Overview and Recommendations
The long record of climate on the Earth is valuable in understanding how the climate system works. The study of past climates can conveniently be viewed on three time scales: (1) the last 10,000 years; (2) the period back to 2 million years ago (Ma)—the Pleistocene Epoch, which witnessed the rise of man; and (3) the pre-Pleistocene period prior to 2 Ma. Much attention has already been devoted to the records of climatic change in the first two intervals; this study is devoted to the third, the vast (billions of years) pre-Pleistocene period of Earth history. The climate record is contained in strata and rocks of the continental crust and in sediments of the oceans. By examining this record we can learn much about the long-term changes in climate, what the state of climatic normalcy has been back into remote periods of geologic time, and what some of the factors are that perturb this state and by how much. We offer here a summary of achievements in understanding ancient climates and of current research problems and make recommendations concerning profitable avenues for future research.
Climate is the result of flow of the air and ocean system on the rotating Earth in accordance with the laws of physics. These fluids flow and interact within a geographic setting determined primarily by the arrangements of land and sea, the orientation of mountains and lowlands on continents, the depths of the oceans, and the
location of ocean gateways. The total system is complex, and climate is the result of the interplay between the air, ocean, and land on a range of time scales. These scales range from the short period, dealt with in weather forecasts, to long-period changes resulting from changes in atmospheric composition and the drift of continents on the mobile lithosphere.
Meteorologists are currently developing computer models of the atmosphere that give increasingly accurate descriptions of the global circulation. Through the use of such models in the decades ahead, they will explain the flow of air and the movement of weather patterns. In addition, oceanographers are attempting to understand the way the ocean works. In time their models can be combined with atmospheric models leading to general-circulation models of both the air and the ocean waters. Most changes within the ocean system take place over many centuries, so the study of oceanography aimed at such time scales as well as over shorter times must be melded into the framework of the weather and climate models, which deal with changes over days, weeks, months, and years. Paleo-oceanography, concerned with describing the ocean system back farther into time, can be expected to reveal how longer-term components of the ocean play their part in the way the climate system works.
Geologists and geophysicists, employing the concepts of plate tectonics and continental drift, are now reconstructing past arrangements of lands and seas for the last several hundred million years. These reconstructions are providing boundary conditions for investigations by paleo-oceanographers and paleoclimatologists. In the near future they will lay out the framework for computer modeling of ancient climates. With enough information from sediments on the seafloor and strata within the continents, an iterative approach promises to disclose how changing land-sea patterns and topography and bathymetry influence the flow of ocean waters, which, in turn, strongly influence the circulation pattern of the atmosphere. By comparing differences in the factors controlling past climates we will learn of the checks and balances in the system and of the complex feedbacks, overshoots, and dampenings. We will learn what physical mechanisms bring the perturbed climatic system back to a near-steady state. The climatic record contained in strata will provide information on what past climates were like, on how quickly they changed, and on how large these changes were. Nature has performed a large number of experiments during the course of Earth history, and the geologic record contains the outcome of these experiments.
In this Overview, we summarize key factors for the understanding of climates in particular periods in the pre-Pleistocene and various approaches (synoptic, time-series, and event analyses) for investigating ancient climates. We also make several specific recommendations for future investigations. Detailed discussions of recent advances in understanding ancient climates appear in the chapters that follow this Overview.
Our principal goal is to encourage specialists in all the earth sciences to find ways that their disciplines can contribute to understanding the complex atmosphere-oceanland interactions that constitute the climate system and its variations through time.
WHY STUDY ANCIENT CLIMATES?
Climate, its variation, and its change are closely linked to the growth of our crops, to our well-being, and to the economy. Consequently, predictions or forecasts of short-term climate changes promise substantial benefits. Recognizing this potential, Congress recently enacted the National Climate Program Act, which calls for “a well-defined and coordinated program in climate-related research, monitoring, assessment of effects, and information utilization” (U.S. Congress, Public Law
95–367, 1978). Improvement in the understanding of climate has become urgent in view of the potential for inadvertent climate modification by our industrial society. Heat and particulates are released to the atmosphere and affect climate both locally and regionally, and perhaps even globally. The greatest potential global impact is from carbon dioxide release (NRC Geophysics Study Committee, 1977; NRC Climate Research Board, 1979; NRC Climate Board, 1982). At present, the rate of carbon dioxide added to the atmosphere is about 0.7 percent per year of the atmospheric carbon dioxide content. The input of industrial carbon dioxide has been called a “geophysical experiment” of unprecedented scope—an experiment whose course and outcome is unplanned and unknown. On the whole, a global warming over the next century is indicated: a doubling of atmospheric carbon dioxide would raise the global temperature by some 2 to 3°C (Hansen et al., 1981; NRC Climate Board, 1982).
To obtain the necessary perspective on present climatic conditions, we study ancient climates. How stable is the present climate, and how has it changed with time? How fast might the present climate respond to pertubation? What might be the effects of rapid change of climate on the biosphere? Any changes in the biosphere, consisting namely of changes brought about by human activities in agriculture and in the forests, will have profound economic and political consequences. Because modeling of the climate system over time spans more than a decade is still rudimentary, appraisal of rates of climatic change need to be extracted from the historical and geologic record.
