Finding Time

Geochronologists establish precise dates for events that occurred eons ago

The Hell Creek region of Montana, a happy hunting ground for paleontologists, features sedimentary strata from 67 to 64 million years ago. (Grant Groberg)
The Hell Creek region of Montana, a happy hunting ground for paleontologists, features sedimentary strata from 67 to 64 million years ago. (Grant Groberg)

When did all the dinosaurs die? The short answer is that they went extinct about 66 million years ago (MYA). But no one can check that date by consulting a history book or glancing at a calendar or a clock.

So how is it possible to establish a date for the death of dinosaurs that long ago? For that matter, how was it possible to pinpoint the age of the famous fossil hominid Lucy, unearthed in East Africa in 1974, at 3.18 MYA?

Many fossil bones look much alike, whether their owners were striding around East Africa 200,000 years ago or three million years ago. But with an accurate age, a fossil bone or a stone tool can fit into a meaningful niche in our developmental timeline. Without an accurate date, it’s just another old fossil. Or another old rock.

Geochronologists are the scientists who know how to find the answers to such questions. Geochronology is a fiendishly complex subdiscipline of geology that relies upon a deep understanding of both geology and geography—and of physics, chemistry, soil science, mathematics, and computer science. Locked as we all are in our span of three-score years and 10, we cannot think realistically about a million years of time, let alone a billion years.

Geochronologists routinely deal with time spans in the millions of years. It is how they see Earth and how they think. Among the most accomplished geochronologists in the world are the scientists at the Berkeley Geochronology Center (BGC), an independent nonprofit institution located just off the campus of the University of California. They are the ones who established those dates noted above.

BGC, led by its founding geochronologist and chief scientist, Paul Renne, has five additional principal scientists— Alan Deino, Warren Sharp, Roland Mundil, Greg Balco, and David Shuster—plus several research associates, a lab manager, some technical staffers, and a few postdoctoral fellows. In their state-of-the-art laboratory, they work to establish significant dates and ages in such broad subjects as extinctions, human evolution, and climate change. They have made major contributions to archaeology, paleontology, geology, and paleoanthropology. Currently, they are addressing these challenges:

What, precisely, were the global events that led to the extinction of dinosaurs? Was it the giant meteor impact on what is now the Yucatán Peninsula 66.04 MYA? Or the massive volcanic eruptions in what is now India? Or some combination of those catastrophic events?

Where, exactly, is the San Andreas Fault located in Southern California?

What do ostrich eggshells reveal about human behavior in Middle Stone Age Africa?

Was the emergence of modern humans in East Africa related to climate changes?

Precise dates enable scientists to address larger questions of causality: not just when something happened, but what actually happened, how it happened, and even why. These questions are not merely of academic interest. The proximate cause of the extinction of dinosaurs—and of about 75 percent of all life at the time—was a climate catastrophe.


Geochronologists go about their work in the field and then in the laboratory. The field may be anywhere from California to Ethiopia, from Montana to Antarctica. Fieldwork involves seeking samples of rocks, minerals, and other materials that can be dated, often while crawling around on the ground in inhospitable—even dangerous—conditions. Think scorpions, snakes, heatstroke, and occasional boy soldiers toting AK-47s.

One of the most important fieldwork sites in the world lies in eastern Montana, in the aptly named Hell Creek region, a badlands south of the Missouri River. There, in the jumble of hills, buttes, canyons, and dry washes, sedimentary strata spanning from 67 to 64 MYA are plainly visible in layers of browns, yellows, blacks, and other earth tones. To casual observers, these features are just part of the colorful landscape of the arid West. To an experienced geologist, the layers can be read like timelines in a great text that shows how Earth was formed. The time span visible on the sides of the buttes and canyons in this area covers the late life history and abrupt death of the dinosaurs and beyond. A happy hunting ground for paleontologists, it is the richest dinosaur fossil area in the world, the place where, starting in 1902, many of the world’s mega-dinosaurs were unearthed, including 11 fossil Tyrannosaurus rex remains. Additional T. rex fossil remains were discovered there in the summer of 2016.

Ancient fossil bones cannot be dated directly. Instead, their ages are determined by analyzing the radioactive elements in the minerals present in the exact geological stratum in which they are found. It is that mineral context, not the fossils, that gets dated. Those ages are further “constrained” by dating the strata above and beneath the object fossils.

