanding to answer whether they are experiencing instability, meaning GL retreat 10-30 ka BP is of interest. The British and Irish Ice Sheet (BIIS) had large areas grounded below sea-level. BRITICE-CHRONO (B-C) was set up in 2012, comprising land-based observations in Great Britain and Ireland and cruises around these islands. Observations reconfirmed the BIIS was a marine ice-sheet, with GLs at the continental shelf edge. GL retreat commenced on the shelf edge before 20 ka BP and post-LGM warming. B-C surveyed transects parallel with paleo-ice-flow and found variable retreat rates before and after the LGM. These raise the questions (i) why did retreat commence before warming started? (ii) what caused variable retreat rates? B-C seeks to identify when and where the MISI occurred – is rapid retreat an indication of the MISI operating as contrasted with strong climate forcing? Interplay of ice growth, glacio-isostatic adjustment and sea-level change is one aspect; another is the presence of ice shelves, which inhibits the MISI. Factored in is the strength of climate forcing. Modelling teams at Sheffield/Delft, BAS, Leeds and Bristol/PSU are active, focussing on informing geological interpretation. B-C will make geodata available, from ‘raw’ data to forms filtered by geological insight. We will present the spatial and temporal range of the data, modelling underway and results obtained. GL dynamics played a fundamental role in Northern Hemisphere deglaciation.

Richard Hindmarsh [#10, 14]

T. Lane, Ø. Paasche*, B. Kvisvik, K. Adamson, Á. Rodés, H. Patton

The dynamical behaviour and imprint of the interior of the Fennoscandian Ice Sheet (FIS) during deglaciation is poorly constrained, with piecemeal reconstructions inferred from intermittent marginal positions and remote isostatic loading histories. Here we combine geomorphological mapping with surface exposure dating from central Norway to provide direct evidence of interior ice sheet dynamics. We find that high-elevation (>1,600 m) sites became ice free before 16 ka, and moraines were deposited at 11.6 ± 0.2 ka and 11.8 ± 0.2 ka during the Younger Dryas (YD). Our results demonstrate that sectors in southern central Norway contained extensive ice-free tracts prior to the YD, contrary to previous hypotheses of persistent, cold-based ice domes. These observations suggest that FIS draw-down was far more rapid and dynamic than previously thought, and occurred concurrently with patterns of marginal retreat. Present empirical and model reconstructions fail to capture such rapid interior down-wastage, leading to considerable uncertainty in FIS volume changes and concomitant eustatic sea-level contribution.  Moreover, such dynamic behaviour in past ice sheets portends to the potential rapid interior draw-down of the Greenland and Antarctic ice sheets if they continue on their current deglaciation trajectory.

Øyvind Paasche [#All]

F. Muschitiello* et al.

Constraining the response time of the climate system to changes in North Atlantic Deep Water (NADW) formation is fundamental to improving climate and Atlantic Meridional Overturning Circulation predictability. Here we report a new synchronization of terrestrial, marine, and ice-core records, which allows the first quantitative determination of the response time of North Atlantic climate to changes in high-latitude NADW formation rate during the last deglaciation. Using a continuous record of deep water ventilation from the Nordic Seas, we identify a 400-year lead of changes in high-latitude NADW formation ahead of abrupt climate changes recorded in Greenland ice cores at the onset and end of the Younger Dryas stadial (YD), which likely occurred in response to gradual changes in temperature- and wind-driven freshwater transport. We suggest that variations in Nordic Seas deep-water circulation are precursors to abrupt climate changes and that future model studies should address this phasing.

Francesco Muschitiello [#13]

Climate and volcanism interact across a range of timescales, lengthscales and forcing mechanisms.

The smallest range considers weeks with km-scale resolution: individual eruptions placing aerosols, CO₂, sulphur and halogens (and other chemicals) into the atmosphere.  Important variables include the height of injection, the atmospheric layer the volcanic emissions persist in, consequent chemical reactions and the lifetime of the emissions.

The largest range considers hundreds of Ka with global resolution: global volcanism as an exchange mechanism between mantle and surface reservoirs.  In particular, the atmospheric CO2 concentration could be influenced by variations in volcanic activity, be it Large Igneous Provinces or the pressure-driven volcanism associated with changing ice sheets and sea-level.

This session includes contributions using both instrumental records and physical models to improve our knowledge across these scales.

L. M. E. Percival* et al.

Numerous perturbations in Earth’s climate and surface environment, both long- and short-term in nature, have taken place throughout the 4.54-billion-year history of the planet. In the last 300 million years, a large number of geologically abrupt climate warming events have taken place, which caused major disturbances to the global environment, and sometimes mass extinctions. Establishing the causes and development of past climate change events is of great use when considering the considerable current and future effects of anthropogenic warming. Similarly to the modern day, past warming events are thought to have been driven by a major flux of carbon dioxide (CO2) to the atmosphere; however, the precise origin of those CO2 emissions remains debated. Large-scale volcanic gas emissions are frequently proposed as being the chief source of the carbon, potentially aided by the heating and degassing of volatile-rich country rocks intruded by magmas that gave rise to this volcanism. Volcanogenic CO2 emissions are supported by a strong coincidence of the dates of past climate events with volcanism linked to the formation of Large Igneous Provinces. However, uncertainties in those dates mean that a direct tracer of volcanism in sedimentary records of past climate change is preferable. Mercury (Hg) is a trace volcanic gas emitted to the atmosphere, and volcanism is a major natural source of the element to the Earth’ surface. Hg has a relatively long atmospheric residence time of about 1 year, allowing for hemispheric–global distribution before it is removed and ultimately deposited in sediments. Here, Hg data from a number of sedimentary records of major climate warming are presented, together with information on the ocean-atmosphere CO2 inventory, and correlations between the two volcanic gases (Hg and CO2) shown. Such correlations strongly support hypotheses of a major role for volcanic CO2 in causing past climate warming events.

