Future climate may strongly warm during the next decades and centuries in response to anthropogenic greenhouse gas (GHG) emissions, specifically carbon dioxide (CO2). The Earth’s climate did considerably change already since the start of the industrialization featuring, for example, signicant global surface warming, ice loss and sea level rise (IPCC 2013). The term "Anthropocene" has been coined to describe the large human impact on the Earth system, which is now comparable to the natural influences. Atmospheric CO2 and methane (CH4) concentrations, for example, have reached levels that are unprecedented for at least the last 800,000 years (Fig. 1). However, large natural climate excursions have been also observed in the past (Fig. 1). Most prominent are the glacial cycles and these need to be understood and simulated with comprehensive Earth System Models (ESMs) in order to reliably project the future climate. Here we focus on the last glacial cycle.
Figure 1: (a) The 800,000-year records of atmospheric carbon dioxide (red; ppm) and methane (green; ppb.) from the EPICA Dome C ice core, together with a temperature reconstruction (relative to the average of the past millennium). Concentrations of greenhouse gases in the modern atmosphere are highly anomalous with respect to natural greenhouse-gas variations. (b) The carbon dioxide and methane trends from the past 2,000 years. See details in Brook (2008) and Lüthi et al. (2008).
The CO2-concentration has risen from its preindustrial value of about 280 ppmv to its current concentration of about 400 ppmv. It is well known that the increase in atmospheric greenhouse gases must lead to global surface warming due to an enhanced greenhouse effect. Consistent with the increase in GHG concentrations, the globally averaged surface air temperature (SAT) increased by 0.85°C during the period 1880-2012 (IPCC 2013). The recent increase in the radiative forcing by well mixed greenhouse gases and SAT is clearly unusual when considering the whole last millennium (Fig. 2). Further, the globally averaged sea level rose by about 20 cm since 1900 (IPCC 2013), which is due to the thermal expansion of the sea water and melt water contributions from mountain glaciers, ice caps and continental ice sheets. A large regional variation, however, was observed in the climate trends during the 20th and early 21st century (IPCC 2013), reflecting the effects of natural variability, which superimposes the anthropogenic trends, and regionally varying feedbacks. Although it is unquestionable that the climate of the planet is changing (Figs. 1 and 2), it remains controversial how much of the observed climate change can be attributed to anthropogenic forcing and also how the climate will evolve in the future in response to natural and anthropogenic forcing.
Figure 2: Last-millennium simulations and reconstructions. (a) 850–2000 PMIP3/CMIP5 radiative forcing due to volcanic, solar and well mixed greenhouse gases. (b) 850–2000 PMIP3/CMIP5 simulated (red) and reconstructed (shading) NH temperature changes. The thick red line depicts the multi-model mean while the thin red lines show the multi-model 90% range. See details in IPCC 2013 (Box TS.5, Figure 1a, b).
An inherent feature of the climate system is its variability. The climate of the Earth varies naturally on a broad range of time scales, where the amplitude of climate changes generally increases with the time scale. The most prominent examples of major climate excursions during the late Pleistocene, i.e. during the last about 650,000 years, are the glacial cycles with a period of approximately 100,000 years (Fig. 1), a waxing and waning of cold and warm climates which are referred to as glacials and interglacials, respectively. It is generally accepted (Imbrie et al. 1992) that the glacial cycles are forced by changes in the Earth’s orbital parameters (Hays et al. 1976), known as the Milankovitch theory which describes the effects of changes in the Earth's orbit around the sun on the geographic distribution of the insolation. The globally averaged SAT change between glacials and interglacials amounts to about 5-6°C, while globally averaged sea level changes are of the order of 100m (see Fig. 5 for the rise during the last deglaciation) and variations in atmospheric CO2 on the order of about 100 ppmv (Fig.1).
