The current rate of atmospheric increase of CO2—~3 ppmv per year (IPCC, 2007)—is an order of magnitude or more greater than the increase in atmospheric CO2 during the last deglaciation, when rapid retreat of northern hemisphere ice sheets led to rates of sea level rise of up to 5 m per century (Stanford et al., 2006). The last time the atmosphere contained CO2 levels comparable to today’s values, during the Pliocene, surface temperatures were on average ~3°C warmer, the Greenland ice sheet collapsed, and sea level rose by up to 30 m (Pagani et al., 2010; Seki et al., 2010). With the combination of continued burning of fossil fuels and the additional contribution of greenhouse gases to the atmosphere through positive feedbacks in the climate system, future atmospheric CO2 levels could exceed 1,000 ppmv (Kump et al., 2009)—levels well above the stability threshold values for continental ice on Earth (Hansen et al., 2008). In fact, it is necessary to look back at least 34 million years—prior to the current icehouse—to examine climate change under such CO2 levels. In this context, the magnitude and rate of the present greenhouse gas increase place the climate system in what could be one of the most severe increases in radiative forcing of the global climate system in Earth history.
To fully evaluate climate forcing feedbacks and tipping points that may characterize Earth’s future, and to better understand climate change impacts and recovery, it is necessary to examine the records from past warm periods when there were similar magnitudes and rates of greenhouse gas forcing. The deep-time paleoclimatology record contains a rich archive of such warm worlds, and the associated transitions into and out of “greenhouse” conditions. For example, climate reconstruc-
tions of the end-Paleocene (~55 Ma), mid to late Cretaceous (~120 to 90 Ma), end-Triassic (~200 Ma), and Late Paleozoic (~300 to 251.2 Ma)—all periods associated with the massive release of greenhouse gases to the atmosphere—reveal dramatic changes in oceanic conditions and terrestrial climates. These changes brought about extensive restructuring of marine and terrestrial ecosystems that in many cases involved mass extinctions. These deep-time records also reveal that some of the feedbacks in the climate system may be unique to warmer worlds—and thus are not archived in more recent paleoclimate records—and accordingly might be expected to become increasingly relevant with continued warming. In particular, long-term feedbacks that are typically active on millennial scales are likely to become important at the human timescale, leading to substantial and abrupt (years to centuries) climate modifications. Reconstructions of past climates show that civilization has evolved in an anomalously stable period unrepresentative of the climate system’s natural variability. Therefore, refining current understanding of climate dynamics (e.g., the range, rates, and magnitudes of feedbacks and change) during past periods of global warming, particularly times associated with epic deglaciations, is critical for assessing future risks. Improved understanding of climate dynamics will also aid efforts to mitigate the impact of continued warming on regional hydroclimates and water resources, ice sheet and sea level stability, and the health of marine and terrestrial ecosystems. Exciting research opportunities to help accomplish this task exist in the untapped potential of the deep-time geological record.
This report identifies a six-element research agenda designed to describe past climate variability and to better constrain how Earth’s climate system has responded to episodes of changing greenhouse gas levels. The knowledge gained by this scientific agenda will be important for addressing questions regarding the projected rise in atmospheric CO2 and the societal implications of this rise. The report also describes the research infrastructure necessary for successful implementation of the deep-time paleoclimatology agenda, as well as an education and outreach strategy designed to broaden our collective understanding of the unique perspective that the full range of the geological record provides for future climate change.
Improved Understanding of Climate Sensitivity and CO2-Climate Coupling
Determining the sensitivity of Earth’s mean surface temperature to increased greenhouse gas levels in the atmosphere is a key requirement for estimating the likely magnitude and effects of future climate change. The current understanding of climate sensitivity, defined on the basis of
modern data and relatively recent paleoclimate records (≤20,000 years), is associated with large uncertainty (1.5 to ≥6°C). Positive feedbacks typically considered to have been active on longer timescales, but that may become increasingly relevant with continued warming, are not considered in these estimates. An improved definition of long-term equilibrium climate sensitivity—including more refined constraints on its lower boundary—over the full range and timescales of past radiative forcing is a major research priority. An associated focus is on gaining an improved understanding of how climate feedbacks and their role in amplifying climate change have varied with changes in greenhouse gas forcing. Accomplishing this objective will require the development of more accurate and precise paleo-CO2 and paleotemperature proxies, as well as the development of new proxies for the full range of greenhouse gases. A complementary requirement is for high-resolution and high-precision time-series records, based on integrating multiproxy techniques. Data-model comparisons are needed to rigorously test non-CO2 forcing mechanisms of global warming, as well as to refine the understanding of how the Earth’s climate system would respond to increasing levels of atmospheric CO2.