Many important materials owe their origins to the interplay of climatic variables both in the oceans and on land. Petroleum and natural gas, for example, have at times and at places originated where wind-driven upwelling of deep ocean water encouraged the flourishing of microscopic algae, resulting in organic-rich muds on the seafloor. Marine phosphate deposits, important as fertilizers in crop production, were also largely associated with regional upwelling. Bauxite and laterites, prime sources of aluminum and nickel, respectively, resulted from extensive deep weathering of crystalline rocks under warm and humid conditions. Manganese deposits and coal also require specific climatic conditions for their origin. Coal deposits derive from swamps, generated in a humid climate. Much of our knowledge about past climates results from the search for such economic deposits, and our ability to locate additional resources will be improved by an increased understanding of climates throughout Earth history.
The geologic record shows that surface temperatures on Earth have not been too different over most of the Precambrian and Phanerozoic times from those of today (Frakes, 1979). Sedimentary rocks deposited 3500 Ma and since show that water has been predominantly present in its fluid state in the past as it is today. The air-ocean system has been driven by the Sun’s energy as it is today. Life flourished and kept evolving within the seas and later upon the lands. Climate has been remarkably stable when viewed in this time perspective.
Paleoclimatic investigations can be grouped into three general time periods on the basis of the precision with which the climatic record can be read, the tools and techniques available for study, and the applicability to future climate forecasts. Studies of the most recent period make use of instrumental, historical, and high-resolution proxy records and are the most definitive. Proxy data, such as tree rings, glacial records from ice cores, and pollen distributions, have extended our detailed climate knowledge back some 10,000 years ago. The goals and challenges of these
types of investigations have been covered in previous reports, for example, Understanding Climatic Change (NRC U.S. Committee for the Global Atmospheric Research Program, 1975).
The second period is the Pleistocene Epoch, a time of alternating glacial and interglacial stages that extends back to about 2 Ma. Major advances in understanding the climate of this interval have resulted from quantitative approaches similar to those used in the CLIMAP (Climate, Long-Range Investigations, Mapping and Prediction) program. Recent results of such investigations are summarized by Imbrie and Imbrie (1979).
The subject of this report is the third period, the pre-Pleistocene (Figure 1), a period that encompasses the entire geologic record older than 2 million years (m.y.). The goals pursued through the study of pre-Pleistocene climate are as varied as those of historical geology itself, as indicated by the range of the specific contributions in this study.
PRE-PLEISTOCENE CLIMATOLOGY: A MATTER OF SCALE
The central task of paleoclimatology is to describe the climatic patterns throughout geologic time and to understand the trends, cycles, and discontinuities in these patterns—in short, to find out how the climate system works.
Weather elements include temperature, precipitation (as either rain or snow), wind strength and direction, cloud cover, and humidity; climate consists of weather patterns in time and space. In human affairs, climate usually refers to the sum of weather patterns over periods of months to a few years. Changes in some weather elements from one decade to the next, such as changes bearing on the occurrence of droughts, are of course significant. In reconstructing past climates from the geologic record, however, we are forced to integrate over longer and longer intervals as we go back in time. In general, the further back we go, the less sure we are of the ages of
our sample and of the spans of time they represent. Because of the increasing number of unknowns in progressively older records, the type of questions we can hope to answer must change with the age of the record. We can conveniently consider the available record of pre-Pleistocene climates on five different time scales.
The most detailed climatic records can be constructed for the time during which the continents were essentially in their present positions and, as today, the polar areas had snow and ice. The onset of northern continental glaciations, about 3 Ma (Berggren, 1972; Shackleton and Opdyke, 1977) marks the beginning of this period. Essentially, the climatic mechanisms of the Pleistocene (1.7 Ma to the present) apply.
The second scale applies to the time during which we have an adequate sedimentary deep-sea record. This period spans about the last 100 million years. Paleontology and geochemistry of the deep-sea sediments provide detailed climatic signals on a global basis. In addition, the positions of continents and the morphologies of ocean basins—both of which affect the atmospheric and oceanic circulation—can be reconstructed from seafloor paleomagnetic information. Because there are more uncertainties on this scale than on that in 1, above, our understanding of climatic change within this period will necessarily be less complete. However, some of the questions that can be asked regarding changes, variations, and repetitions over relatively long periods will lead to new insights not derivable from Pleistocene studies.
The third scale applies to the last 200 Ma or the time since the disruption of Pangaea, the early Mesozoic supercontinent that made up of most of the Earth’s land masses. Although there are several reconstructions of Pangaea, the differences refer largely to detail: the broad outlines of paleogeography are reasonably well known (McElhinny and Valencio, 1981). A deep-sea sedimentary record exists for some of this period but is much less complete than that of the 0–100 Ma record.
The fourth scale includes the entire Phanerozoic, nearly 600 m.y., the time for which we can read climatic zonation from biogeography, aided by geology and geophysics (positioning of the continents), geochemistry (mapping of climate-sensitive deposits), and interpretations based on data from strata incorporated in the continents. It includes all of Paleozoic time.