Lawrence Percival [#15]

P. Huybers*, P. Liautaud, C. Proistosescu, B. Boulahanis, S. M. Carbotte, R. F. Katz, C. Langmuir

Abyssal hills are ubiquitous seafloor features that form at mid-ocean ridges. A hypothesis that their relief represents changes in melt supply driven by sea- level variations was contested on the grounds that faulting independently pro- duces abyssal hill spacing. Driving a fault model with periodic changes in melt supply resulting from late-Pleistocene sea-level variations, however, leads to a prediction that ridges spreading at intermediate rates respond to 100 ky variations in sea level and faster-spreading ridges to 41 ky variations. This prediction is supported by new findings of 100 ky and 41 ky spectral peaks in a global analysis of 207 bathymetry sections covering the late Pleistocene and of a linear relationship between spreading rate and increased abyssal hill spacing at fast-spreading ridges. A combination of faulting and magma supply variations influenced by changes in sea level can explain the global distribution of abyssal hill spacing on late-Pleistocene crust.

Peter Huybers [#15, 17,18]

J. M. A. Burley*, R. F. Katz, P. Huybers

The Late Pleistocene glacial cycles show a mismatch between the timeseries of orbital forcings (predominantly 20 & 40 ka) and sea level (~100 ka). These >40 ka cycles in sea level require that internal dynamics in the Earth system create a glacial response that is not linearly related to insolation. MOR CO2 emissions are driven by sea-level change with a tens-of-thousands-of-years timescale that may enable this glacial response. Available data does not make convincing statements on the probability of this hypothesis, so we have instead looked towards a dynamical model to elucidate plausible Earth behaviour. This work considers a reduced-complexity Earth system model that can run for a million years, resolving orbital forcing at a daily resolution and allowing CO2 to evolve over time. We show that glacially-driven variations in volcanic CO2 emissions are a plausible physical mechanism to create a 100 ka Earth system response to orbital forcing. We also show some of the surprising physical insights that come from the model's resolution and coupling of dynamical scales not previously considered: that CO2 drives the [slight] majority of sea level change in a glacial, and that the efficacy of any one of obliquity, precession and CO2 in changing the climate varies significantly depending on the states of the other two drivers (eg. sea-level sensitivity to CO2 changes by 50% depending on obliquity and precession)

Jonathan Burley [#15]

H.-B. Fredriksen*, M. Rypdal

I will demonstrate how we can use linear response models to quickly compute the temperature response to a given forcing record. I will also discuss the limitations of these models. For volcanoes, the main challenge is that there are large uncertainties associated with translating the aerosols to a global forcing record measured in W/m^2. It is also a challenge for linear modeling that volcanic forcing may affect mainly one hemisphere. In addition, I will present some work in progress on an oscillatory response model for the North Atlantic only. This is a simple dynamical model suggesting how temperature oscillations like the Atlantic Multidecadal Oscillation (AMO) may arise, in connection with changes in the Atlantic Meridional Overturning Circulation (AMOC).

Hege-Beate Fredriksen [#15]

H. Brenna*, S. Kutterolf, M. Mills, K. Krüger

The Los Chocoyos eruption (Magnitude ~8, dated to 84 kyrs before present) was one of the largest volcanic eruptions during the past 100,000 years. Originating from present-day Guatemala, the eruption formed the current stage of the large Atitlán caldera. Los Chocoyos released more than ~1100 km3 of tephra and the eruption is used as a widespread stratigraphic marker during that time. The ash layers can be found in marine deposits from offshore Ecuador to Florida over an area of more than 10 million square kilometers. Using a new estimate of erupted magma mass and recent volatile measurements we estimate that the Los Chocoyos eruption released >1045 Mt SO2, ~1200 megatons of chlorine, and ~2 megatons of bromine, which classifies it as a super-size eruption. Considering these volatile emissions, the eruption must have caused massive effects on the atmosphere, climate and environment at that time, e.g. pronounced and long lasting ozone depletion with impacts on surface ultraviolet radiation,. We will present results of the impact of volatile injections from the super-size Los Chocoyos eruption on atmospheric composition, chemistry and radiation. We will use the newly developed coupled chemistry climate model CESM2(WACCM) taking the combined effect of both sulfur and halogen interactively into account. The model results will be compared with a sulfur-rich only volcanic eruption. The analysis will focus in particular on halogen and ozone chemistry, radiation, atmospheric circulation and surface climate changes.

Hans Brenna [#15]

What can be learned from periods in the past that were warmer than today? How can these insights be used to constrain future warming and related climate processes and feedbacks?

In this session, the presenters discuss understanding gained from studying past warm climates. From assessing the stability of the (Atlantic) overturning circulation and its relation to freshwater forcing due to changing Northern Hemispheric ice sheets during the Quaternary, to interactions between the Southern Ocean and the Antarctic ice sheet in the late Miocene and Pliocene. From seeking to establish the conditions under which the Greenland ice sheets could exist during the warm Oligocene, to understanding the consequences of a bipolar world on climate dynamics. And from deciphering atmospheric responses to extreme greenhouse gas perturbations during the Paleocene-Eocene Thermal Maximum, to understanding other abrupt and high magnitude climate events in the more recent past.

In an introductory talk, we will briefly present Cenozoic climate change in order to put the presented work in a geological context, and address the challenges and promises of using knowledge from past warm periods for constraining future climate.

D. Holmes, T. Babila, U. Ninneman, G. Bromley, S. Tyrell, E. Rohling, M. Curran, A. Morley*

Recent findings suggests that we may have overestimated the stability of the AMOC to global warming due to model bias in future climate simulations and ignored thermal buoyancy forcing as a possible trigger to initiate AMOC collapse. Similarly, the focus on either large freshwater events or the presence of large northern hemisphere ice sheets as prerequisites for AMOC shut/slow-downs in palaeoceanographic investigations curtails our understanding of mechanisms and boundary conditions that characterize possible AMOC disruptions during warm climates. Using a core site located in the Rockall Trough, we present new evidence for an abrupt high magnitude climate event that occurred at the end of MIS 11, 390ka ago. We reconstructed sea surface temperatures using foraminifer assemblage counts and deep water flow strength using end-member modelling of grain size distributions. Two sharp cooling steps abruptly terminate a period of warm climate and mark the passage of the Arctic front over the Rockall Trough. The evolution of the deep water flow end-member suggests that a decrease in overflow into the Rockall Trough occurred at least three centuries before temperatures cooled abruptly at the surface and remained low throughout the event. Our results provide evidence that abrupt changes in North Atlantic climate were coeval with changes in ocean circulation changes, ice volume, and sea-level rise during a time when the climate system was not yet in a glacial state, challenging the ice volume (IV) threshold hypothesis (e.g. McManus et al. 1999).