Climate Dynamics of Hot Tropics and Warm Poles
Recent climate modeling and deep-time paleoclimatology studies have demonstrated that the long-standing paradigm that the temperatures of tropical climates do not rise significantly during warm periods because of some type of temperature buffering mechanism is probably incorrect. Consequently, the mechanisms and feedbacks in the modern climate system that have controlled tropical and polar surface temperatures—ultimately leading to the existing relatively high pole-to-equator thermal gradient—may not operate in warmer worlds. A decreased latitudinal gradient in the future, which would almost certainly be associated with polar sea ice and continental ice sheet losses, would change atmospheric wind patterns and, in turn, ocean circulation—all having potential detrimental effects through teleconnections (Hay, 2010). To refine knowledge of the processes and climate feedbacks that may influence surface temperatures under higher atmospheric pCO2, it is important that high-temporal-resolution, higher-precision proxy time series be developed across latitudinal transects, with a focus on reconstructing terrestrial-marine linkages. This will require a greatly increased effort in high-precision geochronological dating, coupled with substantially more spatially resolved proxy records. A more comprehensive understanding of the limits of tropical climate stability, the origin of anomalous polar warmth, and an understanding of how a weaker thermal gradient is established and maintained in warmer climate regimes will require further climate model development and deep-time
data-model comparisons. These comparisons would also provide a much needed test of the efficacy of model projections of future climate.
Sea Level and Ice Sheet Stability in a Warm World
Study of the current icehouse climate state has provided better constraints on CO2 and surface temperature threshold levels for ice sheet stability (Pagani et al., 2005, 2010; Pearson et al., 2009; Seki et al., 2010). Large gaps, however, remain in the understanding of ice sheet dynamics, with resulting limitations on the applicability of current coupled climate-ice sheet models. These issues highlight the uncertainties that still exist in projections regarding the timescales at which ice sheets would respond to continued warming and in understanding the influence of feedbacks not revealed by recent paleoclimate records or considered by future climate model projections (e.g., the projections used in IPPC, 2007). Consequently, the magnitude of sea level rise, once climate equilibrium is reached, remains elusive despite deep-time paleoclimate evidence that it could be substantially higher than model projections (Rohling et al., 2009). To markedly improve the understanding of climate–ice sheet–sea level dynamics relevant to a warming Earth, it will be necessary to probe deeper into Earth’s history to the periods of truly catastrophic ice sheet collapse that accompanied past icehouse-to-greenhouse transitions. To fully exploit such deep-time archives will require radiometrically constrained and spatially resolved marine, paralic, and terrestrial records for both high and low latitudes. In addition, improved methods for deconvolving temperature and seawater δ18O from proxy records are needed, as well as targeted efforts to couple land-ice component models with complex global climate models that are capable of integrating the atmospheric hydrological cycle.
Understanding the Hydrology of a Hot World
There is broad scientific consensus that one of the largest impacts of continued CO2 forcing would be major regional climate modifications, with the likelihood of substantial societal impacts (e.g., water shortages, flooding). The insights gained from reconstructing the processes and climate feedbacks that influence surface temperatures under higher atmospheric pCO2 levels are an important element of this research agenda, particularly because of the sensitivity of climate to small changes in high-latitude and tropical surface temperatures as a consequence of teleconnections. The deep-time geological record provides a critical and unique component of research focused on this issue, because it is the only source of information regarding how marine-terrestrial carbon and water cycle dynamics have influenced the global climate system during periods of radiative forcing
comparable to those projected for the future, including periods of unipolar glacial or fully deglaciated greenhouse conditions. This will require a greatly expanded effort to develop linked marine-terrestrial records that are spatially resolved and of high temporal resolution, precision, and accuracy. New and improved quantitative estimates of paleoprecipitation, paleoseasonality, paleoaridity and humidity, and paleosoil conditions (including paleoproductivity) are critical components of this effort.