The fifth scale is applicable to Precambrian time and includes the entirety of Earth history except for the last one seventh. Climatic information exists for this scale but is scarce and commonly imprecise. On this scale, we see the chemistry of the atmosphere change in response to the evolution of organisms. On the whole, oxygen content increases and CO2 content decreases.
On each of the time scales, we are interested in learning what were the inputs to the climate system, its boundary conditions, its inner workings, and the way it expressed itself in the geologic record. The levels of detail of description and of understanding differ considerably for the various scales considered. For the most remote intervals, such as those within the Precambrian, statements about typical temperature ranges on the surface of the Earth and about limits on atmospheric composition and on solar radiation may be all that can be hoped for. For the Paleozoic Era, fossils provide clues, and it is possible both to recognize climatic zonations and to estimate the rates of change of such zonations. Yet even within these remote times, there may be glimpses into climatic cycles, based on continuous sequences of finely layered rocks. We may be able, for example, to test the concept of the Sun being a stable star for thousands of millions of years.
Continental configurations of the Mesozoic Era are reasonably well known. However, the geologic-geochemical setting during this era is different enough from
today’s so that we cannot unambiguously reconstruct pole-to-equator temperature gradients or humidity patterns. From the middle Cretaceous on, we have the advantage of extensive deep-sea records of the necessary time constants that contain detailed climatic signals of global significance. The record of the last 100 Ma will therefore be most productive in our search for the sequences of climatic change and their causes.
The Tertiary Period, and especially the Neogene part, is well represented by deepsea sediments, and it is possible to extract information on details such as temperature anomalies within latitudinal bands.
Paleoclimatic investigations can be conveniently characterized by three approaches: synoptic studies, time-series studies, and event and episode analysis, as summarized below.
One way to study ancient climates is to select convenient intervals of past geologic time and to assemble all pertinent information bearing on climate within that interval, that is, to study synoptic intervals or time slices. The length of the time slice selected is determined by the time control available, and it should be short enough so that climate change is minimal through its duration. Difficulties of correlation and the fact that the record is more complete for younger periods than for older ones mean that time intervals are longer and less truly synoptic as we go back in time. The crucial ingredients of synoptic analyses are reliable reconstructions of continental positions and accurate stratigraphic correlation from one continent to another and from continents to ocean basins.
The comparison of reconstructed climatic zones with present-day belts has been a traditional approach in such studies. For example, interpretation of the latitudinal distribution of coral reefs, evaporite deposits, coal deposits, and glacial moraines has relied on present-day analogies; this is also true for studies of the biogeography of plant and animal fossils. This analog approach involves the establishment of boundary conditions such as the distributions of continental landmasses, continental shelves, and ocean basins; the presence and extent of snow and ice (Figure 2); the overall pattern of surface albedo (insofar as it is dependent on snow and plant cover and the area of the seas and oceans); and the sizes and fluxes among the important carbon reservoirs influencing the carbon dioxide content of the atmosphere. The distribution of land heights and the location of mountain chains are also part of the boundary conditions. The distribution of climatic indicators and of theoretically expected belts of temperature and humidity can then be compared and the reasons for mismatches within a given synoptic interval discovered. Once such understanding is achieved, synoptic maps acquire predictive value. Not only can climatic zones be completed by interpolation, but their probable extent and associated sedimentary deposits of zones, not yet discovered, may also be predicted.
In studying the older geologic record we are commonly hampered by the lack of present-day or Pleistocene analogs. In the absence of a physicochemical understanding of ocean-atmosphere-land interaction, we need such analogies to provide guidance. Quantitative simulation of paleoclimates proceeds from using the present as an analog rather than from first principles of fluid flow. To understand the full functioning of the system, we need a greater array of climate conditions than the Pleistocene can yield. Nor does the Late Cenozoic yield enough analogs for more remote times. Although a synoptic analysis of the early Pliocene and late Miocene can be constructed from existing data, it is insufficient as an analog for earlier times. Oceanic circulation was thermally driven in Cenozoic times, but there is evidence (Chapter 7) that it was driven by salinity gradients in the Mesozoic Era.
For late Precambrian intervals, intercontinental correlation is imprecise and continental masses are difficult to define and to place relative to one another and to the equator. For older periods within the Precambrian, the time slices become ever thicker and the identification of sites of sediment deposition become less precise. Investigations in the Precambrian must therefore focus increasingly on the statistical significance of climatic indicators. As an analogy to such research in the fragmentary Precambrian record, consider what might be learned of the Earth’s climate today if it were visited by an unmanned automated spacecraft and lander. Random samples from the present Earth’s surface would be revealing. We probably would be able to reconstruct (1) the relative abundance of ocean, shelf, and land environments; (2) the fact that we live in an ice age but with warm tropics; and (3) the presence of considerable diversity in soil types. We might even be able to arrange these soil types in a series from those for which mechanical processes of soil formation predominate to those for which chemical processes prevail.