Audrey Morley [#09]

G. Ramstein*, N. Tan, J. Ladant, C. Dumas, P. Bachem, E. Jansen

For long, our planet has not experienced large and perennial ice sheets. Indeed, since the last and huge glaciation occurring during the carboniferous Permian boundary (320-270 Ma), the cryosphere did not play a major role on climate changes for more than 200 million years. Since 40 million years, the long-lasting CO2 decrease allowed first the onset of Antarctica ice sheet (34 Ma ago) and a very asymmetric climate dynamics with only one ice sheet located at the southern pole favoring deep-water convection mostly in the southern hemisphere. Second, 30 million years were necessary to decrease the CO2 and achieve values low enough to trigger the inception of the Greenland (Deconto et al., 2008) Recent modeling studies (Tan et al., 2018 and Willeit et al., 2015) demonstrated how pCO2 evolution, into a narrow window, was pivotal to maintain the Greenland ice sheet during Plio-Pleistocene boundary. This talk will focus mostly on the consequences of the establishment of two ice sheets in each hemispheres towards a more symmetric climate world. We will also show how these new conditions favored large Northern hemisphere glacial-interglacial oscillations. Moreover, this cooling transition has also major impact on the dispersal of our human ancestors in the tropics modifying drastically the hydrological cycle in the tropics. We will conclude speculating on the long-term future impact of cryosphere evolution on climate dynamics.

Gilles Ramstein [#11]

R. P. Caballero-Gill*, L. A. Hinnov

The latest Miocene-Pliocene Epoch (~7 Ma to 2.7 Ma) was a pivotal time in Earth’s history when the global landscape and climate underwent the latest transition from “greenhouse world” to “icehouse world”. Understanding the mechanisms behind this transition is one of the grand challenges in Cenozoic paleoclimatology. The Southern Ocean is a key player in the climate system, impacting ocean circulation and the redistribution of heat, nutrients, and freshwater across latitudes, and as a result is capable of influencing global climate dynamics. Perhaps more importantly, the bulk of ice volume, sea ice, and albedo changes during the Miocene-Pliocene Epoch likely originated in the Southern Ocean. Unfortunately, there are only a very few continuous records of Southern Hemisphere paleoclimatic variability that cover the Epoch. Here we present new high-resolution proxy records that monitor sea surface temperature (alkenone unsaturation), haptophyte productivity (C37total), lithostratigraphic variations (photoscans), and intermediate water density and chemistry (stable isotopes from benthic foraminifera) from DSDP Site 594 (east of New Zealand) within the Southern Ocean. We evaluate the presence of Milankovitch forcing signals in these records, which provide a high-resolution astrochronology, and discuss the paleoclimate responses that occurred during the transition.

Rocio Caballero-Gill [#11]

J.-E. Chu*, A. Timmermann

The Paleocene-Eocene Thermal Maximum (PETM) was one of the most dramatic climate events in Earth’s history. Around 55 million years ago a massive amount of carbon was rapidly released into the atmosphere, causing global warming and ocean acidification. There is an ongoing debate as to what triggered this event and whether its climatic response can serve as an analog for a future projected quadrupling of the atmospheric CO2 concentrations. To gain a deeper understanding of atmospheric dynamical processes and their response to extreme greenhouse gas perturbations we analyzed two high-resolution AGCM simulation (50 km horizontal resolution) which were run for post-PETM (Eocene) and PETM boundary conditions. The simulated temperature pattern agrees well with paleo-proxy reconstructions; in particular, the lack of strong meridional temperature gradients is well captured. In response to the applied PETM boundary conditions, the AGCM simulates a massive reorganization of atmospheric flow and its instabilities. Small-scale weather phenomena, such as tropical cyclones play a much stronger role in global energy and water cycles compared to today’s climate, especially over the Arctic region. As a result of low extratropical baroclinicity, hurricanes can penetrate far into the high and polar latitudes. Our results document that extreme hothouse conditions are likely to trigger a transition to a new atmospheric regime.

Jung-Eun Chu [#17]

P. M. Langebroek*

The Eocene-Oligocene transition (~34 Ma) is one of the major climate transitions of the Cenozoic era. Atmospheric CO2 decreased from the high levels of the Greenhouse world (>1000 ppm) to values of about 600-700 ppm in the early Oligocene. High latitude temperatures dropped by several degrees, causing a large-scale expansion of the Antarctic ice sheet. Concurrently, in the Northern Hemisphere, the inception of ice caps on Greenland is suggested by indirect evidence from ice-rafted debris and changes in erosional regime. However, ice sheet models have not been able to simulate extensive ice on Greenland under the warm climates of the Eocene-Oligocene transition. We show that elevated bedrock topography is key in solving this inconsistency. During the late Eocene / early Oligocene, Greenland bedrock elevations were likely higher than today due to tectonic and deep-Earth processes related to the break-up of the North Atlantic and the position of the Icelandic plume. When allowing for higher initial bedrock topography, we do simulate a large ice cap on Greenland under the still relatively warm climate of the early Oligocene. Ice inception takes place at high elevations in the colder regions of north and northeast Greenland; with the size of the ice sheet being strongly dependent on the climate forcing and the bedrock topography applied.

Petra Langebroek [#11]

U. Ninnemann*, E. V. Galaasen, N. Irvali, Y. Rosenthal, H. F. Kleiven, D. A. Hodell

Thermohaline Circulation (THC) plays a crucial role in ventilating the ocean interior, in the ocean’s sequestration of carbon, and influences regional climate and sea level. Because of this, the sensitivity of THC to future warming and freshwater forcing is of major concern. Past changes in NADW, a major constituent of THC, have been documented in association with abrupt, millennial-scale, climate oscillations during glacial cycles. This association has led to wide speculation about how the mode of THC could be sensitive to, and play a role in, abrupt climate changes. However, proxy reconstructions generally find little to no evidence for large millennial-scale reductions in NADW influence during interglacial periods, leading to a general consensus that deep water ventilation is vigorous and relatively stable during warm climate intervals—potentially contributing to the relative climate stability of these warm periods. In contrast to this apparent multi-millennial stability, recent monitoring of the Atlantic meridional overturning circulation has revealed high-frequency variability within the period of observation, including in the dense components of the overturning contributing to NADW. Yet little is known about THC variability and sensitivity on more intermediary (e.g. multi-decadal to centennial) timescales during past warm (interglacial) climates, posing challenges for prediction and attribution of future circulation change and its impacts. 