The high correlation between globally averaged SAT and atmospheric CO2 seen in Fig. 1a suggests a tight coupling between the physical climate system and the carbon cycle on multi-millennial time scales such that perturbations in either component will be reinforced by the other. Let us consider the marine component of the global carbon cycle. The marine carbon uptake depends among other factors on the sea surface temperature (SST), because the solubility of carbon dioxide is a strong inverse function of SST. The latter is largely governed by the strength of the greenhouse effect that in turn depends on the amount of atmospheric CO2. Further, the oceanic carbon uptake also depends on biological processes, but little is known about biologically-dependent mechanisms of CO2 uptake and how they vary with the physical and chemical ocean state, and with time scale. The role of dust for the ocean’s ecology on glacial time scales needs to be investigated in this context. Unresolved issues in the global carbon cycle also pertain to its terrestrial component. We need to better understand soil and vegetation feedbacks, and the time scale at which weathering of CO2 becomes important. Finally, we need to understand the connections between the different biogeochemical cycles. It is well known, for instance, that the cycles of nitrogen and carbon are strongly coupled with each other owing to the metabolic needs of organisms for these two elements. Suffice it to say that a reliable projection of the future climate obviously requires a throrrough understanding of the biogeochemical cycles, in particular the global carbon cycle, on wide range of time scales and their interactions with the physical climate system. The simulation of the last glacial cycle with its large excusrions in both physical and biogeochemical parameters offers a unique opportunity to test our Earth System Models.
However, climate variability exists on all time scales, from seasonal and interannual to multi- millenial, and climatic processes are intimately coupled, so that understanding variability at any one time scale requires some understanding of the whole spectrum (Huybers and Curry 2006). Discussion of long-term climate variability is commonly divided between deterministic and stochastic components often associated with spectral peaks and the continuum (the “red” background), respectively. Spectral analysis of surface temperature variability (Fig. 3) indicates that the annual and Milankovitch energy are linked with the continuum, and together represent the climate response to insolation forcing. Whereas the annual and Milankovitch-period climate variations yield information primarily about the forcing, the continuum tells us about the dynamical processes governing climate variability.
The existence of two separate scaling regimes above and below centennial periods suggests the presence of distinct controls on the climate variability. First, the climate system has memory associated with the oceans that causes high-frequency variability to accumulate into progressively larger and longer-period variations (termed the stochastic climate model, see e.g. Hasselmann 1976). Second, a Milankovitch-driven low-frequency response transfers spectral energy toward higher frequencies, possibly involving nonlinear ice-sheet dynamics. These low- and high-frequency surface temperature responses appear to be of nearly equal magnitude at centennial time scales, the time scale at which the effects of global warming become prominent. This is why we have to understand the processes that lead to climate variability over a wide range of time scales, if we wish to skillfully project the climate during the next centuries. Because the annual and Milankovitch cycles are deterministic, a greater understanding of their influence on the continuum may enhance the skill of climate predictions for the 21st century or the entire millennium.
Figure 3: Patch-work spectral estimate using instrumental and proxy records of surface temperature variability (top panel), and insolation at 65°N (bottom panel, see Huybers and Curry 2006 for details).
The climate of the last glacial period has been extremely variable, with rapid and strong cold and warm events (Fig. 4). The origin of the cold events, termed Heinrich events (HE), is still controversial (see e.g. Alvarez-Solas and Ramstein 2011 and references therein), but both ocean dynamical processes and ice sheet instability are presumably critical to their development. Various mechanisms have been proposed to explain the cause of Heinrich events. Most center on the variability of the Laurentide ice sheet, but others suggest that the instability of the West Antarctic Ice Sheet may have played a triggering role. The quantification of the relative roles of Northern and Southern Hemisphere processes in glacial-interglacial transitions and in glacial climate variability is an important scientific question we wish to address.
One particular major cold event in the Greenland ice core that attracted a lot of attention is the Younger Dryas, an event also referred to as the “Big Freeze” (Berger 1990). It was a relatively brief (1,300 ± 70 years) return to very cold climatic conditions which occurred between approximately 12,800 and 11,500 BP. The Younger Dryas event is thought to have been caused by the collapse of the North American ice sheets leading to enormous amounts of fresh water entering the North Atlantic thereby slowing the Atlantic Meridional Overturning Circulation (AMOC), although several competing theories have also been proposed. Yet the nature of the Younger Dryas remains a mystery. Was the Younger Dryas a one-time event or an ‘‘accident’’ of the last termination? Or, are Younger Dryas-type events an intrinsic feature of climate change during glacial-interglacial cycles as hypothesized by Sima et al. (2004) and Broecker et al. (2010)?
