Like the lapis lazuli gem it resembles, the blue, cloud-enveloped planet the we recognize immediately from satellite pictures seems remarkably stable. Continents and oceans, encircled by an oxygen-rich atmosphere, support familiar life-forms. Yet this constancy is an illusion produced by the human experience of time. Earth and its atmosphere are continuously altered. Plate tectonics shift the continents, raise mountains and move the ocean floor while processes not fully understood alter the climate.

Such constant change has characterized Earth since its beginning some 4.5 billion years ago. From the outset, heat and gravity shaped the evolution of the planet. These forces were gradually joined by the global effects of the emergence of life. Exploring this past offers us the only possibility of understanding the origin of life and, perhaps, its future.

Scientists used to believe the rocky planets, including Earth, Mercury, Venus and Mars, were created by the rapid gravitational collapse of a dust cloud, a deation giving rise to a dense orb. In the 1960s the Apollo space program changed this view. Studies of moon craters revealed that these gouges were caused by the impact of objects that were in great abundance about 4.5 billion years ago. Thereafter, the number of impacts appeared to have quickly decreased. This observation rejuvenated the theory of accretion postulated by Otto Schmidt. The Russian geophysicist had suggested in 1944 that planets grew in size gradually, step by step.

According to Schmidt, cosmic dust lumped together to form particulates, particulates became gravel, gravel became small balls, then big balls, then tiny planets, or planetesimals, and, nally, dust became the size of the moon. As the planetesimals became larger, their numbers decreased. Consequently, the number of collisions between planetesimals, or meteorites, decreased. Fewer items available for accretion meant that it took a long time to build up a large planet. A calculation made by George W. Wetherill of the Carnegie Institution of Washington suggests that about 100 million years could pass between the formation of an object measuring 10 kilometers in diameter and an object the size of Earth.

The process of accretion had significant thermal consequences for Earth, consequences that forcefully directed its evolution. Large bodies slamming into the planet produced immense heat in its interior, melting the cosmic dust found there. The resulting furnace--situated some 200 to 400 kilometers underground and called a magma ocean--was active for millions of years, giving rise to volcanic eruptions. When Earth was young, heat at the surface caused by volcanism and lava ows from the interior was intensified by the constant bombardment of huge objects, some of them perhaps the size of the moon or even Mars. No life was possible during this period.

Beyond clarifying that Earth had formed through accretion, the Apollo program compelled scientists to try to reconstruct the subsequent temporal and physical development of the early Earth. This undertaking had been considered impossible by founders of geology, including Charles Lyell, to whom the following phrase is attributed: No vestige of a beginning, no prospect for an end. This statement conveys the idea that the young Earth could not be re-created, because its remnants were destroyed by its very activity. But the development of isotope geology in the 1960s had rendered this view obsolete. Their imaginations red by Apollo and the moon ndings, geochemists began to apply this technique to understand the evolution of Earth.

Dating rocks using so-called radioactive clocks allows geologists to work on old terrains that do not contain fossils. The hands of a radioactive clock are isotopes--atoms of the same element that have different atomic weights--and geologic time is measured by the rate of decay of one isotope into another [see "The Earliest History of the Earth," by Derek York; Scientific American, January 1993]. Among the many clocks, those based on the decay of uranium 238 into lead 206 and of uranium 235 into lead 207 are special. Geochronologists can determine the age of samples by analyzing only the daughter product--in this case, lead--of the radioactive parent, uranium.

Panning for zircons
ISOTOPE GEOLOGY has permitted geologists to determine that the accretion of Earth culminated in the differentiation of the planet: the creation of the core--the source of Earth's magnetic field--and the beginning of the atmosphere. In 1953 the classic work of Claire C. Patterson of the California Institute of Technology used the uranium-lead clock to establish an age of 4.55 billion years for Earth and many of the meteorites that formed it. In the early 1990s, however, work by one of us (Allègre) on lead isotopes led to a somewhat new interpretation.