The Hell Creek Formation (in geology, a formation is a large distinct layer of rock) is truncated by the visible boundary line between two major geologic periods, the Cretaceous and the Paleogene. That stratum is known as the K-Pg (formerly the K-T) boundary, and its well-established date is 66.04 MYA. It marks one of the five great extinctions of life on Earth, a time when an estimated 60 to 80 percent of all animal life perished. The time of this extinction is a matter of observation: beneath that layer, many fossil species including dinosaurs have been discovered. Above it, no fossil dinosaurs have been found.

In the summer of 2017, Renne made his sixth field trip in five years to Hell Creek, in the company of an amateur geologist, John Rea. “There is no other place in the world,” Renne says, “where we see terrestrial ecosystems through that time interval with the degree of resolution that we have in Montana.” Every day for two weeks in mid-July, with temperatures topping 100 degrees, Renne and Rea explored selected small sites in a vast shadeless landscape of dusty broken hills and buttes.

This was Renne’s latest effort to refine the dating of one of the known precipitating events of the K-Pg extinction, when a large meteor smashed into Earth at an estimated 45,000 miles per hour. Named the Chicxulub impact, for the Mexican village nearest where it hit, the four-to-six-kilometer-wide object punched through Earth’s crust, left a 110-mile-wide crater, and created lasting havoc in the global environment. It also left a thin layer of an element rare on Earth but common in asteroids and meteorites—iridium—all over the world. That thin gray streak of iridium, first identified in 1980 by the father-and-son team of Luis Alvarez, a prominent physicist, and Walter Alvarez, a geologist at Berkeley, pegged the sedimentary layer corresponding to the Chicxulub impact. The ubiquity of iridium affirmed that the event was global.

At sites in Montana where the iridium layer is visible on the surface, Renne and Rea are seeking in volcanic ash the presence of sanidine crystals, which are the preferred material for a dating method known as argon-40/argon-39. Sanidine is a volcanically deposited feldspar. Its crystals contain traces of an inert gas, argon, and appear to the naked eye as a pinpoint sparkle of reflected sunlight. In this formation, sanidine is commonly seen in black layers of lignite, an ancient immature coal that looks and feels like crumbling charcoal. The geologists are seeking crystals adjacent to the iridium layer. “We have more than 50 individual ash beds here that occur within about a million-year interval above and below the K-T boundary,” Renne says. “So we can dig with amazing precision.”

At each site, Renne and Rea squat or stretch out on the ground and scrape away the surface of eroded clay and siltstone, and with geologist’s hammers (Renne prefers to use a larger ice ax), drag out handfuls of hitherto-unexposed lignite, looking for layers of ash as thin as one millimeter. They examine each dusty handful first with the naked eye and then with the 12-power hand lens they carry. Upon finding a sample that is a keeper, they spend up to several hours extricating the ash from the lignite, eventually filling a gallon cloth bag, writing an ID number on the bag with a black Sharpie, and placing it carefully in their backpacks. Meticulous note takers, they then record the precise GPS coordinates of the site. It is dirty, tedious work that in this landscape calls for alertness to scorpions.

One afternoon, Renne and Rea sprawl in the scree on a precarious slope of a windswept butte, scraping and searching in the lignite for hours under a merciless sun, until Renne sits up and says, “Bingo!” He is peering through his scope at a handful of ash. In it, he has found sanidine crystals that are a useful size and shape for 40-Ar/39-Ar dating.


Paul Renne, the director and chief scientist of the Berkeley Geochronology Center, looks for sanidine crystals in lignite in the Hell Creek Formation.


“Argon-argon is our bread and butter,” Renne says. “It works from about 2,000 years ago back to the early solar system and is among the most precise dating tools available.” Argon works well as a timekeeper because it is the product of the long-term decay process of potassium, which is an abundant radioactive element often found in volcanic deposits and especially concentrated in the mineral sanidine. Potassium decays over a lengthy half-life to form an isotope, argon-40.

“We know the rate at which the conversion from potassium-40 to argon-40 happens,” Renne says, “and knowing that rate, we just have to measure how much of the ‘daughter’ isotope is present, compared to the ‘parent.’ It’s the ratio of the two, which is completely independent of the actual amount present. That’s the beauty of it. And that is true for all the different radioactive processes that we use for dating.”