Figure 4: (a) The GRIP oxygen isotopic profile (blue) with respect to depth at GRIP. Isotopic values (δ18O) are expressed in ‰ with respect to Vienna Standard Mean Ocean Water (V-SMOW). (b) The NGRIP oxygen isotopic profile (red) with respect to depth at NGRIP. The GRIP record (blue) has been plotted on the NGRIP depth scale using the rapid transitions as tie points (c) The difference between the NGRIP and GRIP oxygen isotopic profiles plotted above on the GRIP2001/ss09sea time scale (black). The record is compared to a record representing sea level changes (green) and a 10-kyr smoothed oxygen isotope profile from NGRIP (red). See details in NGRIP members (2004).
A number of rapid and strong warming events followed by slow cooling are also seen in the Greenland temperature record, and these are termed Dansgaard-Oeschger (DO) events (Fig. 4). Again, the origin of these rapid events is controversial. Although the effects of the DO events are largely constrained to ice cores taken from Greenland, there is evidence to suggest DO events have been globally synchronous. The so-called bipolar seesaw (Stocker 1998) is one possibility to explain the coherence between variations in the Northern and Southern Hemisphere. DO events are quasi-periodic (Schulz 2002) and possibly involve major re-organizations of the global thermohaline circulation, specifically the AMOC (Rahmstorf 2002). If the basic ocean state is close to a bifurcation, separating a stable AMOC from one supporting oscillatory behavior, DO events may be excited simply by stochastic or a weak periodic external (solar) forcing (Bond et al. 1997 and Bond et al. 2001). Both the (cold) Heinrich and (warm) Dansgaard-Oeschger events are welcome test cases for comprehensive ESMs. Earth System Models of Intermediate Complexity (EMICs) are limited in this respect and unable to simulate the full range of variability. However, EMICs can provide valuable guidance pertaining to initialization of and selection of model parameters in comprehensive ESMs.
A lot of research during the recent years was devoted to the understanding of the climate of the last millennium (Fig. 2) and to conduct projections of the climate of the 21st century, the latter within the framework of the Coupled Model Intercomparison Project (CMIP). Considerable progress was achieved concerning these time periods, although a number of key questions remain open. The physics controlling centennial and longer-timescale surface temperature variability are distinct from that on shorter time scales, raising questions of whether climate models able to represent interannual to decadal variations will adequately represent the physics of centennial and longer-timescale climate variability. Different climate models simulate different levels of centennial variability, but the models generally tend to underestimate centennial variability. This was revealed by comparing the CMIP5- PMIP3 last-millennium simulations (e.g., Bothe et al. 2013) of Northern Hemisphere surface temperature with proxy data reconstructions (IPCC 2013, Chapter 9). The result is not surprising given the above discussion and suggests that the inclusion of more comprehensive physics and biogeochemistry will considerably improve the models.
There is a large uncertainty about the near-term climate change to be expected during the 21st century and the longer-term change during the entire millennium (IPCC 2013). This uncertainty is due to three factors: scenario uncertainty, model (response) uncertainty, and natural variability (Hawkins and Sutton 2009). The understanding of the natural variability is of primary importance when modeling and assessing the effects of anthropogenic climate change. On the one hand, the existence of natural variability makes the detection of anthropogenic climate signals a challenge, on all spatial scales: from local to even global. An example is the naturally varying deep ocean heat uptake which may be a reason for the current hiatus in global surface warming (IPCC 2013). On the other hand, we have to thoroughly understand the dynamics of the natural variability and disentangle the underlying processes on a large variety of temporal and spatial scales, because the natural variability itself may change in response to global warming.