Understanding Tipping Points and Abrupt Transitions to a Warmer World
Studies of past climates and climate models show that Earth’s climate system does not respond linearly to gradual CO2 forcing, but rather responds by abrupt change as it is driven across climatic thresholds. Modern climate is changing very rapidly, and there is a possibility that Earth will soon pass thresholds that will lead to even more rapid changes in Earth’s environments. Consequently, the question of how close Earth is to a tipping point, and when it could transition into a new climate state, is of critical importance. Because of their proven potential for capturing the dynamics of past abrupt changes, intervals of tipping-point climate transitions in the geological record—including past hyperthermals—should be the focus of future collaborative paleoclimate, paleoecological, and modeling studies. Such studies should lead to an improved understanding of how various components of the climate system responded to abrupt transitions, in particular during times when the rates of change were sufficiently large to imperil diversity. This research will also help determine whether there exist thresholds and feedbacks in the climate system of which we are currently unaware, especially in warm worlds and past icehouse-to-greenhouse transitions. Moreover, targeting such intervals for more detailed investigation is a critical requirement for constraining how long any abrupt climate change might persist.
Understanding Ecosystem Thresholds and Resilience in a Warming World
Both ecosystems and human society are highly sensitive to abrupt shifts in climate, because such shifts may exceed the tolerance of organisms and, consequently, have major effects on biotic diversity as well as human investments and societal stability. Modeling future biodiversity losses and biosphere-climate feedbacks, however, is inherently difficult because of the complex, nonlinear interactions with competing effects that result in an uncertain net response to climatic forcing. How rapidly biological and physical systems can adjust to abrupt climate change is a
fundamental question accompanying present-day global warming. An important tool to address this question is to describe and understand the outcome of equivalent “natural experiments” in the deep-time geological record, particularly where the magnitude and/or rates of change in the global climate system were sufficiently large to threaten the viability and diversity of species, which at times led to mass extinctions. The paleontological record of the past few million years does not provide such an archive because it does not record catastrophic-scale climate and ecological events. As with the other elements of a deep-time research agenda, improved dynamic models, more spatially and temporally resolved datasets with high precision and chronological constraint, and data-model comparisons are all critical components of future research efforts to better understand ecosystem processes and dynamic interactions.
Four key infrastructure and analytical elements will be required to implement this high-priority research agenda.
Improved Proxies and Multiproxy Records
Refinement of existing and development of new mineral and organic proxies for environmental and ecological parameters, coupled with an enhanced effort to chronologically calibrate targeted intervals with high-precision radiometric ages, are critical requirements for developing the spatially resolved, multiproxy paleoclimate and paleoecological time series described in the research agenda.
Despite exponential advances in the development of paleoclimate proxies over the past two decades, the precision and accuracy of existing organic and mineral paleotemperature and paleo-CO2 proxies are compromised by their calibrations to extant analogues, by incompletely understood biological and environmental controls on geochemical signatures, and/or by their sensitivity to postdepositional alteration. Moreover, paleobarometer proxies are limited to CO2, and there is a need for the existing very limited complement of proxies for estimating past terrestrial climatic conditions to be expanded and refined. A focused effort to improve existing proxies and develop new proxies is at the core of the proposed research agenda, in particular where the level of precision and accuracy—and thus the degree of uncertainty in inferred climate parameter estimates—can be quantified and significantly reduced. Such an effort will need to be
highly collaborative, requiring calibration studies in modern marine and terrestrial environments as well as laboratory systems. The Critical Zone Observatories initiative funded by the National Science Foundation may offer opportunities to integrate such calibration studies into existing observatories. Ultimately, comparison studies of plant-mineral proxy estimates that are characterized by differing sensitivities and uncertainties are necessary to test the veracity and sensitivity of each of the proxies. Proxy development efforts must be complemented by studies that apply emerging imaging and analytical technology to critically evaluate the effects of postdepositional alteration on the compositions of isotopic and geochemical proxies.