As Patterson argued, some meteorites were indeed formed about 4.56 billion years ago, and their debris constituted Earth. But Earth continued to grow through the bombardment of planetesimals until some 120 million to 150 million years later. At that time--4.44 billion to 4.41 billion years ago--Earth began to retain its atmosphere and create its core. This possibility had already been suggested by Bruce R. Doe and Robert E. Zartman of the U.S. Geological Survey in Denver two decades ago and is in agreement with Wetherills estimates.

The emergence of the continents came somewhat later. According to the theory of plate tectonics, these landmasses are the only part of Earth's crust that is not recycled and, consequently, destroyed during the geothermal cycle driven by the convection in the mantle. Continents thus provide a form of memory because the record of early life can be read in their rocks. Geologic activity, however, including plate tectonics, erosion and metamorphism, has destroyed almost all the ancient rocks. Very few fragments have survived this geologic machine.

Nevertheless, in recent decades, several important nds have been made, again using isotope geochemistry. One group, led by Stephen Moorbath of the University of Oxford, discovered terrain in West Greenland that is between 3.7 billion and 3.8 billion years old. In addition, Samuel A. Bowring of the Massachusetts Institute of Technology explored a small area in North America--the Acasta gneiss--that is thought to be 3.96 billion years old.

Ultimately, a quest for the mineral zircon led other researchers to even more ancient terrain. Typically found in continental rocks, zircon is not dissolved during the process of erosion but is deposited in particle form in sediment. A few pieces of zircon can therefore survive for billions of years and can serve as a witness to Earths more ancient crust. The search for old zircons started in Paris with the work of Annie Vitrac and Jol R. Lancelot, later at the University of Marseille and now at the University of Nmes, respectively, as well as with the efforts of Moorbath and Allgre. It was a group at the Australian National University in Canberra, directed by William Compston, that was nally successful. The team discovered zircons in western Australia that were between 4.1 billion and 4.3 billion years old.

Zircons have been crucial not only for understanding the age of the continents but for determining when life rst appeared. The earliest fossils of undisputed age were found in Australia and South Africa. These relics of blue-green algae are about 3.5 billion years old. Manfred Schidlowski of the Max Planck Institute for Chemistry in Mainz studied the Isua formation in West Greenland and argued that organic matter existed as long ago as 3.8 billion years. Because most of the record of early life has been destroyed by geologic activity, we cannot say exactly when it rst appeared--perhaps it arose very quickly, maybe even 4.2 billion years ago.

Stories from gases
ONE OF THE MOST important aspects of the planet's evolution is the formation of the atmosphere, because it is this assemblage of gases that allowed life to crawl out of the oceans and to be sustained. Researchers have hypothesized since the 1950s that the terrestrial atmosphere was created by gases emerging from the interior of the planet. When a volcano spews gases, it is an example of the continuous outgassing, as it is called, of Earth. But scientists have questioned whether this process occurred suddenly--about 4.4 billion years ago when the core differentiated--or whether it took place gradually over time.

To answer this question, Allègre and his colleagues studied the isotopes of rare gases. These gases--including helium, argon and xenon--have the peculiarity of being chemically inert, that is, they do not react in nature with other elements. Two of them are particularly important for atmospheric studies: argon and xenon. Argon has three isotopes, of which argon 40 is created by the decay of potassium 40. Xenon has nine, of which xenon 129 has two different origins. Xenon 129 arose as the result of nucleosynthesis before Earth and solar system were formed. It was also created from the decay of radioactive iodine 129, which does not exist on Earth anymore. This form of iodine was present very early on but has died out since, and xenon 129 has grown at its expense.

Like most couples, both argon 40 and potassium 40 and xenon 129 and iodine 129 have stories to tell. They are excellent chronometers. Although the atmosphere was formed by the outgassing of the mantle, it does not contain any potassium 40 or iodine 129. All argon 40 and xenon 129, formed in Earth and released, are found in the atmosphere today. Xenon was expelled from the mantle and retained in the atmosphere; therefore, the atmosphere-mantle ratio of this element allows us to evaluate the age of differentiation. Argon and xenon trapped in the mantle evolved by the radioactive decay of potassium 40 and iodine 129. Thus, if the total outgassing of the mantle occurred at the beginning of Earths formation, the atmosphere would not contain any argon 40 but would contain xenon 129.