The next level in the understanding of climate variability will be to disentangle the processes which determine the spectrum of the variability up to multi-millennial time scales including glacial- interglacial transitions. This is one of the grand challenges in climate research and requires the development and optimization of comprehensive Earth System Models (ESMs), the synthesis of the available data, and a high degree of collaboration between different communities. The understanding of climate variability on this broad range of time scales has also a large relevance for projecting future climate. Clearly, the climate fluctuations on glacial-interglacial time scales have been at least an order of magnitude larger than the changes observed during the 20th century (Fig. 5), but will this also be the case for the 21st century and beyond?
Transient simulations of the last termination with ESMs are one focus of the fist phase. Deciphering the evolution of global climate from the end of the Last Glacial Maximum (LGM) approximately about 19,000 BP to the early Holocene 11,000 BP presents an outstanding opportunity for understanding the transient response of Earth's climate system to external and internal forcings. During this interval of global warming, the decay of ice sheets caused globally averaged sea level to rise by approximately 80m (Fig. 5); terrestrial and marine ecosystems experienced large disturbances and range shifts; perturbations to the carbon cycle resulted in a net release of the greenhouse gases (Clark et al. 2012). However, the climate evolution was far away from being gradual. For example, periods are seen in the sea level reconstructions which depict increases of a few meters per century; events which are termed melt water pulses (MWPs, Fig. 5).
Figure 5: (a) Sea level reconstructed from corals recovered in long holes drilled onshore and offshore Tahiti Island. Coral depths are expressed in meters below present sea level. Grey and colored symbols show respectively coral samples collected in onshore holes and in offshore holes. (b) Magnified view of the MWP-1A. See Deschamps et al. (2012) for details.
All this raises questions about the the future climate. An example is future sea level rise. The rates of globally averaged sea level rise for selected time periods during the deglaciation are presented in Fig. 6 and compared with those during selected time periods during the late Holocene (IPCC 2013). We note the overall higher rates of global mean sea level change characteristic of times of transition between glacial and interglacial periods. The rate of sea level rise during MWP-1A was on the order of 40-50 mm/year; as opposed to the present rate obtained from satellite altimetry which is about 3 mm/year during 1993-2012. One pressing question is that of whether rates typical for past melt water events can occur in the future under anthropogenic global warming. In order to answer such questions we must understand and model with comprehensive Earth System Models the past changes in the rate of sea level rise. Another important question concerning sea level rise is the role of Antarctica during interglacials. Recent studies suggest that the sea level rise during the Eemian warm period was not exclusively driven by the Greenland ice sheet, but had a major contribution from the Antarctic ice sheet (NEEM community members 2013). Thus the Antarctic ice sheet cannot be ignored when projecting sea level rise during the 21st century and beyond.
Figure 6: (a) Estimates of the average rate of global mean sea level change (in mm yr-1) for five selected time intervals: last glacial-to-interglacial transition; Meltwater Pulse 1a (see Fig. 6); last 2 millennia; 20th century; satellite altimetry era (1993–2012). Blue columns denote time intervals of transition from a glacial to an interglacial period, whereas orange columns denote the current interglacial period. Black bars indicate the range of likely values of the average rate of global mean sea level change. (b) Expanded view of the rate of global mean sea level change during three time intervals of the present interglacial. IPCC 2013 (FAQ 5.2).
In summary, we have to develop comprehensive Earth System Models capable of simulating the climate evolution and the variability characteristics during the last glacial cycle and to perform reliable climate projections for the future. Simulating such a long time period will allow addressing the following key questions of climate research. What is the likelihood that changes that normally occur only on the very long multi-millennial time scales can materialize within this century or the next few centuries as suggested by some state-of-the-art climate models, if greenhouse gas emissions continue to grow unabatedly (IPCC 2013)? So far, most model studies have made an attempt to assessing equilibrium climate sensitivity (ESS), defined as the globally averaged equilibrium SAT change in response to a doubling of the preindustrial CO2-concentration, although the response of the very slow components of the Earth system such as ice sheets and weathering of CO2 are not represented in the models. Hence the question of the “true” climate system sensitivity, or Earth system sensitivity, remains unanswered. In fact, the issue of climate sensitivity may pose a nonlinear problem, and ESS may depend on the climate state itself. The successful simulation of the spectrum of climate variability during the last glacial cycle will allow a more reliable assessment of as to whether a regime shift in the variability may also occur in response to global warming during the next centuries. Will the future climate be more or less variable and what is the probability of rapid climate transitions? Could polar ice sheets collapse catastrophically, as in the past and documented by the melt water pulses described above? How quickly can sea level rise under present and future climate conditions? We shall be able to answer these questions only if we have understood and modeled in detail the climate of the full last glacial cycle.