Here we present new high-resolution NW Atlantic proxy reconstructions characterizing NADW ventilation (benthic foram d¹³C) and surface climate (foram assemblage, Mg/Ca) for multiple late Pleistocene interglacial periods. Our results reveal that subpolar climate of the current interglacial was the mildest of the past five interglacials. The warmer conditions experienced during previous interglacials were generally associated with reduced Greenland Ice Sheet volume. Large reductions in NADW ventilation occurred during each interglacial, revealing a previously unrecognized degree of warm-climate instability in deep Atlantic ventilation. The most pronounced and longest-lasting reductions in NADW ventilation occurred during the warmest interglacial in our records, MIS 9e.  In addition, multi-centennial reductions persisted throughout much of MIS 11c. Comparison to corrected ocean tracer fields (d¹³C _DIC) shows that the current ocean is well ventilated relative to interglacial natural variability (e.g. near the Holocene maximum), in seemingly sharp contrast to proxy-based reconstructions suggesting a recent anomalous decline in deep circulation and highlighting the difference between kinetics and ventilated renewal of deep water. Taken as a whole, our results reveal that the current interglacial was relatively mild with regard to both subpolar climate and to variability in deep water ventilation and may not be generally representative of late Pleistocene warm periods. In short, centennial-scale reductions in NADW ventilation, similar in magnitude to the glacial ocean changes, were frequent in the past when subpolar warmth was akin to what we may face in the near future.

Ulysses Ninnemann [#09]

Climate and weather interact with orography and landscape on a range of timescales. In this session we take a journey through some bizarre and idealized paleo- and future- Worlds, conducting model experiments to uncover the importance of the interactions between the land and our climate. Within this topic, the impact of mountains and plateaus on climate has always been the hottest. We will hear about how the presence of the Mongolian mountains impacts the tropospheric and stratospheric circulation more than the taller and more extensive Tibetan plateau and Himalaya, and also exert a greater impact on the Atlantic meridional overturning circulation than the American Rockies. Even a gap in a mountain range can impact climate largely - we will travel to the Sierra Madres mountains in central America to fill in the Tehuantepec Gap, and explore the role it plays in enhancing tropical cyclogenesis over the eastern Pacific. But mountains are not the only aspect of land that influences climate - properties of the land, and the land-ocean distribution also play key roles. For a moment, we will speed away from our semi-realistic Worlds and SLAM into a Simplified Land-Atmosphere Model. Here, we will learn how summertime temperature variability is impacted by evaporation from the soil, including the importance of a negative feedback between vapor pressure deficit and soil moisture that plays an increasingly large role in limiting evaporation as the land gets wetter. Back in a (somewhat!) more realistic setting, we will push the American continent back towards its African and European friends, and explore the climate of South America when the Atlantic ocean was half its current width - we will discover a much drier continent than present day as much less water vapor reaches South America due to the reduction in transit time across the Atlantic ocean. Lastly we will leave the past behind and look to the future, to study patterns of warming under climate change, and understand more about the causes of variability in regional warming uncertainty.

R. H. White*, D. S. Battisti, J. M. Wallace

We use a state of the art climate model to explore the impacts of mountains ranges on the large-scale circulation of the troposphere, stratosphere, and the ocean. We present the similarities between the original results on this topic from the 1980s and our results with a modern climate model (WACCM). We then show the surprises and new insights that can come from further exploration of these model simulations. This includes the strong zonal mean response that orography produces, and the importance of this zonal mean response for Earth’s stationary waves. We present evidence that the effect of the ‘Mongolian mountains’ on both tropospheric and stratospheric circulation dominates over the effect of the taller and more extensive Tibetan plateau and Himalaya. Without the presence of the ‘Mongolian mountains’ Sudden Stratospheric Warmings (events which also impact the surface climate, particularly in the Atlantic) decrease in frequency by over a factor of 4. Lastly, we look at the climate of Europe, with a focus on the oceanic Atlantic Meridional Overturning Circulation (AMOC). Contrary to expectations, the mountains with the largest impact on the AMOC, and subsequently temperatures of Northern Europe, are not the North American Rockies, rather, they are the mountains in Asia - the ‘Mongolian mountains’, Himalaya, and Tibetan plateau. We also present results showing the impacts of model resolution on the simulated response to mountains.

Rachel White [#12]

J. Baldwin*, J. Tralie, G. Vecchi

Near the Gulf of Tehuantepec in Central America there is a large gap in the Sierra Madres mountains. Observational studies have suggested that this gap plays a key role enhancing tropical cyclogenesis over the eastern Pacific by funneling strong winds from the Gulf of Mexico and creating vorticity anomalies. Here we systematically evaluate this hypothesis using an atmosphere-ocean coupled Global Climate Model GFDL CM2.5-FLOR, which has a relatively high-resolution tropical cyclone-permitting atmosphere (grid scale ~50 km), and somewhat lower resolution ocean (grid scale ~1 degree). Simulations are run with and without the Tehuantepec Gap, and with and without SSTs nudged to eliminate the role of low-frequency atmosphere-ocean coupling. With SSTs nudged, the Tehuantepec orographic gap does indeed increase tropical cyclone genesis through increasing low-level vorticity over the eastern Pacific. However, with atmosphere-ocean coupling, SSTs decrease over the eastern Pacific in the presence of the gap, decreasing tropical cyclone potential intensity and, in turn, genesis potential. Including the atmosphere-only and coupled responses, we find the Tehuantepec gap in the net has little influence on tropical cyclone density over the eastern Pacific. Dynamics of the SST changes, and contrasts between the conclusions of our work and prior observational studies will be discussed.