The major challenge facing an investigator who wants to measure such ratios of decay is to obtain high concentrations of rare gases in mantle rocks because they are extremely limited. Fortunately, a natural phenomenon occurs at mid-ocean ridges during which volcanic lava transfers some silicates from the mantle to the surface. The small amounts of gases trapped in mantle minerals rise with the melt to the surface and are concentrated in small vesicles in the outer glassy margin of lava ows. This process serves to concentrate the amounts of mantle gases by a factor of 104 or 105. Collecting these rocks by dredging the seaoor and then crushing them under vacuum in a sensitive mass spectrometer allows geochemists to determine the ratios of the isotopes in the mantle. The results are quite surprising. Calculations of the ratios indicate that between 80 and 85 percent of the atmosphere was outgassed during Earths rst one million years; the rest was released slowly but constantly during the next 4.4 billion years.

The composition of this primitive atmosphere was most certainly dominated by carbon dioxide, with nitrogen as the second most abundant gas. Trace amounts of methane, ammonia, sulfur dioxide and hydrochloric acid were also present, but there was no oxygen. Except for the presence of abundant water, the atmosphere was similar to that of Venus or Mars. The details of the evolution of the original atmosphere are debated, particularly because we do not know how strong the sun was at that time. Some facts, however, are not disputed. It is evident that carbon dioxide played a crucial role. In addition, many scientists believe the evolving atmosphere contained sufficient quantities of gases such as ammonia and methane to give rise to organic matter.

Still, the problem of the sun remains unresolved. One hypothesis holds that during the Archean eon, which lasted from about 4.5 billion to 2.5 billion years ago, the suns power was only 75 percent of what it is today. This possibility raises a dilemma: How could life have survived in the relatively cold climate that should accompany a weaker sun? A solution to the faint early sun paradox, as it is called, was offered by Carl Sagan and George Mullen of Cornell University in 1970. The two scientists suggested that methane and ammonia, which are very effective at trapping infrared radiation, were quite abundant. These gases could have created a super-greenhouse effect. The idea was criticized on the basis that such gases were highly reactive and have short lifetimes in the atmosphere.

What controlled co?
IN THE LATE 1970s Veerabhadran Ramanathan, now at the Scripps Institution of Oceanography, and Robert D. Cess and Tobias Owen of Stony Brook University proposed another solution. They postulated that there was no need for methane in the early atmosphere because carbon dioxide was abundant enough to bring about the super-greenhouse effect. Again this argument raised a different question: How much carbon dioxide was there in the early atmosphere? Terrestrial carbon dioxide is now buried in carbonate rocks, such as limestone, although it is not clear when it became trapped there. Today calcium carbonate is created primarily during biological activity; in the Archean eon, carbon may have been primarily removed during inorganic reactions.

The rapid outgassing of the planet liberated voluminous quantities of water from the mantle, creating the oceans and the hydrologic cycle. The acids that were probably present in the atmosphere eroded rocks, forming carbonate-rich rocks. The relative importance of such a mechanism is, however, debated. Heinrich D. Holland of Harvard University believes the amount of carbon dioxide in the atmosphere rapidly decreased during the Archean and stayed at a low level.

Understanding the carbon dioxide content of the early atmosphere is pivotal to understanding climatic control. Two conicting camps have put forth ideas on how this process works. The rst group holds that global temperatures and carbon dioxide were controlled by inorganic geochemical feedbacks; the second asserts that they were controlled by biological removal.

James C. G. Walker, James F. Kasting and Paul B. Hays, then at the University of Michigan at Ann Arbor, proposed the inorganic model in 1981. They postulated that levels of the gas were high at the outset of the Archean and did not fall precipitously. The trio suggested that as the climate warmed, more water evaporated, and the hydrologic cycle became more vigorous, increasing precipitation and runoff. The carbon dioxide in the atmosphere mixed with rainwater to create carbonic acid runoff, exposing minerals at the surface to weathering. Silicate minerals combined with carbon that had been in the atmosphere, sequestering it in sedimentary rocks. Less carbon dioxide in the atmosphere meant, in turn, less of a greenhouse effect. The inorganic negative feedback process offset the increase in solar energy.