Simulating the climate including the spectrum of its variability during the last glacial cycle constitutes a major enterprise in several respects. We describe the major challenges in the following:
It is widely accepted that glacial-interglacial transitions are forced by changes in the orbital parameters. In order to simulate such transitions, we have to model the growth and demise of the continental ice sheets, and this in fully interactive mode where only the orbital forcing is specified. Interactive ice sheet modeling, however, is still at its infancy. Many processes are not well understood such as the surging nature of ice sheet motion or the interactions with the ocean and ice shelves. We have to develop comprehensive ice sheet models which also simulate the regionally confined fast processes which are crucial to rapid climate variations. We can do this initially by considering only the physical Earth system and specifying changes in biogeochemistry as boundary conditions. This is still a major challenge: there is to date no successful example of the simulation of either the built-up or demise of an ice sheet in realistic configuration and in interactive mode using comprehensive ESMs. Successful interactive ice sheet modeling also depends, for example, on the quality of simulating the hydrological cycle which is so important to the glacier surface mass balance. Atmospheric general circulation models, however, still have serious problems in simulating cloud processes and regional precipitation patterns (IPCC 2013).
Physical system: internal variability
The second challenge concerns the internal variability, as the long-term climate evolution has not been smooth, as demonstrated, for example, by the Greenland ice core records depicting the regional climate of the last 100,000 years (Fig. 4). Although the orbital forcing is of multi-millennial scale, the temperature response in Greenland obviously is not. Several features can be noted in the Greenland records. First, the glacial climate was much more variable than the climate of the present warm period, the Holocene, which started 11,600 years before present (BP). In fact, the Greenland climate during the Holocene was remarkably stable, with only one major fluctuation around 8,200 BP. The sudden transition in the spectrum of the variability is reminiscent of a bifurcation or regime shift. The reason for this huge difference in variability between the glacial climate and that of the Holocene is largely unknown, but clearly reflects nonlinear behavior. Furthermore, glacial climate variability is dominated by a single mode, resulting in well expressed co-variations of the proxies with a fundamental 1470- year signal (Dansgaard-Oeschger cycles). In contrast, there is no compelling evidence for a dominant and persistent centennial-to-millennial climate cycle during the Holocene (Schulz et al. 2004). Interglacial climate variations seem to co-vary less pronounced among the different proxies than those of the last glacial period, suggesting the simultaneous activity of independent climate modes, each characterized by its own natural period in the range of approximately 400-3000 years. The reason for this difference in the variability between the glacial and Holocene climate is also unclear.
Second, the glacial climate depicted a number of extremely rapid and strong climate transitions. The extraordinary colds events seen in the Greenland temperature record are linked to Heinrich (HE) events (Heinrich 1988, MacAyeal 1993, Bond et al. 1993, Alley and MacAyeal 1994). During HE events, armadas of icebergs broke off from glaciers and traversed the North Atlantic. The icebergs contained rock mass eroded by the glaciers, and as they melted this matter was dropped onto the sea floor as ice rafted debris. During Heinrich events, huge volumes of fresh water entered the North Atlantic Ocean, thereby considerably slowing the Atlantic Meridional Overturning Circulation (AMOC) and reducing poleward heat transport, eventually causing widespread cooling in the Northern Hemisphere. Pronounced warm events were also reconstructed from proxy data, which are termed Dansgaard-Oeschger (DO) events.