Jane Baldwin [#17]

L. V. Zeppetello*, D. S. Battisti, M. Baker

The relationship between evaporation and soil moisture is understood to be extremely important to the distribution of summertime temperatures. Observational analysis cannot determine the relationship between soil moisture and evaporation, and climate models do not have consistent representations of these two quantities. One dominant conceptual framework for describing soil moisture’s impact on evaporation was proposed by Budyko in 1961. This framework uses a critical value of soil moisture to determine the dominant control on evaporation: below the critical value, evaporation scales linearly with soil moisture; above it, evaporation is governed purely by atmospheric demand. We have developed the simple land atmosphere model (SLAM) as a tool for studying land atmosphere interaction. Using this model, we find that the relationship between soil moisture and evaporation is indeed dependent on climatological soil moisture; soil moisture and atmospheric demand play different roles in regulating evaporation depending on how wet the soil is. However, a critical value of soil moisture above which the dominant control switches is not required to explain this behavior. The different control regimes are due to a negative feedback between vapor pressure deficit and soil moisture that plays an increasingly large role in limiting evaporation as climatological soil moisture increases. This feedback explains the different soil moisture regimes without complex parameterizations and has important implications for summertime temperature variability in a warming climate.

Lucas Vargas Zeppetello [#18]

X. Liu*, R. White, and D. S. Battisti

We examined two factors that contribute to the early Eocene climate of tropical South America: a narrower Atlantic ocean and enhanced CO2. For this study, We used two atmospheric general circulation models (ECHAM 4.6 and CAM5) coupled to a slab ocean. Experiments show that, to first order, narrowing the Atlantic decreases the precipitation of tropical South America, whereas increasing atmospheric CO2 tends to increase the precipitation. The early Eocene climate in our model is a near-linear contribution of change in atmospheric CO2 concentration and change in Atlantic geometry, with a dominant contribution from the latter, resulting in a drier-than-today tropical South America during the early Eocene. Using water budget analysis, we find that the precipitation reduction induced by narrowing the Atlantic is mainly due to the reduction of water vapor flux entering South America across the northeast and east boundaries which, in turn, is due to a reduction in the amount of water evaporated from the ocean as air travels from Africa to the South American continent. It is not due to changes in atmospheric circulation. In fact, there is no dramatic atmospheric circulation change around South America even though the Atlantic Ocean is shrunk to less than half its modern width. Our study provides a step towards a dynamical understanding of how the Eocene climate of South America differs from today’s.

Xiaojuan Liu [#12]

The uncertainty in projections of climate change at the end of the 21st Century due to burning of fossil fuels is roughly equally due to uncertainty in greenhouse gas emissions and to the response of the climate to those emissions. Concerning the latter, it is well known that the climate models share a similar global pattern of warming. Here, I will show that the differences in the projections of warming across the models also share a common pattern of variability. The physical significance of the common difference pattern is explored. Once the common difference pattern is accounted for, there is little uncertainty left in the projections. Implications for numerical downscaling will be discussed.

David Battisti [#All]

The Earth’s climate is externally forced diurnally, annually, and over tens of thousands of years by the sun. However, within these timescales lies centennial and millennial climate variability, for which external forcing is often unclear or absent, and yet proxy evidence suggests was large and pervasive in the past. Abrupt climate variability on these timescales is critical to human civilizations. Millennial variability during the Quaternary appears to have been both global in extent and, in some places, extraordinarily abrupt. Yet many aspects of these events, and the processes that cause them, remain enigmatic. In the absence of external forcing, this variability may be internal and involve complex interactions between different aspects of the Earth system including the cryosphere, oceans, atmosphere, and geosphere. These timescales can be a challenge to study and often involve a variety of techniques including paleo-proxies, modeling, and theory. In this session we will see a range of studies investigating centennial and millennial climate variability and processes and their causal mechanisms. This session will cover Dansgaard–Oeschger events, Heinrich events, the importance of ocean variability, tropical hydroclimate, and sea ice. Talks will present a range of data from models and proxy records, and from the Arctic to Antarctic.

H. J. Andres*, L. Tarasov

Explaining the presence of abrupt climate cycles of large amplitude during Marine Isotope Stage 3 and the last deglaciation and their absence during the Last Glacial Maximum and the Holocene is an elusive problem. However, with the detection of self-sustained oscillations of large amplitude in paleoclimate simulations [1,2,3], it is now possible to test some of the proposed hypotheses in models. Common features of modelled oscillations so far include the abrupt release of accumulated heat under extensive sea ice cover in the Nordic Seas and changes to the transport of accumulated surface salinity in the subtropical gyre. Whether the same mechanisms are occurring in all of the different models exhibiting this type of variability is not yet clear, though, especially since the complexity and resolution of these models differ. In other areas of climate research, unforced oscillations arising in climate models are characterized as internal modes of variability (e.g. the North Atlantic Oscillation) through simple metrics that capture the key phenomena involved and allow the timescales and patterns of the variations in different models to be compared. Similarly, we characterize the key phenomena involved in D-Olike oscillations arising in the relatively simple and low-resolution Earth System Model (ESM), Planet Simulator (PlaSim) and define a simple metric to describe them. We then use this metric to compare these oscillations against those performed using the more complex ESM CESM1.2 in order to better understand the key requirements for such phenomena.1. Peltier, W. R. and G. Vettoretti, 2014, doi:10.1002/2014GL061413.2. Brown, N. and E. D. Galbraith, 2016, doi:10.5194/cp-12-1663-2016.3. Zhang, X. et al, 2014, doi:10.1038/nature13592.

Heather Andres [#16]

R. C. J. Wills*, W. R. Gray, D. S. Battisti, C. Proistosescu, L. Thompson, K. C. Armour, D. L. Hartmann, J. W. B. Rae

Low-frequency climate variability generally arises from a slow dynamical processes integrating higher-frequency stochastic forcing. Our goal is to identify these slow integrators by analyzing the anomaly patterns that evolve on the longest timescales. For this purpose we use Low-Frequency Component Analysis (LFCA), to isolate the modes of variability with the highest ratio of low-frequency to high-frequency variance. By applying LFCA to unforced 500-year integrations of CMIP5 coupled climate models, we identify the patterns of lowest frequency variability in sea-surface temperature (SST) and show that they are associated with variability in the Atlantic Meridional Overturning Circulation (AMOC), Southern Ocean residual overturning circulation, and North Pacific subpolar gyre circulation. By comparing with Pleistocene climatic changes, we suggest that the same mechanisms that govern unforced atmosphere-ocean internal variability govern responses of the atmosphere-ocean system to external forcing (e.g., greenhouse gasses, ice sheet changes, orbital changes). AMOC and Southern Ocean circulation changes have been invoked by a number of studies to explain past climatic changes, but the North Pacific subpolar gyre has not been investigated in this context. Using δ18O data from ocean sediment cores, we show how an intensified North Pacific subpolar gyre circulation at the Last Glacial Maximum (LGM) sustained a relatively warm climate in the subpolar North Pacific. Using PMIP3 climate models, we show that this was a response to ice sheet induced changes in wind-stress forcing, analogous to the physics whereby the Pacific Decadal Oscillation (PDO) in modern Pacific SSTs results from stochastic variability in wind-stress forcing over the North Pacific. This suggests that there are preferred patterns of ocean variability and change that are relevant across a continuum of timescales, from decadal to centennial and beyond.