This solution contrasts with a second paradigm: biological removal. One theory advanced by James E. Lovelock, an originator of the Gaia hypothesis, assumed that photosynthesizing microorganisms, such as phytoplankton, would be very productive in a high carbon dioxide environment. These creatures slowly removed carbon dioxide from the air and oceans, converting it into calcium carbonate sediments. Critics retorted that phytoplankton had not even evolved for most of the time that Earth has had life. (The Gaia hypothesis holds that life on Earth has the capacity to regulate temperature and the composition of Earth's surface and to keep it comfortable for living organisms.)

In the early 1990s Tyler Volk of New York University and David W. Schwartzman of Howard University proposed another Gaian solution. They noted that bacteria increase carbon dioxide content in soils by breaking down organic matter and by generating humic acids. Both activities accelerate weathering, removing carbon dioxide from the atmosphere. On this point, however, the controversy becomes acute. Some geochemists, including Kasting, now at Pennsylvania State University, and Holland, postulate that while life may account for some carbon dioxide removal after the Archean, inorganic geochemical processes can explain most of the sequestering. These researchers view life as a rather weak climatic stabilizing mechanism for the bulk of geologic time.

Oxygen from algae
THE ISSUE OF CARBON remains critical to how life inuenced the atmosphere. Carbon burial is a key to the vital process of building up atmospheric oxygen concentrations--a prerequisite for the development of certain life-forms. In addition, global warming is taking place now as a result of humans releasing this carbon. For one billion or two billion years, algae in the oceans produced oxygen. But because this gas is highly reactive and because there were many reduced minerals in the ancient oceans--iron, for example, is easily oxidized--much of the oxygen produced by living creatures simply got used up before it could reach the atmosphere, where it would have encountered gases that would react with it.

Even if evolutionary processes had given rise to more complicated life-forms during this anaerobic era, they would have had no oxygen. Furthermore, un ltered ultraviolet sunlight would have likely killed them if they left the ocean. Researchers such as Walker and Preston Cloud, then at the University of California at Santa Barbara, have suggested that only about two billion years ago, after most of the reduced minerals in the sea were oxidized, did atmospheric oxygen accumulate. Between one billion and two billion years ago oxygen reached current levels, creating a niche for evolving life.

By examining the stability of certain minerals, such as iron oxide or uranium oxide, Holland has shown that the oxygen content of the Archean atmosphere was low before two billion years ago. It is largely agreed that the present-day oxygen content of 20 percent is the result of photosynthetic activity. Still, the question is whether the oxygen content in the atmosphere increased gradually over time or suddenly. Recent studies indicate that the increase of oxygen started abruptly between 2.1 billion and 2.03 billion years ago and that the present situation was reached 1.5 billion years ago.

The presence of oxygen in the atmosphere had another major bene t for an organism trying to live at or above the surface: it ltered ultraviolet radiation. Ultraviolet radiation breaks down many molecules--from DNA and oxygen to the chlorouorocarbons that are implicated in stratospheric ozone depletion. Such energy splits oxygen into the highly unstable atomic form O, which can combine back into O2 and into the very special molecule O3, or ozone. Ozone, in turn, absorbs ultraviolet radiation. It was not until oxygen was abundant enough in the atmosphere to allow the formation of ozone that life even had a chance to get a root-hold or a foothold on land. It is not a coincidence that the rapid evolution of life from prokaryotes (single-celled organisms with no nucleus) to eukaryotes (single-celled organisms with a nucleus) to metazoa (multicelled organisms) took place in the billion-year-long era of oxygen and ozone.

Although the atmosphere was reaching a fairly stable level of oxygen during this period, the climate was hardly uniform. There were long stages of relative warmth or coolness during the transition to modern geologic time. The composition of fossil plankton shells that lived near the ocean oor provides a measure of bottom water temperatures. The record suggests that over the past 100 million years bottom waters cooled by nearly 15 degrees Celsius. Sea levels dropped by hundreds of meters, and continents drifted apart. Inland seas mostly disappeared, and the climate cooled an average of 10 to 15 degrees C. Roughly 20 million years ago permanent ice appears to have built up on Antarctica.