In the collection of such data from the Precambrian sedimentary record, the over-representation or underrepresentation of certain regions or certain rock types needs evaluation. Such distortions come both from differential preservation of rock types and from accidents of exposure and location. Improvement is also needed in understanding the systematic physical and chemical changes that can modify the sedimentary record after deposition and the effect of these postdepositional changes on the distortion of the climatic information. In addition, depositional conditions may not have analogs in the present because the chemical composition of the oceans and atmosphere was different from now. We suggest that studies in organic geochemistry, with its abundance of compounds and pathways, have been underutilized in providing understanding of the postdepositional changes of climate indicators. The diagenetic processes themselves may be climate controlled.
The composition of the early Precambrian atmosphere was significantly different
from the composition of the present atmosphere. From astrophysical theory and by comparison with the inferred evolution of other stars, it is assumed that the Sun was somewhat less luminous when the Earth was young (3000 or 4000 Ma). The concept that the Sun has continuously increased its heat output (by about 30 percent) has led to a search for a balancing factor to keep the Earth’s temperature hospitable. If the Earth’s atmosphere had its current composition throughout geologic time, calculations (Budyko, 1969; Schneider and Dickinson, 1974) predict that the planet would have been permanently glaciated since near its beginning and would not have thawed yet. If the early atmosphere were rich in carbon dioxide, an enhanced greenhouse effect would have kept the mean surface temperature of the Earth above freezing.
The composition of some Precambrian strata indicates that the early atmosphere was poor in oxygen (Cloud, 1965). This has been inferred from study of the banded-iron formations found in ancient rocks over the world (about 2200 Ma). The transport of the iron without the concurrent movement of other materials (e.g., Al2O3) to its depositional sites implies a carbon dioxide content of the Precambrian atmosphere approximately double that of the present atmosphere (Rubey, 1951; James, 1966). A similar conclusion for atmospheric composition has been drawn from the Precambrian gold-uranium-pyrite occurrences typified by the Witwatersrand (South Africa) and Jacobina (Brazil) deposits. These have been interpreted as placers. However, the detrital uranium-bearing mineral and the pyrite would have been oxidized and made soluble in the surface waters if the atmosphere were rich in oxygen as it is today (Hutchinson, 1981).
The Paleozoic and Mesozoic Eras
For each geologic epoch within the Paleozoic and Mesozoic Eras, several synoptic intervals hold promise for satisfactory worldwide climate correlation. Within such time slices, those data on the geochemistry of sedimentary rocks and fossils that bear on climate need to be systematically collected and synthesized. For each time slice the former position of continents needs to be reconstructed, incorporating paleomagnetic and other data. Such reconstructions should be augmented by geophysical, geologic, and pertinent compositional information from sedimentologic investigations. The sizes of the error margins also need evaluation. If this is done properly, each reconstruction can start where the previous one left off, thus adding further confidence. Estimates of the latitudinal gradients of temperature and humidity, for each time slice, will emerge from interpretation of the distribution of climate indicators with the continental reconstructions.
Such synoptic analyses are currently being attempted for several periods in the Paleozoic and Mesozoic. See, for example, time-slice analysis in Chapter 21 for Silurian-Devonian times and in Chapter 17 for the Jurassic. Similar analyses have been attempted by Gray and Boucot (1979), Habicht (1979), and Ziegler et al. (1979).
The Cenozoic Era
As we move toward the present, the requirements of reconstruction of place and time are met with comparative ease. The reconstruction and interpretation of climate, and especially of ocean-atmosphere interactions, can now move from the level of narrative toward one of numerical simulation. Such simulations have been attempted for the last glacial maximum in the Pleistocene with significant success (CLIMAP Project Members, 1976).
We need synoptic intervals for an Earth with no (or minor) northern ice caps. One time slice closest to the present and useful in this regard is a warm peak in the early
Pliocene. Further back in time, another instructive interval occurred within the early Miocene, when the Antarctic ice cap was small. Stratigraphic correlations are probably adequate within Miocene and Pliocene sequences, and data coverage is good because sediments of these ages are available over wide areas of the seafloor, as well as on continental shelves.
Thick continuous stratigraphic sections consisting of numerous thin beds may record climate variability if the bedding is rhythmic or repetitious and the thickness, or some other characteristic, is the response to a seasonal climate signal. Statistical analysis of the recurring signal may disclose variations in climate on a seasonal, annual, or even longer basis. Well-dated continuous sequences invite such analysis, and the investigation of several nearby sequences may enable the sorting out of local or regional responses from those of broader or even global significance.
Time-series analysis can also employ data from many localities if the rocks or sediments can be accurately dated and stratigraphically correlated. Ancient ocean temperatures may be deduced from oxygen isotopic analyses (Figure 3) and are the type of data appropriate to time-series study. Such analyses show the gross variations of a climatic parameter with time, and continuous stratigraphic sections show more detailed variations, which may reflect differences in climatic driving forces.
With long quasi-rhythmic sequences in hand, variability can be extracted both from complete time series and from random sampling within a given section. Where there is order along the time axis, so that a pattern or repetition is recognized, climate cycles may be present. For example, the extraction of such cycles from the Pleistocene record of deep-seas cores by frequency analysis has shown that climate has responded to orbital variations in the recent geologic past (Imbrie and Imbrie, 1980).