To get to that ratio of daughter to parent, the sanidine crystals are separated from the surrounding materials in the laboratory via gravity and flotation. The thousands of tiny individual crystals are picked out by hand with a moistened camel-hair brush. A lab technician selects the clearest and most angular crystals, reducing the number for analysis to 20 to 100 crystals.

After further processing, a laser gradually heats the sample to release the argon gas from the crystal. The gas is introduced into a mass spectrometer, which tallies the individual isotopes present. The result is a date in years, with a calculated plus-or-minus margin of error. The software that governs this process, named Mass Spec, was designed by Alan Deino, one of BGC’s other principal scientists. The center distributes its proprietary software free of charge to other geochronologists around the world.

Over the 20-plus years since BGC was founded, geochronologists have worked to refine their processes to arrive at ever-more precise dates. The initial effort to date the Chicxulub impact put it at 65 MYA, plus or minus nearly a million years. That large margin of error limits the date’s value for correlating other ancient events or determining causality. But after many procedural refinements, the margin of error has been significantly reduced.

“In 2013, we published the most precise age yet established for the impact,” he says. “That’s 66.04 million years, plus or minus 10,000 years.”

For more than a decade, Renne and his colleagues have been trying to further refine the date of the dinosaurs’ demise, and to theorize what actually killed them off, by determining more precise dates for the two most likely trigger events, the Chicxulub impact and the ancient flood-volcanic activity in India known as the Deccan Traps. Most scientists agree now that the worldwide extinction is related to both events, but establishing causality is both tricky and important. Tricky because there are some indications that the dinosaur die-off began prior to the Chicxulub impact, and because the Deccan volcanism, which occurred on a scale unimaginable in our historical experience, persisted over a period of 800,000 years. Important because both events profoundly compromised Earth’s climate for millennia in ways that made survival impossible for most life at that time.

On his most recent trip to the Hell Creek Formation, Renne passed one steamy afternoon searching for sanidine in a lignite seam above and near the location of the most recent adult T. rex find: the fossil bones of the Tufts-Love Rex, named for two Burke Museum paleontology volunteers, Luke Tufts and Jason Love. They were scouting sites in a remote canyon when they noticed dinosaur bones jutting from a sandstone hillside. The T. rex  skull and about 20 percent of the bones were excavated in 2016 by a team led by the University of Washington and the Burke Museum, and their excavations resumed in the summer of 2017. Initial estimates indicate that this T. rex dates to about 66.3 MYA, not long before the K-Pg mass extinction. For Renne, this was an opportunity to nail down more precisely the death date of this one dinosaur.

Renne points out that no one is ever going to find and date the last dinosaur. “How lucky would we have to be? For every outcrop, every exposure of this scenario, there are billions of times more material still under the surface that’s not exposed.”

Further, he notes, “there is a lot of discussion about how quickly the extinction happened. Many paleontologists think the dinosaurs and some of the early mammals were actually in decline as much as a million years before the K-T boundary.” In other words, “Was everything going just fine for the dinosaurs right up until the Chicxulub impact and then boom? Or were things sort of struggling and the impact just kind of finished them off?”

Much more is now known about the Chicxulub impact, its global climatic consequences, and the extinction of dinosaurs than is known about the scale, duration, and atmospheric consequences of the prolonged Deccan Traps eruptions. “We can clearly relate the end of the dinosaurs to an impact,” says Renne. But about the Deccan Traps? “We also know that this massive pile of lavas in India had erupted over that same time interval. But where in that pile is the time horizon that corresponds to the impact and the mass extinction?”

The two cataclysmic events may even be related in some complex ways, Renne and his colleagues hypothesize—and it is noteworthy that some earlier mass extinctions coincided with catastrophic volcanic events—but answers will depend on accurate dating of the Deccan Traps.


BGC geochronologists commonly work on several dating challenges at once with different dating methods. Mostly they work with teams of scientists exploring events, phenomena, and evolutionary developments in the ancient past. The results of these scholarly efforts culminate in papers in such major journals as Science, Nature, Geology, the Proceedings of the National Academy of Sciences, the Journal of Human Evolution, the Journal of Geophysical Research, and many more—hundreds of such scholarly papers since 1994.

Warren Sharp, who has long worked at BGC on dating projects related to human origins, earthquakes, and paleoclimates, is focused on events “in the recent past, the past 500,000 years.” Among his subjects of inquiry: the evolution of modern human behaviors in Africa.