Rather abrupt and large changes in global sea level have been observed during the transition from the last glacial maximum into the Holocene. The strongest of these pulses, the so called melt water pulse 1A or MWP-1A, led to sea level rise of about 20 m in approximately 340 years coeval with the onset of the Bölling warm period (Deschamps et al. 2012). Some studies that investigated the spatial patterns or “fingerprint” of sea level rise during the last deglaciation attribute much of the melt water to Antarctica. Different sources of ice melt leave geographically distinctive sea level patterns, because their ice unloading histories and gravitational pull between the shrinking ice masses and ocean vary. On the other hand, geological data indicate significant deglaciation in Antarctica starting only toward the end of MWP-1A, which suggests that most of the melt water originated from the breakup of Northern Hemisphere ice sheets. Gregoire et al. (2012) suggest, on the basis of ice sheet modeling, a mechanism in which the separation of the Laurentide and Cordilleran ice sheets in North America can produce a melt water pulse corresponding to MWP-1A.
The third challenge is to understand and simulate the changes in the biogeochemical cycles during the last glacial cycle. Of particular importance are the carbon and methane cycles. Fluctuations in physical properties (e.g. temperature) and the GHG concentrations basically varied in phase on multi-millennial and longer time scales (Fig. 1). A major issue in this respect is the cause of the variation of the CO2- concentration between interglacial and glacial periods on the order of 100 ppmv (Fig. 1). Comprehensive models are unable to explain such a large change, which demonstrates the lack in our understanding of the coupling between the physics and biogeochemistry. According to our present understanding, there is no single mechanism that could explain the range of the glacial-interglacial CO2 change (e.g. Sigman and Boyle 2000). Changes in ocean solubility or circulation, in marine productivity due to enhanced dust input, in marine CaCO3 budget or terrestrial biosphere, they all fail to account individually for the large CO2 change. The changes in the GHG concentrations constitute an important positive feedback, in the sense that an initial temperature perturbation in response to the Milankovitch forcing is reinforced by a change in the GHG concentrations which influence the atmospheric greenhouse effect.
One special concern with respect to anthropogenic climate change is the fate of the marine carbon sink, especially in the high-latitude regions. Changes in deep water production rates and oceanic mixing as anticipated for a warmer world will slow down the vertical mixing of water saturated with CO2. This will decrease the rate of marine carbon uptake and thus increase atmospheric CO2. Eventually, this will - through an enhanced greenhouse effect - increase the rate of global warming. However, we still lack capabilities to realistically simulate both the processes of deep water formation and the more biologically-dependent mechanisms of CO2 uptake in our models. Simulating the changes in GHG concentrations during the last glacial cycle with comprehensive ESMs requires a detailed process understanding to enhance our models. If we pass this test, we can have more confidence in the biogeochemical feedbacks incorporated in our ESMs used to conduct climate projections.
The fourth challenge concerns the synthesis of proxy data. Over the last decades, enormous amounts of individual climatic records from marine and terrestrial sediments, ice-cores, cave deposits, corals, tree rings and other accreted materials have been generated. These records are individually valuable, but their full potential is only realized when they are considered collectively. Their global synthesis can yield robust constraints on the actual sequence of events, magnitude of change and the spectrum of variability in the climate system as it occurred in the course of the last glacial cycle. As such, the synthesis will thus provide essential means of benchmarking simulations with comprehensive ESMs. To achieve such synthesis, individual proxy-records must be harmonized and aligned to a common time scale.
One important aspect of the work is described in the following example: Uncertainties due to the attribution of the proxy values and uncertainties resulting from temporal alignment of records must be quantified, reduced as far as possible and explicitly integrated into the subsequent data-model comparison. This represents a significant challenge for the portion of the record beyond the reach of the radiocarbon dating method as well as for the synchronization of marine and terrestrial records and requires the development of intelligent metrics capable of process-oriented diagnosis of model simulations against noisy proxy data associated with complex uncertainty.
Comprehensive Earth system models require enormous computer resources. This demand for computer resources naturally is amplified by the long-time horizon when simulating a full glacial cycle. Moreover, the research proposed in this project includes or introduces some aspects which are challenging also from the viewpoint of computational sciences as applied mathematics and informatics. Among these challenges are coupling schemes between different model components, coupled simulations on long time horizons and fine model level as well as sensitivity analysis and uncertainty quantification for model assessment and improvement.