About two million to three million years ago the paleoclimatic record starts to show signi cant expansions and contractions of warm and cold periods in 40,000-year or so cycles. This periodicity is interesting because it corresponds to the time it takes Earth to complete an oscillation of the tilt of its axis of rotation. It has long been speculated, and recently calculated, that known changes in orbital geometry could alter the amount of sunlight coming in between winter and summer by about 10 percent or so and could be responsible for initiating or ending ice ages.

The warm hand of man
MOST INTERESTING and perplexing is the discovery that between 600,000 and 800,000 years ago the dominant cycle switched from 40,000-year periods to 100,000-year intervals with very large uctuations. The last major phase of glaciation ended about 10,000 years ago. At its height 20,000 years ago, ice sheets about two kilometers thick covered much of northern Europe and North America. Glaciers expanded in high plateaus and mountains throughout the world. Enough ice was locked up on land to cause sea levels to drop more than 100 meters below where they are today. Massive ice sheets scoured the land and revamped the ecological face of Earth, which was ve degrees C cooler on average than it is currently.

The precise causes of the longer intervals between warm and cold periods are not yet sorted out. Volcanic eruptions may have played a signi cant role, as shown by the effect of El Chichón in Mexico and Mount Pinatubo in the Philippines. Tectonic events, such as the development of the Himalayas, may have inuenced world climate. Even the impact of comets can inuence short-term climatic trends with catastrophic consequences for life [see "What Caused the Mass Extinction? An Extraterrestrial Impact," by Walter Alvarez and Frank Asaro; and "What Caused the Mass Extinction? A Volcanic Eruption," by Vincent E. Courtillot; Scientific American, October 1990]. It is remarkable that despite violent, episodic perturbations, the climate has been buffered enough to sustain life for 3.5 billion years.

One of the most pivotal climatic discoveries of the past 30 years has come from ice cores in Greenland and Antarctica. When snow falls on these frozen continents, the air between the snow grains is trapped as bubbles. The snow is gradually compressed into ice, along with its captured gases. Some of these records can go back more than 500,000 years; scientists can analyze the chemical content of ice and bubbles from sections of ice that lie as deep as 3,600 meters (2.2 miles) below the surface.

The ice-core borers have determined that the air breathed by ancient Egyptians and Anasazi Indians was very similar to that which we inhale today--except for a host of air pollutants introduced over the past 100 or 200 years. Principal among these added gases, or pollutants, are extra carbon dioxide and methane. Since about 1860--the expansion of the Industrial Revolution--carbon dioxide levels in the atmosphere have increased more than 30 percent as a result of industrialization and deforestation; methane levels have more than doubled because of agriculture, land use and energy production. The ability of increased amounts of these gases to trap heat is what drives concerns about climate change in the 21st century [see "The Changing Climate," by Stephen H. Schneider; Scientific American, September 1989].

The ice cores have shown that sustained natural rates of worldwide temperature change are typically about one degree C per millennium. These shifts are still signi cant enough to have radically altered where species live and to have potentially contributed to the extinction of such charismatic megafauna as mammoths and saber-toothed tigers. But a most extraordinary story from the ice cores is not the relative stability of the climate during the past 10,000 years. It appears that during the height of the last ice age 20,000 years ago there was 50 percent less carbon dioxide and less than half as much methane in the air than there has been during our epoch, the Holocene. This nding suggests a positive feedback between carbon dioxide, methane and climatic change.

The reasoning that supports the idea of this destabilizing feedback system goes as follows. When the world was colder, there was less concentration of greenhouse gases, and so less heat was trapped. As Earth warmed up, carbon dioxide and methane levels increased, accelerating the warming. If life had a hand in this story, it would have been to drive, rather than to oppose, climatic change. It appears increasingly likely that when humans became part of this cycle, they, too, helped to accelerate warming. Such warming has been especially pronounced since the mid-1800s because of greenhouse gas emissions from industrialization, land-use change and other phenomena. Once again, though, uncertainties remain.