Researching the Middle Stone Age and Later Stone Age, roughly 300,000 to 30,000 years ago, the time when earlier hominid species in Africa evolved into Homo sapiens, Sharp is dating long-inhabited cave and rock shelter sites in Kenya and South Africa. He and colleagues from Harvard and the University of Bergen are seeking to learn whether archaeological evidence of evolving human cognition—such as improved stone tools, uses of ocher for adornment or decoration, the trade in shell beads—are coincident with climate changes. Relating such changes to human behaviors requires precise dating of widely scattered habitation sites in Africa. Sharp is using a long-established method, uranium-series dating, and applying it in a novel way to a common archaeological material.

The material used in this case is ostrich eggshell, and the “nuclear clock” is the decay of uranium-238 to thorium-230. Ostrich eggs were a common food in those times and places, and the ancient eggshell fragments, which Sharp notes are “pretty sturdy little things,” contain traces of uranium picked up when they were buried. The samples, painstakingly gathered with dental picks, are eggshell fragments a few centimeters square. For the dating to be valid, the exact location of each sample in the habitation site must be identified in three dimensions, by stratigraphic layer and by GPS-derived x-y coordinates.

Then back in Berkeley, they turn a section of each sample into an aerosol and run it through one of the lab’s eight specialized mass spectrometers, along with a known quantity of “tracer” atoms for calibrating the sensitivity of each analysis. The mass spectrometer counts and compares the ions of uranium-238 (the parent element) and thorium-230 (the daughter isotope) in each sample. As in the other radiogenic methods, the ratio of the two elements plus the known rate of decay make it possible to calculate how long it took the thorium-230 to build up—and thus calculate the age of the sample. They typically process 30 to 40 samples per site, which yield an average date with a precision of plus or minus about 1,000 years. That is precise enough to begin to link the age of a habitation site with the physical evidence and with known information about the climate and vegetation in which those early Homo sapiens lived. The age findings do not prove cause and effect, but the linkage is clear enough to warrant ongoing research on when and possibly how Homo sapiens achieved that level of cognitive behavior.

Best Ways to Get a Date

Most geochronological dating is radioisotopic dating, so called because some elements include unstable radioactive isotopes that decay over time into “daughter” isotopes at known and constant rates. “If you know the decay rate,” Paul Renne says, “then all you need to know to calculate an age is the ratio of parent to daughter isotope in a sample.”

Another important factor in geochronological dating is stratigraphy, the study of layers of sediments, rocks, and other materials that were deposited over time on the Earth’s surface. Those “layer cake” strata, commonly seen on mountainsides and road cuts, can be dated by various methods.

The most useful dating methods are these:

Radiocarbon dating is based on the rate of decay from carbon-14 to nitrogen-14. Carbon-14 is present in all living organisms and all organic matter. When an organism dies, its uptake of carbon-14 ceases, and its age can be determined by measuring the ratio of carbon-14 to nitrogen-14 in a sample. But radiocarbon dating is effective only for the past 50,000 years.

Uranium (U)-lead (Pb) dating is based on the decay of uranium isotopes U-238 and U-235 through many intermediate isotopes to the stable element lead. It is most useful for dating one million to billions of years before the present.

Uranium-series dating relies upon the many different half-lives of radioactive daughter isotopes of uranium. The decay from one isotope to another can be used to date materials over long spans of time. This is a very precise method, useful for paleoanthropological dating over the past 600,000 years.

40-Potassium (K)/40-Argon (Ar) dating is based on the common presence of potassium in volcanic rock. The 40-K decays to 40-Ar at a known rate, and accumulation of 40-Ar begins at the time of volcanic eruption. So a date can be calculated based upon the ratio of the two isotopes in a sample. This method is effective for dates as old as billions of years.

40-Ar/39-Ar dating is a refinement of K/Ar dating that resolves some of its limitations, largely superseding it.

Thermoluminescence (TL) and Optically stimulated luminescence (OSL) are comparatively new dating methods that bypass the age limitations of carbon-14 dating and the U-series and argon dating dependence on volcanic deposits. Their effective age range is 1,000 to 500,000 years.

The principle for both methods is the same: many minerals absorb natural radiation from their immediate environment and from cosmic rays. Once buried, radiation—that is, negatively charged electrons—is trapped in the crystals of buried minerals. It builds up over time, like a battery being charged. Samples, commonly quartz and potassium feldspar, must be obtained in complete darkness to avoid light contamination. When they are excited by heat or light, they emit visible light (luminescence) equal to the total radiation absorbed by the sample since burial. Age is determined by multiplying total radiation in the sample by the known dose rate.