Robert Jnglin Wills [#12]

K. Kleiven*, U. Ninnemann

The equatorward ventilation of Southern Hemisphere extratropical water masses to the low latitude thermocline has been proposed as a window through which the high latitude ocean can modulate tropical climate on anything from decadal to orbital timescales. This hypothesis is founded largely on the observation that tropical thermocline waters originate mostly in the Southern Hemisphere and that computer simulations suggest property anomalies in these source regions can advect through the intermediate ocean, “the ocean tunnel” to influence tropical SST. However, few observational records of extratropical ocean changes are available to assess their impacts on multi-decadal and longer timescales. Here we add to the observational record using new decadally resolved planktonic and benthic foraminiferal isotopic records spanning MIS 3 (20-50 ka) from the Chilean slope ODP Site 1233, located on the northern margin of the Antarctic Circumpolar Current. Thus the site is ideally situated to reconstruct both near surface and AAIW variability in the high southern latitudes. On centennial to millennial timescales, changes in intermediate water properties track those in the near surface albeit with a reduced amplitude—confirming the idea that changes in the extratropical ocean effect the oceanic tunnel on these timescales. The new benthic and plantic foraminiferal isotope results demonstrate that variations in intermediate ocean properties and climate of the southeast Pacific closely align with those recorded in the EPICA ice core from Dronning Maud Land. Such abrupt, synoptic scale changes in Antarctic climate and dynamics will have potentially widespread climatic and biogeochemical consequences along the downstream flowpath of AAIW. Such abrupt and concurrent shifts in extratropical surface and intermediate ocean properties are most likely driven by changes in the major dynamical systems of Southern Hemisphere ocean and atmospheric circulation which govern their properties today (the Westerlies and the ACC).

Kikki Kleiven [#09]

K. H. Nisancioglu*

DO-events, the abrupt warming events seen repeatedly during the Last Glacial, are the most extreme climate changes found in the Greenland ice cores. The repeated warming of up to 15 degrees in the course of a few decade are particularly clear during MIS3 (~60-30ka BP). The cold stadial periods, preceding each DO-event, exhibit large variations in duration and structure. The dynamics governing these cold stadials, as well as the abrupt transitions to warm interstadials, has alluded climate scientists since their discovery in the 70s. However, new proxy and model data provide an increasingly detailed view of these events. Here, we will summarize existing reconstructions and simulations of MIS3 with a focus on the cold stadials and the contrast between stadials with H-events (H-stadials) and stadials without H-events (DO-stadials). Studies suggest H-stadials exhibit two phases, different from the relatively stable cold phase of most DO-stadials. Here we investigate whether this could be caused by a delayed release of icebergs and meltwater to the North Atlantic, during a period of extensive Arctic and Nordic Seas sea ice cover, giving a significant reduction in ocean overturning circulation towards the end of Greenland stadials.

Kerim Nisancioglu [#All]

T. Sun*, A. Lawman, T. Shanahan, P. DiNezio, K. Gomez, N. Piatrunia, C. Sun, X. Wu, M. Kageyama, U. Merkel, B. Otto-Bliesner, A. Abe-Ouchi, G. Lohmann, J. Singarayer, X. Zhang

We explore the response of tropical climate to abrupt cooling of the North Atlantic (NA) during Heinrich Stadial 1 (HS1) combining paleoclimate proxies with model simulations. A total of 180 published paleoclimate records from tropical locations are used to categorize whether HS1 was wetter, drier, or unchanged relative to a deglacial baseline state. Only records with sufficient resolution to resolve HS1 and sufficient length to characterize the deglacial trend are considered. This synthesis reveals large-scale patterns of hydroclimate change relative to glacial conditions, confirming previously reported weaker Indian summer monsoon, a wetter southern Africa, and drying over the Caribbean. Our synthesis also reveals large-scale drying over the Maritime continent as well as wetter conditions in northern Australia and southern tropical South America. Our reinterpretation of the available proxy data reveals far more complexity and uncertainties for equatorial East Africa, a region that appears to straddle a pattern of dryer conditions to the north and wetter conditions to the south. Overall, these patterns of hydroclimate change depart from a southward shift of the Inter Tropical Convergence Zone (ITCZ), particularly outside the tropical Atlantic. We explore mechanisms driving these changes using a multi-model ensemble of “hosing” simulations performed relative to glacial conditions. The best-agreeing models indicate that cooling over the tropical NA is essential to communicate the response to the global tropics. This cooling is communicated to the Pacific over the Panama isthmus resulting in an El Niño-like pattern accompanied by drying over the Maritime Continent – as seen in the proxy data. Together these results show a dominant role for altered tropical SST gradient driving changes in tropical rainfall, and a lesser role for inter-hemispheric shifts in the ITCZ.