Time-series studies, including both variability analysis and frequency analysis, are possible when continuous global climate signals dominate or are amplified by regional ones. Because of the likelihood of continuous deposition, pelagic sediments from the deep sea, and even older sequences uplifted and exposed on land, are the most promising candidates for time-series study. Well-dated sections provide the most readily interpretable data, but long sections that are only approximately dated may reveal useful repetitions interpretable as having resulted from climate changes. Although the pelagic record is more complete from the mid-Cretaceous and younger seafloor (because of the destruction of ocean crust at subduction zones), ancient sections even as old as Precambrian warrant study. Times in which interesting climatic signals have been observed are the early Pliocene (preceding northern ice caps), the middle Eocene (weak latitudinal temperature gradient), and the late Cretaceous (equable temperature gradient and possibly high carbon dioxide content in the atmosphere).
Pleistocene deep-sea cores show that cyclicity is commonly associated with more than one climate-associated parameter. Phase differences between such parameters contain information, as yet poorly understood, about cause-and-effect chains in the ocean-atmosphere-biosphere-lithosphere system. For example, the coherences and offsets between oxygen isotope cycles, carbon isotope cycles, and carbonate cycles have implications for the sequence of events—such as the buildup and melting of ice, changes in temperature, changes in carbonate sedimentation, and fluctuations in atmospheric carbon dioxide.
Fluctuations of sea level accompany the waxing and waning of continental glaciers. However, nonclimatic factors, such as vertical crustal movements and the rates of seafloor spreading, also cause eustatic sea-level changes. These eustatic changes need to be separated from those caused by climate—in fact they can affect climate. Sea-level fluctuations change the albedo by altering the ratio between exposed land surface and the surface area of inland seas and lakes. Because water on the average is a better absorber of light than is dry land, when sea level is high, increased absorption of solar radiation results. In addition there is an increase in the surface area from which water can enter the atmosphere through evaporation. Hence, humidity rises and there is an increase in the greenhouse effect, which warms the near-surface layers of the atmosphere. Such complex feedback relations as these are still incompletely documented or understood. Are times of high sea level invariably warmer and more equable than those of low sea level? What are the causes of sea-level change? How has sea level changed through geologic time?
A period-by-period and region-by-region compilation of sea-level variation for the continental shelf areas of the world has been initiated by geologists of the petroleum industry employing the techniques of seismic stratigraphy (see Chapter 15). The key to success is exact stratigraphic correlation, using well-dated sections. These data in turn need to be dovetailed with climatic indicators.
EVENT AND EPISODE ANALYSIS
Notable events in Earth history, such as those associated with the termination of the Cretaceous Period, the fragmentation of Gondwana, and widespread volcanism, have left discrete geologic records that will certainly be useful in connection with climatic investigation. Many events may actually occur over longer periods and are more properly referred to as episodes. The analysis of such climatic transitions involves study of conditions prevailing before and after, as well as during, the event. Even though rapid climatic changes are deemed to be caused by single forcing factors, the signals may be overwhelmed by background factors and from competing causes. Careful analysis is required to sort out the various contributing causes of the
event. Sea-level change stimulates feedback through a chain of processes that need elucidation. Basic questions regarding many important geologic events are whether they were externally (e.g., extraterrestrially) caused or forced, and how important internal feedback is in the system.
Event analyses of climatic records are successful when gaps in the record are absent or subordinate. A sedimentary sequence whose continuity is established can be characterized statistically by the distribution of events within it. By studying each event in detail, using various climate indicators, it may be possible to distinguish classes of events (Berger et al., 1981).
Times of rapid climate change need to be studied in detail on a worldwide basis if we are to understand the factors that can cause such climatic transitions. At the time of the Cretaceous termination (about 63 Ma), for example, differences in composition of deep-sea sediments (a short-lived increase in iridium concentration) accompanied a marked biological event; this supports the hypothesis of a large meteoritic or cometary impact. Other key times in geologic history that should be elucidated by paleoclimatic investigations are (1) the Albian-Cenomanian (about 100 Ma) transition, (2) the Eocene-Oligocene boundary (38 Ma), (3) the mid-Miocene (15 Ma), (4) the end-of-Miocene (6 Ma), and (5) the mid-Pliocene (3 Ma). A multifaceted approach involving biostratigraphy, magnetostratigraphy, stable isotopes, trace element analysis, and other methods is needed for each of these events and several others.
Some episodes in the distant geologic past may have taken place over millions of years. For example, what happened climatologically at the onset of the Phanerozoic? What caused the appearance of calcareous shells at approximately this time? Was there a change in seawater chemistry or an increase in predation pressure and competition once the first hard shells evolved? In the Precambrian, events may have taken tens of millions of years to rise, culminate, and perhaps to fade—possibly comparable to the long-term cooling in the Cenozoic. Conditions favorable for the formation of some banded-iron formations have not existed for nearly 2 billion years. The origin of intercalations of chert and iron ore in these banded-iron formations is still a subject of research. Perhaps changing climatic and geochemical conditions hold the clue: What were the factors that made deposition of banded-iron formations possible? What environmental condition changed when their deposition ceased?