Nevertheless, most scientists would agree that life could well be the principal factor in the positive feedback between climatic change and greenhouse gases. There was a rapid rise in average global surface temperature at the end of the 20th century [see illustration on opposite page]. Indeed, the period from the 1980s onward has been the warmest of the past 2,000 years. Nineteen of the 20 warmest years on record have occurred since 1980, and the 12 warmest have all occurred since 1990. The all-time record high year was 1998, and 2002 and 2003 were in second and third places, respectively. There is good reason to believe that the decade of the 1990s would have been even hotter had not Mount Pinatubo erupted: this volcano put enough dust into the high atmosphere to block some incident sunlight, causing global cooling of a few tenths of a degree for several years.

Could the warming of the past 140 years have occurred naturally? With ever increasing certainty, the answer is no.

The box at the right shows a remarkable study that attempted to push back the Northern Hemisphere's temperature record a full 1,000 years. Climatologist Michael Mann of the University of Virginia and his colleagues performed a complex statistical analysis involving some 112 different factors related to temperature, including tree rings, the extent of mountain glaciers, changes in coral reefs, sunspot activity and volcanism.

The resulting temperature record is a reconstruction of what might have been obtained had thermometer-based measurements been available. (Actual temperature measurements are used for the years after 1860.) As shown by the confidence range, there is considerable uncertainty in each year of this 1,000-year temperature reconstruction. But the overall trend is clear: a gradual temperature decrease over the first 900 years, followed by a sharp temperature upturn in the 20th century. This graph suggests that the decade of the 1990s was not only the warmest of the century but of the entire past millennium.

By studying the transition from the high carbon dioxide, low-oxygen atmosphere of the Archean to the era of great evolutionary progress about half a billion years ago, it becomes clear that life may have been a factor in the stabilization of climate. In another example--during the ice ages and interglacial cycles--life seems to have the opposite function: accelerating the change rather than diminishing it. This observation has led one of us (Schneider) to contend that climate and life coevolved rather than life serving solely as a negative feedback on climate.

If we humans consider ourselves part of life--that is, part of the natural system--then it could be argued that our collective impact on Earth means we may have a signi cant co-evolutionary role in the future of the planet. The current trends of population growth, the demands for increased standards of living and the use of technology and organizations to attain these growth-oriented goals all contribute to pollution. When the price of polluting is low and the atmosphere is used as a free sewer, carbon dioxide, methane, chlorouorocarbons, nitrous oxides, sulfur oxides and other toxics can build up.

Drastic changes ahead
IN THEIR REPORT Climate Change 2001, climate experts on the Intergovernmental Panel on Climate Change estimated that the world will warm between 1.4 and 5.8 degrees C by 2100. The mild end of that range--a warming rate of 1.4 degrees C per 100 years--is still 14 times faster than the one degree C per 1,000 years that historically has been the average rate of natural change on a global scale. Should the higher end of the range occur, then we could see rates of climatic change nearly 60 times faster than natural average conditions, which could lead to changes that many would consider dangerous. Change at this rate would almost certainly force many species to attempt to move their ranges, just as they did from the ice age/interglacial transition between 10,000 and 15,000 years ago. Not only would species have to respond to climatic change at rates 14 to 60 times faster, but few would have undisturbed, open migration routes as they did at the end of the ice age and the onset of the interglacial era. The negative effects of this significant warming--on health, agriculture, coastal geography and heritage sites, to name a few--could also be severe.

To make the critical projections of future climatic change needed to understand the fate of ecosystems on Earth, we must dig through land, sea and ice to learn as much from geologic, paleoclimatic and paleoecological records as we can. These records provide the backdrop against which to calibrate the crude instruments we must use to peer into a shadowy environmental future, a future increasingly inuenced by us.

CLAUDE J. ALLGRE and STEPHEN H. SCHNEIDER study various aspects of Earths geologic history and its climate. Allgre is professor at the University of Paris and directs the department of geochemistry at the Paris Geophysical Institute. He is a foreign member of the National Academy of Sciences. Schneider is professor in the department of biological sciences at Stanford University and co-director of the Center for Environmental Science and Policy. He was honored with a MacArthur Prize Fellowship in 1992 and was elected to membership in the National Academy of Sciences in 2002.