Tianyi Sun [#18]

M. F. Jensen*, K. H. Nisancioglu, M. A. Spall

During the last glacial period, when climate on Greenland is known to have been extremely unstable, sea ice is thought to have covered the Nordic seas. The dramatic excursions in climate during this period, seen as large abrupt warming events on Greenland and known as Dansgaard–Oeschger (DO) events, are proposed to have been caused by a rapid retreat of Nordic seas sea ice. Here, we study the sensitivity of sea ice in the Nordic Seas to changes in Atlantic water temperature using an eddy-resolving configuration of the Massachusetts Institute of Technology general circulation model with idealized topography. We show that a full sea ice cover and Arctic-like stratification can exist in the Nordic seas given a suffi- ciently cold Atlantic inflow and corresponding low transport of heat across the Greenland–Scotland Ridge. Once sea ice is established, continued sea ice formation and melt efficiently freshens the surface ocean and makes the deeper layers more saline. This creates a strong salinity stratification in the Nordic seas, similar to today’s Arctic Ocean, with a cold fresh surface layer protecting the overlying sea ice from the warm Atlantic water below. There is a nonlinear response in Nordic seas sea ice to Atlantic water temperature with simulated large abrupt changes in sea ice given small changes in inflowing temperature. This suggests that the DO events were more likely to have occurred during periods of reduced warm Atlantic water inflow to the Nordic seas. In addition, additional freshwater experiments show self-sustained oscillations in sea-ice cover without a change in forcing.

Mari. F. Jensen [#14]

D. Choudhury*, A. Timmermann, F. Schloesser, D. Pollard

In our study, we present new model simulations with a recently developed three-dimensional coupled climate – ice-sheet model (LOVECLIM – Penn State University ice-sheet model) covering the period from 240 thousand years ago (ka) to 170ka (MIS 7 to MIS 6). A series of initial sensitivity experiments reveals the presence of multiple climate – ice-sheet equilibria and run-away effects. To overcome unrealistic ice-sheet growth, we adjust several global parameters (such as climate sensitivity etc.) and enhance the basal sliding coefficient over the Hudson bay. Our simulations suggest such regional scale adjustments to affect the global response to orbital variations and being crucial for realistically simulating the amplitude of extreme glaciation events. More realistic simulations also show the emergence of millennial scale variability. We further test the hypothesis that millennial-scale dynamics play a pivotal role in ice-sheet growth/decay on orbital timescales.

Dipayan Choudhury [#18]

A. K. Faber*, C. Li, B. Vinther, C. Guo, K. H. Nisancioglu

Greenland ice cores allow us to investigate the past climate conditions at the Greenland Ice Sheet. The number of deep ice core drillings in recent years have allowed a deeper insight into the glacial conditions and the characteristics of Dansgaard-Oeschger (D-O) events. In particular the improved spatial distribution of ice core locations provide possibilities for a deeper investigation of how the signatures of D-O events varies across the Greenland Ice Sheet. In this talk we investigate the ice core records during D-O events and compare with a set of model simulations of glacial climate conditions from The Norwegian Earth System Model (NorESM). Particularly we focus on how the complex interplay between different climate components such as ice sheets, atmospheric circulation and sea ice might be recorded in the Greenland ice cores. This makes it possible to investigated whether models and ice core data paint a consistent picture of the Greenland climate conditions during D-O events and create the potential to explore the dynamics of the D-O events, as recorded by Greenland ice cores, in greater detail.

Anne-Katrine Faber [#11,17]

The seasonal cycle forces the largest observed variations in climate, and its signal is felt through the ocean, atmosphere, and cryosphere. The 2017 ACDC summer school explored the role of the seasonal cycle in each of these components of the Earth system. We learned how it governs the tropical climate by driving the monsoons and via interactions with ENSO, the extra-tropics by modulating the jet streams and storm track, and how it controls interactions between the ocean’s mixed layer and thermocline. On longer timescales, we looked at how its modification on orbital timescales results in glacial cycles. 

This session showcases a range of the impacts of the seasonal cycle, including its effect on both the ocean and the atmosphere. Using an oceanic observational network, the importance of considering differing seasonality in oceanic and biological processes is highlighted. The CMIP5 simulations are used to examine the projections of summertime temperature variance as well as wintertime interactions between Arctic sea-ice loss and Eurasian winter climate. Also, the seasonal monsoonal circulation is explored using an idealised modelling approach. These studies demonstrate the importance of understanding the changing dynamics over the seasonal cycle and modelling it correctly.

Z. Erickson*, A. Thompson

Vertical exchange of tracers such as carbon in the ocean is mediated by small scale (submesoscale; ~1-20 km) dynamics and instabilities. Destabilizing buoyancy fluxes and deeper mixed layers in the winter allow for greater levels of available potential energy that are released in a variety of submesoscale instabilities. Because of their small scale they are difficult to observe, yet these instabilities can lead to large vertical velocities and may induce substantial vertical transport of carbon. Here I use observations of fluorescence, a proxy for chlorophyll and biologically-fixed carbon, from autonomous underwater vehicles called Seagliders to characterize the seasonality of carbon export from these instabilities in different ocean regions and seasons. A key factor is the vertical stratification of the water column at the base of the mixed layer, which is dramatically reduced during winter, allowing the wintertime mixed layer to act as a window to the interior ocean. However, the seasonality of biological production is often not aligned with the seasonality of instabilities, limiting total export by this mechanism. Using relationships derived from regional Seaglider studies, we perform similar analyses on the global scale using Argo data.

Zach Erickson [#17]

D. Chan, A. Cobb*, D. S. Battisti, P. Huybers

In this study, we examine the biases in summertime surface temperature variance simulated by 30 of the CMIP5 models for the historical period (1986-2015); in many regions, the models have too much temperature variance compared to observations. We then show that in many regions, the projected change in the variance of temperature at end of the 21st century (RCP 8.5) scales as the amplitude of the bias in temperature variance in the current climate. Finally, we use the scaling relationships and the biases in historical simulations to correct the model projections. In general, the larger the bias in summertime temperature variance in the historical period, the greater the projected increase in the variance the model will project in the future scenario. This is especially true in Western Europe, where this relationship is shown to be statistically significant across large regions. In the original multi-model projection (RCP 8.5), much of Europe and parts North America and Canada are expected to experience a significant increase in summertime surface temperature variance as much as 1C^2. However, when the multi- model bias and the cross-model scaling relationship are considered, the new projections reduce the increase of temperature variance in these regions by up to 0.3C^2. This highlights the importance of taking into account model performance in the current climate when analyzing any simulations predicting the future climate.