NEW OPPORTUNITIES IN PALEOCLIMATOLOGY
Climate is determined by two major classes of factors. The first class constitutes the “static setting.” This is the arrangement of continents and ocean basins, which can be considered fixed on a short time scale. The second class of factors involves what may be called the “dynamic setting.” This includes oceanic and atmospheric circulation, heat transfer, albedo distributions, biosphere patterns, and atmospheric concentration of infrared absorbing gases. A goal of paleoclimatology is to interrelate the static and dynamic settings and determine how they operate together.
New opportunities arise from our present understanding of plate tectonics and continental drift. For times back to about 200 Ma, we know the geography of the static setting and can concentrate on the dynamics of the fluids within it. We now have a set of concepts and methods developed by climatologists working on Pleistocene and modern climates. These concepts and methods can be tested for applicability to pre-Pleistocene climates and promise to reveal much about why long-term climate changes have occurred.
Two examples illustrate the opportunities. The first is the map of the surface of the Earth at the time of the last maximum of glaciation, 18,000 years ago, based on integration of paleoclimatic information from land and from the seafloor (CLIMAP
Project Members, 1976; Denton and Hughes, 1981). Paleotemperature and albedo distributions are presented on a scale fine enough to be useful for numerical modeling of the paleoclimate (Gates, 1976; Manabe and Hahn, 1977).
The second example concerns time-series analysis in which a long climatic record, such as a long deep-sea core (Figure 4), is examined for repetitions and cycles of a climatic signal (Hays et al., 1976; Pisias, 1976). These studies are useful in single stratigraphic sections and offer an escape from many of the frustrations associated with time-slice mapping, where correlations over large distances may be uncertain. Such studies yield insights about the dynamics and variability of climate systems at any one site that cannot be obtained from spatial reconstructions alone. Analysis of Pleistocene climatic records inferred from deep-sea sediments has established relations between orbital parameters and glaciation by identifying astronomical cycles of 19,000 yr, 23,000 yr, 41,000 yr, and perhaps 100,000 yr within the geologic record. A search for these cycles in older records and for other cycles is now possible; they may be present in Mesozoic, Paleozoic, and perhaps even Precambrian sequences. The task is to find long, continuous stratal sequences recording a signal (e.g., thickness variations in limestone-shale couplets) that permits frequency analysis. Even where the time scale is not known accurately, the ratios of different cycles or repetitions may eventually lead to the establishment of such time scales, and the identification of cycles will aid in long-distance correlation.
Existence of both longer and shorter cycles than those orbital cycles so far identified is likely. The analysis of long-period variations in annual layers, if they can be recognized, through geologic time may yield information on the variations of solar radiation incident on the Earth. Unfortunately, there is no simple criterion by which to prove that layering is truly annual. Long-cycle lengths of approximately 250,000 and 450,000 years have been proposed, as well as frequencies of 4–5 m.y., 30 m.y., and 200 m.y. (see Fischer, Chapter 9). Usually such suggestions rest on records that are only two to five times longer than the proposed “cycle,” so that they cannot yet be established quantitatively.
THE SYSTEMS APPROACH
Because climate is the result of the air and ocean circulation and their interactions within a geologic setting, a dynamic fluid system is defined, a system that may lend
itself to systems analysis. Sediments deposited by these fluids record a composite of global signals and regional or local overprints. It would be useful to attempt to separate the global message from the regional modulations and to investigate the sensitivity of the system. Global effects may be rapid with quick response times or with phase shifts between different components. Regional influences may amplify or dampen or even reverse global signals. Climate extremes and rapid transitions from one state to another are probably associated with times when factors influencing global climate are maximized.
Studies that focus on times of instability and rapid climate transition within an overall stable record may reveal why evolution is apparently punctuated by mass extinctions and relatively rapid speciation, such as those marking the end of the Paleozoic and Mesozoic eras. To what extent were climatic extremes due to variations within the system or to factors external to the system, such as those of extraterrestrial origin? Important questions about extraterrestrial factors are the following: How stable has the solar constant been? Has the Sun’s output varied in an oscillatory manner? What influence might large impacts on the Earth (similar to those that produced large craters on the Moon) have had on climate evolution? Have there been times when havoc has been raised in the Earth’s surface environment by huge meteorites falling from the sky, as has been suggested for the end of the Cretaceous by Alvarez et al. (1980)?
Factors that are not related to the fluid system but are related to the dynamics and mobile Earth include tectonic forces and volcanic activity. Continents have moved about and oceans have opened and widened and closed through geologic time. Mountain ranges rose in many orientations that influenced the air circulation and then were worn away. These marked changes in size and shape and positioning of landmasses, oceans, and mountains have dramatically influenced climates. Ocean gateways between landmasses have opened and closed and have altered the flow of ocean currents and modified the mechanisms of heat transport from low to high latitudes. Ancient ice ages during the Phanerozoic (see Figure 2) may be largely the consequence of times when continents were sited in high latitudes so that air-ocean flow brought heavy and long-lasting snowfall, with the result that ice caps grew upon them. The location of landmasses, the level of the sea, and the presence of ocean gateways in crucial positions are several of the boundary conditions that are responsible for the sensitivity of the fluid system to perturbation.