Deep-Time Drilling Transects
The recovery of high-quality cores to provide the sample resolution and preservational quality needed to develop multiproxy archives for key paleoclimate targets across terrestrial-paralic-marine transects and latitudinal or longitudinal transects will require substantially increased investment in scientific continental drilling and continued support for scientific ocean drilling. Continental drilling will permit direct comparison of the marine and terrestrial proxy records that record fundamentally different climate responses (local and regional effects on continents compared with homogenized oceanic signals) and will provide the continuous records necessary for high-resolution dating of critical climate transition intervals.
The requirement for well-preserved and chronologically well-constrained proxy records with high spatial and temporal resolution and precision to analyze environmental and ecological systems in climate transition is a recurrent theme throughout the research agenda. A transect-based deep-time drilling program designed to identify, prioritize, drill, and sample key paleoclimate targets—involving a substantially expanded continental drilling program and additional support for the existing scientific ocean drilling program—is a high priority for implementing the recommended research agenda. Although scientific ocean drilling has provided much of the basis for what is presently known about Neogene climate dynamics and ocean-climate linkages, there is still a pressing need for high-resolution sections that carry clear signals of orbital forcing in older parts of the record, particularly the Paleogene and Cretaceous. Sections representing the greenhouse intervals for climatically sensitive regions are still required, specifically in the Arctic and proximal to Antarctica. Continental drilling of cyclic successions, of extended duration
and with high potential for preservation of volcanic ashes, will greatly expand the opportunity for radiometric and nonradiometric dating and correlation, thereby facilitating comparison of paleoclimate records across marine-paralic-terrestrial gradients as a function of time.
Improved Paleoclimate Modules and Models
An enhanced paleoclimate modeling effort, with a focus on past warm worlds and extreme and/or abrupt climate events, is critical for refining scientific understanding of the complex dynamics of past climates and for producing models that can be adjusted to include forcings or feedbacks not revealed by shallower-time paleoclimate reconstructions.
As critical boundary conditions of the climate system—greenhouse gas concentrations, polar ice mass, distribution of biomes—change in the coming century, calibrations of climate models based on modern systems and the recent past will become increasingly less relevant. The deep-time geological record of past climates and major transitions provides the only test of climate models and their predictions against the range of background conditions most likely to be relevant to Earth’s anticipated future climate state if emissions are not reduced. Modeling of ancient climates characterized by boundary conditions substantially different from those of the present day, however, presents a substantial challenge to the modeling community. In turn, how well such models simulate past climates and feedbacks inferred from deep time influences the community’s confidence in the ability of global climate models to forecast future regional and global climate changes.
To that end, a markedly enhanced effort in deep-time paleoclimate modeling involving development of higher-resolution modules, improved parameterization of conditions relevant to future climate, and an emphasis on paleoclimate model intercomparisons and “next-generation” data-model comparisons is a fundamental component of the proposed research agenda. An increase in model spatial resolution will be required to capture smaller-scale features and regional climate changes comparable in scale to the spatially resolved geological data that can be obtained through continental drilling and proxy development. Deep-time data also uniquely offer the opportunity to carry out model-model-data comparisons for past warm climates characterized by elevated CO2. Such comparisons will permit an assessment of the credibility of the performance and parameterizations of various community models in a way that future climate experiments are presently incapable of doing. Achieving this component of the deep-time initiative will require new tools to facilitate model-data
comparisons (e.g., prognostic modules for proxies, geographic information system–based tools, refined dynamic vegetation models, metadata techniques), dedicated computational resources for deep-time climate simulations, and the development and application of Earth system models of intermediate complexity that can be integrated as subsystem models within more complex three-dimensional Earth system models.
Strategies for Fostering Focused Deep-Time Scientific Interaction
Implementing the research agenda described in this report will require a synergistic research culture among the broad range of disciplines that can contribute to solving the numerous puzzles of deep-time paleoclimatology, focusing on specific paleoclimate time slices as natural laboratories for team-based analyses of deep-time climates and their impact on Earth systems. Establishment of a cultural and technological infrastructure to support team-based projects offers the potential for discoveries unattainable by single-discipline research or even by more conventional integrated efforts.