Alison Cobb [#17]

R. Geen*, H. Lambert, G. Vallis

Simulations with aquaplanets have begun to untangle the connections between the seasonal Hadley cell and ITCZ, and Earth’s monsoons. However, while such experiments have been very successful at improving our understanding of the zonal mean, Earth’s monsoons are regional systems, anchored by continents and topography. An understanding of the controls on the regional behavior of the seasonally evolving overturning circulation remains elusive. In this work, we use an idealized modelling framework, Isca, to study seasonal migrations of tropical precipitation in an experiment in which the Eastern hemisphere is given a slab ocean of mixed layer depth 2m, while the Western hemisphere is given a mixed layer depth of 20m. This provides a very simple zonal 'land-sea' contrast. We find that ITCZ migrations in the zonally asymmetric experiment do not resemble either the 2m or 20m aquaplanet, nor an intermediate mixed layer depth aquaplanet, at any longitude. However, tropical precipitation now evolves following a broadscale lat-lon structure similar to that seen in observations. By analysing the moist static energy and momentum budgets, we investigate which mechanisms for monsoon onset and withdrawal identified in the zonally symmetric case carry over to this more complicated configuration, and explore how current theory can be extended to account for zonal asymmetry. 

Ruth Geen [#17]

S. He*, X. Xu, T. Furevik, Y. Gao, H. Wang, F. Li

Recently, the role of Arctic sea-ice loss on the Eurasian winter climate is a topic of controversy due to the negligible response of midlatitude atmosphere to sea-ice loss forcing in climate models which, however, show consistent Arctic surface warming. Does it mean that neither the Arctic sea-ice loss nor the Arctic warming can impact the Eurasian winter climate? It should be noted that the observed Arctic warming may extend to mid-troposphere rather than be confined to near-surface, which is concurrent with significant atmospheric change at midlatitudes. Therefore, we hypothesize that the Arctic surface warming and Arctic mid-tropospheric warming may be concurrent with different Eurasian climate response. Based on Coupled Model Intercomparison Project Phase 5 archives and long-term reanalysis datasets, we find that the probability for Eurasian winters concurrent with warmer Arctic mid-troposphere to suffer from more than 10 extreme cold days is higher by ~20% than that concurrent with warmer Arctic surface. Because the thermal-wind balance notion at midlatitudes (~60°N) is well established if the Arctic warmer anomaly extends upward mid-troposphere, leading to more Ural blocking events. The weakening of poleward temperature gradient and midlatitude jet stream, the increase of Arctic tropospheric thickness, and the Eurasian cold-lasting days related to the warmer Arctic mid-troposphere is more than two times of that associated with warmer Arctic surface. Multi-models’ simulations with natural forcing suggest that the atmospheric internal variability may contribute to different categories of warmer Arctic and Arctic-midlatitude teleconnection.

Shengping He [#17]

We do not live in a symmetric world. The Antarctic is colder than the Arctic, sea surface temperatures are colder off the west coast of continents than off the east, the North Atlantic is warmer than the North Pacific at the same latitudes, and even the periodicity of glacial-interglacial climates is uncertain. There are multiple reasons for these asymmetries, including hemispheric asymmetry in land fraction, asymmetry in ocean heat transport, atmospheric circulation, planetary albedo, and climate feedbacks. In this session, we will discuss global climate with a focus on regional patterns of climate change. We will examine the interplay between radiative feedbacks and patterns of sea surface warming to understand asymmetries in climate and constrain climate sensitivity and future warming. We will also explore physical mechanisms for Arctic amplification by looking at the lapse rate response of high latitudes to the temperature feedback. Lastly, we will consider asymmetry in Milankovitch periodicity during the Pleistocene by assessing relative contributions of northern and southern deep water sources to the benthic δ δ δ¹⁸O. Asymmetries in the earth system are the fundamental drivers of Earth’s global climate and climate sensitivity. Posing questions through the perspective of asymmetries is a powerful way to learn more about climate dynamics.

K. Armour*, Y. Dong, C. Proistosescu, D. S. Battisti

There is an increasing realization that global climate change depends on the interactions between radiative feedbacks and the spatial structure of sea-surface warming. The pattern of warming itself arises from a combination of regional ocean dynamics and atmospheric processes. Moreover, distinct radiative responses to regional warming at different locations arise from asymmetries in mean-state climate. The result is that Earth's global climate sensitivity to greenhouse gases inherently depends on climate asymmetries. Here we review recent progress in this area and discuss implications for our ability to predict warming over the coming centuries.

Kyle Armour [#11, 13]

M. Henry*

Polar amplification is a common feature of global climate models and of recent climate observations. This faster warming near the surface in the Arctic is explained by either increased atmospheric energy transport or by local feedbacks (surface albedo feedback and temperature feedback). Determining which mechanism is dominant is the subject of ongoing research. By only applying the radiative forcing from increased CO2 to the high latitudes, recent work shows the importance of local feedbacks. In these experiments, the high latitude temperature response is bottom-heavy and atmospheric energy transport convergence to the poles is reduced. This underlines the importance of local processes in generating the polar amplified temperature response to increased CO2. While the surface albedo feedback is active in this recent work, similar warming patterns are found in models with inactive surface albedo feedback. It is thus hypothesized that the temperature response of the high latitude atmosphere to increased longwave optical depth is bottom-heavy when the only active feedback is the temperature feedback and when there is no change in atmospheric energy transport. In this work, we use an analytic model of the high latitude atmosphere in radiative-advective equilibrium to find a physical reason to why the high latitudes have a positive lapse rate response to increased longwave optical depth when the only active feedback is the temperature feedback and there is no change in atmospheric energy transport. A radiative-convective model with an additional temperature tendency term representing atmospheric energy transport and global climate model simulations are then used to confirm these theoretical results.

Matthew Henry [#18]

A. L. Morée*, T. Sun, A. Bretones, E. O. Straume, K. H. Nisancioglu, J. Gebbie

d18O proxy records show a pronounced change in dominant Milankovitch periodicity during the Pleistocene. Precessional signals are less pronounced or even absent in the Early Pleistocene records as compared to the Late Pleistocene. In this modeling study, we show that this may be due to the cancellation of Northern and Southern-sourced d18O water mass signatures at depth. This is a result of the hemispheric anti-phasing of the precession signal during the winter (deep water formation) season, and causes a region in the mid-Atlantic to potentially carry a too weak precessional signal to be picked up by benthic d18O records. The relative contribution and d18O signal of the Northern and Southern deep water sources are key in the potential for cancellation, and we therefore explore a wide range of relative contributions and their effects on cancellation.

Anne Morée [#18]