Paleomagnetism, a widely used non-radioisotopic method of dating, relies on the fact that the Earth’s magnetic field has reversed polarity many times, from north to south and back. Those ancient reversals left detectable magnetic patterns in ferrous minerals. The patterns of normal and reverse polarity have been correlated with dated geological strata over many years, thus establishing a time scale that can constrain dates of sediment depositions and fossils.

—Michael W. Robbins


Another area of deep-time research focuses on paleoenvironment and specifically paleoclimate in two long-running studies with ongoing deep-drilling projects. BGC geochronologist Roland Mundil has been working with the Colorado Plateau Coring Project—a multi-institution geological research effort that drilled a core through millions of years of strata in the Petrified Forest National Park in Arizona. The project focuses on a part of the Triassic Period of 230 to 205 MYA, “the time of the early rise of the dinosaurs and other critters,” says Mundil. To a drilling core extracted from a deep hole in the Petrified Forest National Park, Mundil is applying uranium-lead dating, along with paleomagnetic dating. The sampling material that Mundil uses is zircon, a mineral in the local sandstones. Zircon contains “a bit of uranium,” which, with samples taken from every 10 meters of core, enables the uranium-lead method to provide dates over the millions of years of the Triassic Period.

The goal of this project is to establish a clear, continuous sequence that can be correlated with Triassic geochronology elsewhere in the world, and to gather biotic and environmental data on the pace of ecosystem and evolutionary changes.

On the other side of the world, Mundil’s BGC colleague Alan Deino has been dating stratigraphic sequences with a long-term drilling project in East Africa. It involves boring through a sequence of paleolake basins in the Rift Valley from southern Kenya to northern Ethiopia. These drilling projects reach strata that range in age from 3.3 MYA to the present. In the most recent of these efforts, researchers have sought paleoclimate data and—to quote the title of a 2017 paper (of which Deino is a co-author)—“Their Implications for Understanding the Environmental Context of Hominin Evolution.” The core sequences from these drilling projects include such climate indicators as lake-bed sediments containing pollens and diatoms (planktonic algae). Deino dates these sequences in detail using argon/argon dating, which is possible because of the pervasive ancient volcanic activity in the Rift Valley.

Deino has spent much of his career focused on human evolution and the East African environmental context in which that evolution occurred. Recent data from southern Kenya’s Olorgesailie Basin indicate that the Middle Stone Age was a time of rapid climate changes. It may not yet be possible to pinpoint the times and places of the emergence of Homo sapiens. That is largely because there is scant hominin fossil evidence from this period in East Africa. The strongest such evidence for Homo sapiens’s emergence currently falls into the 300,000–200,000-year range.

Other tantalizing indicators suggest, however, that modern humans may have emerged earlier than those dates. Some genetic calculations point to a period 350,000 to 260,000 years ago. In that time, which includes the transition from what is called the Acheulean culture to the Middle Stone Age, there is substantial and widespread archaeological evidence of major changes—technological, socioeconomic, and symbolic—in hominin behavior in East Africa.

That those behavioral changes are congruent chronologically with periods of rapid environmental and habitat change is spelled out in a remarkable series of three papers published in spring 2018 issues of Science by an international team of scientists associated with the Smithsonian’s decades-long Human Origin Program. BGC’s Deino is the lead author of one paper and a coauthor of the other two.

Together, these papers, which are undergirded by precise dating of samples from Olorgesailie, suggest that the developments in toolmaking, in long-distance trade in raw materials like obsidian, and in symbolic uses of mineral pigments like ocher are indicative of evolutionary changes in early modern humans. The many changes in behavior between 350,000 and 295,000 years ago are at least suggestive of the emergence of Homo sapiens nearly 100,000 years earlier than had previously been established, and they hint at possible linkages between human evolution and ancient climate change.

Permission required for reprinting, reproducing, or other uses.

Michael W. Robbins is a historian and former editor of both Audubon magazine and MHQ: The Quarterly Journal of Military History. His most recent book is Lest We Forget: The Great War.

● NEWSLETTER

Please enter a valid email address
That address is already in use
The security code entered was incorrect
Thanks for signing up