Establishing the scientific collaboration, cross-disciplinary syntheses, widespread and open data exchange, cross-training of scientists and students, and dedicated and focused outreach activities required to address the research agenda described in this report will require the development of natural observatories for team-based studies of important paleoclimate time slices, incorporating climate and geochemical models; capabilities for the development, calibration, and testing of highly precise and accurate paleoclimate proxies; and the continued development of digital databases to store proxy data and facilitate multiproxy and record comparisons across all spatial and temporal scales. Such broad-based and interdisciplinary cultural and technological infrastructure will require acceptance and endorsement by both the scientific community and the funding agencies that support deep-time paleoclimatology and paleobiology-paleoecology studies. Without the addition of targeted new resources—in addition to existing programmatic resources—the scientific breakthroughs that can be made by this broad-based research community will be unlikely to ever come to fruition.
Despite the potential and importance of the deep-time geological record, as articulated throughout this report, the public has minimal appreciation of the relevance of deep-time climates for Earth’s future. This
largely reflects the limited efforts by the scientific community to ensure that the importance and relevance of scientific efforts and results are conveyed to students, teachers, scientific and media partners, policy makers, and the general public. Barriers such as disciplinary jargon (geological time, paleoclimate proxies, and numeric climate models), imperfect interpretations and solutions created by uncertainties in temporal resolution, patchwork spatial resolution, and incompletely calibrated climate proxies, all present significant challenges for conveying complex messages to the general public with sufficient simplification but without losing accuracy. To resolve this issue, a strategy for education and outreach, to convey the lessons contained within deep-time records, should be tailored to the range of specific target audiences:
• K-12 elementary and secondary students. Museums are a key resource for educating students. Involving teachers in scientific endeavors can help demystify science and convey the excitement of scientific discovery, as well as being a method of disseminating scientific information.
• For colleges and universities, distinguished lecture tours, topical summer schools, and the integration of deep-time paleoclimatology into traditional and nontraditional earth science courses offer additional opportunities to convey the relevance of the deep-time record.
• To involve and educate the general public, the deep-time observation and modeling communities have opportunities to break into the popular science realm by emphasizing their more compelling and understandable elements. Immediate opportunities to illustrate “deep-time paleoclimatology in action” to the general public abound, whether the irreversible impact of past major climate changes on life, extreme glaciations and catastrophic deglaciations, or the mysteries of the ocean. The scientific community needs to proactively pursue pathways to the public provided by various multimedia opportunities.
• Potential scientific collaborators from the broader climate science community can obtain increased understanding of the potential offered by paleoclimate data and modeling through the creation or use of forums where scientists from different disciplines exchange information and perspectives. This can be effectively done between disciplines at meetings of broader groups (e.g., American Association for the Advancement of Science) and industry, environmental, ecology, and physical anthropology conferences.
• Policy makers require scientifically credible and actionable data on which to base their policies. Faced with a diversity of opinions, they need credible sources of information. This report and other National Research Council reports attempt to play this role, but in a much broader
sense the scientific community must strive to make the presentation of deep-time paleoclimate information as understandable as possible.
The paleoclimate record contains facts that are surprising to most people. For example, there have been times when the poles were forested rather than being icebound; there were times when the polar seas were warm; and there were times when tropical forests grew at midlatitudes. For the majority of Earth’s history, the planet has been in a greenhouse state rather than in the current icehouse state. Such concepts provide an opportunity to help disparate audiences understand that the Earth has archived its climate history and that this archive, while not fully understood, provides crucial lessons to improving our understanding of Earth’s climate future.
The possibility that our world is moving toward a “greenhouse” future continues to increase as anthropogenic carbon builds up in the atmosphere, providing a powerful motivation for understanding the dynamics of Earth’s past “greenhouse” climates that are recorded in the deep-time geological record. It is the deep-time climate record that has revealed feedbacks in the climate system that are unique to warmer worlds—and thus are not archived in more recent paleoclimate records—and that might be expected to become increasingly relevant with continued warming. It is the deep-time record that has revealed the thresholds and tipping points in the climate system that have led to past abrupt climate change, including amplified warming, substantial changes in continental hydroclimate, catastrophic ice sheet collapse, and greatly accelerated sea level rise. Further, it is uniquely the deep-time record that has archived the full temporal range of climate change impacts on marine and terrestrial ecosystems, including ecological tipping points. An integrated research program—a deep-time climate research agenda—to provide a considerably improved understanding of the processes and characteristics over the full range of Earth’s potential climate states offers great promise for informing individuals, communities, and public policy.