As a consequence, the project poses research questions that are challenging in both climate and computational science. It thus offers an opportunity for – and will substantially benefit from – the development and integration of advanced strategies and techniques from numerical mathematics and computer science. These methods can be summarized under the term model optimization. In this project, optimization refers on the one hand to performance optimization both by introduction of new numerical algorithms and by exploitation of the potential of new hardware development (as e.g. accelerator cards). On the other hand, optimization refers to model quality itself. Methods from mathematics and computer science for model and software coupling as well as a reduction of the model-to-data misfit will improve the reliability and validity of the model results.
Model optimization will take place on different levels. It starts with improvement on the component level, i.e., in the different sub-models (e.g., ocean and atmosphere) of an Earth system model. Then, it will be extended across sub-model interfaces in coupled simulations. From another perspective, model optimization will be applied on different levels of complexity (e.g., for EMICs and comprehensive ESMs). This includes the propagation of sensitivity and uncertainty information between these levels and the development and assessment of reduced order models. Here again, mathematics and informatics will provide useful scientific input.
With these common research questions and scientific interactions, the project can be an example for successful interdisciplinary research between the communities of both climate and computational science.
What are the processes that govern the built-up and termination of ice sheets (with application to sea level)?
What is the dynamics of abrupt climate change? Are there thresholds in the Earth system and can we quantify them?
Can we realistically simulate the spectrum of the variability within a regime?
From forcing to response: what are the most important feedbacks?
What are the processes that govern the changes in greenhouse gas concentrations?
What are the implications for the future climate?
What is the best methodology to optimize comprehensive Earth system models?
Overall, we shall use a hierarchical approach in two ways to tackle these questions. First, a number of models of different complexity will be used to efficiently address the questions. Second, part of the research in the different sectors can be initially performed in parallel and synthesized after a few years.
The research is organized in 4 Working Groups (WGs) which basically reflect the challenges described above:
WG1: Physical system
WG2: Biogeochemical system
WG3: Proxy data synthesis
The central management body is a Steering Group consisting of the coordinators, all Working Group leaders and representatives of research organizations that are not otherwise represented in the Steering Group.
Furthermore, an international Scientific Advisory Board will be established to further strengthen international collaboration and to obtain independent scientific advice on the research agenda.
Milestones and deliverables as defined below in each Work Package (WP) will be continuously monitored by the coordinators. If necessary, the research plan will be adjusted.
The results will be published in international peer-reviewed scientific publications and thus available to the public. If possible, open access options will be used for all publications to comply with the “Berlin Declaration on Open Access to Knowledge in the Sciences and Humanities”, the rules of the Priority Initiative “Digitale Information” of the Alliance of German Science Organisations in Germany and the prevailing regulations of the German Ministry for Education and Research BMBF. The data sets and meta data will be stored with open access option in international data bases which are registered as full members of the ISCU World Data System and will thus be sustainably accessible.
Link to „Förderpolitische Ziele von FONA“
The proposed research fits the goals of the BMBF “Framework Programme Research for Sustainable Development (FONA)” addressing the global challenges posed by climate change and water shortage, the loss of biodiversity, soil degradation and shortage of resources and energy. Particularly relevant are the research areas: Global Change where one goal is “providing reliable analyses and predictions of trends” and System Earth aiming at “an understanding of global cycles and the interaction of geosphere”.
Link “to Horizon 2020”
Horizon 2020 is the research program underlying the implementation of the Innovation Union, a Europe 2020 flagship initiative aimed at securing Europe's global competitiveness. Our research fits in Part III of the Work Programme – Climate action, resource efficiency and raw materials.
Alley RB, MacAyeal DR (1994) Ice-rafted debris associated with binge/purge oscillations of the Laurentide Ice Sheet. Paleoceanography 9: 503–512.
Alvarez-Solasa J, Ramstein G (2011) On the triggering mechanism of Heinrich events. PNAS 108: doi: 10.1073/pnas.1116575108.