Geochemical aspects are also intimately tied to climate, mainly through the carbon dioxide content of the atmosphere; increases in carbon dioxide raise the surface temperature. On a long time scale, the average atmospheric carbon dioxide content may be fixed by weathering reactions and carbon storage in soil, rocks, and seawater. However, on a shorter time scale, the atmospheric reservoir is dominated by the much larger adjoining reservoirs of the ocean, the biosphere-soil complex, and the reactive portion of marine organic and calcareous sediments. As the exchange patterns between these reservoirs change, so must the carbon dioxide content of the atmosphere. Rapid increase or decrease of carbon dioxide content that results from temporary buildup or release of carbonate or organic matter may cause, or be associated with, “climate steps” from one climatic state to another (Berger et al., 1981; Revelle, 1981). The existence of steps suggests the existence of inherent instability within the system; this may be tested by a comparison of climatic variability at times long before the event, approaching the event, and after the event.
FINDINGS AND RECOMMENDATIONS
Interpretation of the history of climates on Earth back in geologic time for more than a billion years indicates that the same principles of oceanic and atmospheric circula-
tion have always operated. Over this long interval, however, the composition of the atmosphere and oceans has changed in ways not yet fully determined. Climatic extremes that we are experiencing at present, and that have occurred for the past million years or so, are probably similar to climatic extremes back into the very remote past. Only at rare intervals do the many variables influencing climate so combine that they perturb the system beyond a quasi-steady state, bringing about a transition to a new temperature regime, but one actually not very different from the previous one.
The record of climate and its changes during the last 50 million years reveals a gradual cooling interrupted by a few pertubations. Fairly rapid transitions or steps resulting in a cooler mean surface temperature of a few degrees Celsius have been followed by partial recoveries. Superimposed rhythms on these fluctuations, detectable in the Pleistocene record but so far not discovered in older rocks, are brought about by systematic changes in the Earth’s axis and orbit.
Understanding of climate and climatic change, with the eventual goal of arriving at a satisfactory theory of climate, requires an improved data base and more intense efforts in analysis and interpretation. Many disciplines are involved, Toward these ends, we recommend the following:
Geologists, during their investigation of rocks, should continue to strive to glean information bearing on past climates. Economic geologists, for example, recognize that certain types of mineral deposits were formed under special climatic conditions. Through an understanding of weathering and other climate-controlled processes they may be able to extract climatic information. Even the most ancient sedimentary rocks may reveal some information on climates.
Marine geologists and geochemists have a new tool in hand—the hydraulic piston corer—and should obtain many more long sediment cores from the seafloor. Valuable paleoclimatic data, highly resolved with respect to time, can come from well-placed cores, especially from localities where global climatic signals are not masked by local sediment influxes. We urge the continuation of programs using hydraulic piston cores in any ocean-drilling effort.
Paleogeographers, working closely with paleontologists, tectonicists, and stratigraphers, should prepare paleogeographic maps of the Earth for all practical time intervals of Earth history. These intervals can be short for time slices in the Tertiary and by necessity must be longer for intervals successively more remote. Such maps will serve as a basis for assembling climatic data and as guides for computer modeling.
Meteorologists and climatologists, working at many different scales, should attempt to arrive at explanations for ancient climates. Although satisfactory computer models do not yet exist to explain all aspects of modern climates, such studies for ancient synoptic intervals can nevertheless serve as guides to explain climatic events. Therefore, models of several types should be applied to the paleogeographic reconstructions of selected times of climatic significance in the geologic past.
Investigators of sea-level changes should be encouraged to complete as accurately as possible a worldwide eustatic sea-level curve for the geologic past. Such a curve will aid greatly in correlation from place to place, when the record of advances and retreats of the sea is identified. If other factors remain unchanged, increased flooding should correlate with increased global temperatures because of albedo reduction and increased evaporation. These inferences can be checked independently against the geologic record.
Long stratal sequences displaying variations should be analyzed using the time-series approach. These sequences may come from long deep-sea cores or from sections exposed on land. Their study may reveal repetitions and cycles in bedding
features, such as thickness or composition, that are related to climate. Among these are orbital and axial phases that influence the amount of radiative energy received from the Sun. With some luck, even ancient Precambrian sedimentary rocks may disclose such information.
Alvarez, L.W., W.Alvarez, F.Asaro, and H.V.Michel (1980). Extraterrestrial cause for the Cretaceous-Tertiary extinctions: Experiment and theory, Science 208, 1095–1108.
Berger, W.H., E.Vincent, and H.R.Thierstein (1981), The deep-sea record: Major steps in Cenozoic ocean evolution, in Symposium and Results of Deep-Sea Drilling, R.G.Douglas, J.Warme, and E.L. Winterer, eds., Soc. Econ. Paleontol. Mineral. Spec. Publ. 32.
Berggren, W.A. (1972). Late Pliocene-Pleistocene glaciation, in Initial Reports of the Deep Sea Drilling Project 12, U.S. Government Printing Office, Washington, D.C., pp. 953–963.