Berger WH (1990) The younger dryas cold spell - a quest for causes. Global and Planetary Change 3: 219-237.
Bond G et al (1993) Correlations between climate records from North Atlantic sediments and Greenland ice. Nature 365: 143-147.
Bond G et al (1997) A Pervasive Millennial-Scale Cycle in North Atlantic Holocene and Glacial Climates. Science 278 (5341): 1257-1266.
Bond G et al (2001) Persistent Solar Influence on North Atlantic Climate During the Holocene. Science 294 (5549): 2130-2136.
Bothe O, Jungclaus JH, Zanchettin D, Zorita E (2013) Climate of the last millenium: Ensemble consistency of simulations and reconstructions. Climate of the Past: 9, 1089-1110. Broecker WS et al (2010) Putting the Younger Dryas cold event into context. Quaternary Science Reviews 29: 1078-1081.
Brook E (2008) Windows on the greenhouse. Nature 453: 291-292.
Clark PU et al (2012) Global climate evolution during the last deglaciation. PNAS, 109(19): E1134-E1142.
Cuffey KM, Clow GD (1997) Temperature, accumulation, and ice sheet elevation in central Greenland throughout the last deglacial transition. J. Geophys. Res. 102: 383-396. Deschamps P (2012) Ice-sheet collapse and sea-level rise at the Bolling warming 14,600 years ago. Nature, 483(7391): 559-564.
Gregoire LJ et al (2012) Deglacial rapid sea level rises caused by ice-sheet saddle collapses", Nature, 487: doi:10.1038/nature11257.
Hasselmann K (1976) Stochastic climate models. Part I. Theory. Tellus 28: 473-485.
Hawkins E, Sutton R (2009) The Potential to Narrow Uncertainty in Regional Climate Predictions. Bull. Amer. Meteor. Soc.: 90, 1095-1107.
Hays JD, Imbrie J, Shackleton NJ (1976) Variations in the Earth's Orbit: Pacemaker of the Ice Ages. Science 194 (4270): 1121-1132.
Heinrich, H (1988) Origin and consequences of cyclic ice rafting in the northeast Atlantic Ocean during the past 130,000 years. Quat. Res.: 29, 142-152.
Huybers P, Curry W (2006) Links between annual, Milankovitch and continuum temperature variability. Nature 441: 329-332.
Imbrie J et al (1992) On the Structure and Origin of Major Glaciation Cycles 1. Linear Responses to Milankovitch Forcing. Paleoceanography 6: 701-738.
IPCC (2007) Climate Change 2007: The Physical Science Basis. Working Group I contribution to the IPCC Fourth Assessment Report (WGI AR4).
IPCC (2013) Climate Change 2013: The Physical Science Basis. Working Group I contribution to the IPCC Fifth Assessment Report (WGI AR5).
Lüthi D et al. (2008) High-resolution carbon dioxide concentration record 650,000–800,000 years before present. Nature 453: 379-382.
MacAyeal DR (1993) Binge/purge oscillations of the Laurentide Ice Sheet as a cause of the North Atlantic's Heinrich events. Paleoceanography 6: 775-784.
NEEM community members (2013) Eemian interglacial reconstructed from a Greenland folded ice core. Nature 493: doi:10.1038/nature11789.
Rahmstorf S (2002) Ocean circulation and climate during the past 120,000 years. Nature 419: 207-214.
Schulz M, Paul A, Timmermann A (2002) Relaxation oscillators in concert: A framework for climate change at millennial timescales during the late Pleistocene. Geophys. Res. Lett. 29, doi:10.1029/2002GL016144.
Schulz M, Paul A, Timmermann A (2004) Glacial–interglacial contrast in climate variability at centennial-to-millennial timescales: observations and conceptual model. Quaternary Science Reviews 23: 2219-2230.
Sima A, Paul A, Schulz M (2004) The Younger Dryas - an intrinsic feature of late Pleistocene climate change at millennial timescales. EPSL 222: 741-750.
Stocker TF (1998) The seesaw effect. Science 282 (5386): 61-62.