Beuf, S., B.Biju-Duval, O.DeCharpal, P.Rognon, O.Gariel, and A.Bennacef (1971). Les Grès du Paléozoique Inférieur au Sahara—Sédimentation et Discontinuities, Evolution Structurale d’un Craton, Inst. Fr. Pétroles-Sci. Tech. Pétroles 18, 464 pp.
Budyko, M.I. (1969). The effect of solar radiation variations on the climate of the Earth, Tellus 21, 611–619.
CLIMAP Project Members (1976). The surface of the ice-age Earth, Science 191, 1131–1137.
Cloud, P.E., Jr. (1965). Significance of the Gunflint (Precambrian) flora, Science 148, 27–35.
Denton, G.H., and T.J.Hughes (1981). The Last Great ice Sheets, Wiley-Interscience, New York, 489 pp.
Douglas, R.G., and F.Woodruff (1981). Deep sea benthic foraminifera, in The Sea, Vol. 7, C.Emiliani, ed., Wiley, New York.
Frakes, L.A. (1979). Climates Throughout Geologic Time, Elsevier, Amsterdam, 310 pp.
Gates, W.L. (1976). The numerical simulation of ice-age climate with a global general circulation model, J. Atmos. Sci. 33, 1844–1873.
Gray, J., and A.J.Boucot, eds. (1979). Historical Biogeography, Plate Tectonics, and the Changing Environment, Oregon State U. Press, Corvallis, Ore., 500 pp.
Habicht, J.K.A. (1979). Paleoclimate, Paleomagnetism, and Continental Drift, AAPG Studies in Geology No. 9, American Association of Petroleum Geologists, Tulsa, Okla., 31 pp.+11 maps.
Hambrey, M.J., and W.B.Harland (1981). Earth’s Pre-Pleistocene Glacial Record, Cambridge U. Press, New York, 1004 pp.
Hansen, J., D.Johnson, A.Lacis, S.Lebedeff, P.Lee, D.Rind, and G.Russell (1981). Climate impact of increasing atmospheric carbon dioxide, Science 213, 957.
Hays, J.D., J.Imbrie, and N.J.Shackelton (1976). Variations in the Earth’s orbit, pacemaker of the ice ages, Science 194, 1121–1132.
Hutchinson, R.W. (1981). Mineral deposits as guides to supracrustal evolution, in Evolution of the Earth, R.J.O’Connell and W.S.Fyfe, eds., Geodynamic Series, Geological Society of America and American Geophysical Union, Washington, D.C., pp. 120–140.
Imbrie, J., and J.Z.Imbrie (1980). Modeling the climate response to orbital variations, Science 207, 943–953.
Imbrie, J., and K.P.Imbrie (1979). Ice Ages—Solving the Mystery, Enslow Publishers, Short Hills, N.J., 224 pp.
James, H.L. (1966). Chemistry of the iron-rich sedimentary rocks, U.S. Geol. Surv. Prof. Pap. 440.
Manabe, S., and D.G.Hahn (1977). Simulation of the tropical climate of an ice age, J, Geophys. Res. 82, 3389–3911.
McElhinny, M.W., and D.A.Valencio, eds. (1981). Paleoreconstruction of the Continents, Geodynamic Series, Geological Society of America and American Geophysical Union, Washington, D.C., 200 pp.
NRC Climate Board (1982). CO2and Climate: A Second Assessment, National Academy Press, Washington, D.C., 72 pp.
NRC Climate Research Board (1979). Carbon Dioxide and Climate: A Scientific Assessment, National Academy of Sciences, Washington, D.C., 22 pp.
NRC Geophysics Study Committee (1977). Energy and Climate, National Academy of Sciences, Washington, D.C., 158 pp.
NRC U.S. Committee for the Global Atmospheric Research Program (1975). Understanding Climatic Change, National Academy of Sciences, Washington, D.C., 239 pp.
Pisias, N.C. (1976). Late Quaternary variations in sedimentation rate in the Panama Basin and the identification of orbital frequencies in carbonate and opal deposition rates, Geol. Soc. Am. Mem. 145, 375–391.
Revelle, R.R. (1981). Introduction to The Sea, Vol. 7, C. Emiliani, ed., Wiley, New York.
Rubey, W.W. (1951). The geologic history of sea water, Geol. Soc. Am. Bull. 62, 1111–1147.
Schneider, S.H., and R.E.Dickinson (1974). Climate modeling, Rev. Geophys. Space Phys. 12, 447–493.
Seyfert, C.K., and L.A.Sirkin (1979). Earth History and Plate Tectonics: An Introduction to Historical Geology, Harper & Row, New York.
Shackleton, N.H., and N.D.Opdyke (1977). Oxygen isotope and paleomagnetic evidence for early northern hemisphere glaciation, Nature 270, 216–219.
U.S. Congress (1978). National Climate Program Act, Public Law 95–367, 95th Congress, September 17, 1978.
Ziegler, A.M., C.R.Scotese, W.S.McKerrow, M.E.Johnson, and R.K.Bambach (1979). Paleozoic paleogeography, Ann. Rev. Earth Planet. Sci. 7, 473–502.