How to kill a dinosaur in 10 minutes

by Paul Braterman

Ten minutes difference, and Earth would still be Planet of the Dinosaurs

We have suspected for some decades that the dinosaurs¹ became extinct as the result of a massive meteorite, an asteroid, hitting the Earth. We have known where the impact site was since 1990, if not before. But it is only last year that we successfully drilled into the impact site, and only now, for the first time, do we really understand why the impact was so fatal. And if the meteorite had arrived ten minutes earlier, or ten minutes later, it would still no doubt have inflicted devastation, but the dinosaurs would still be here and you wouldn't.

Too many suspects

L: Ammonites (Haeckel, 1904, via Wikipedia). Click on this and other images to enlarge

66.1 million years ago, dinosaurs covered the Earth. 66 million years ago, there were none. And not only the dinosaurs, but the pterosaurs in the skies, the long necked plesiosaurs and even the ammonites in the oceans, and 75% of all complex animal life. No terrestrial vertebrate heavier than around 25 kg seem s to have survived. What happened?

There was no shortage of theories. Quite a lot was going on, geologically, at the time. There was massive volcanic activity in India, giving rise to what are known as the Deccan Traps, containing over a million cubic kilometres of basalt. Such major volcanic episodes have been connected with other mass extinctions. They are accompanied by the ejection into the stratosphere of sulphur dioxide, which in turn slowly reacts with atmospheric oxygen and water vapour, forming a haze of sulphuric acid far above the Earth's surface. This partly blocks out the Sun, affecting the plant growth on which almost² all life on Earth depends. We saw such an effect on a much smaller scale with the 1991 eruption of Mt Pinatubo in the Philippines, which caused a two-year slowdown in global warming and reduced crop yields worldwide, far beyond the reach of the actual dust cloud. Continents were on the move, with the reopening of the North Atlantic, polar icecaps reappeared after prolonged absence and led to lowering of sea levels, affecting ocean productivity, and all these things may have added to ecological stress. But all this hardly seems enough for such a wide-reaching die-back, nor does it account for the suddenness of the process.

Above, Mt Pinatubo, caldera formation phase, USGS

And so, we saw the emergence of the asteroid impact hypothesis. The Solar System still contains fragments of rocky material left over from its formation. This material is mainly concentrated in the asteroid belt, between Mars and Jupiter, but small fragments continually come our way. Most of these are tiny particles, which burn up in our atmosphere (meteors or shooting stars). Some are large enough to survive, falling to earth as meteorites, and their chemical composition tells us a great deal about the stuff from which Earth was originally formed. Very occasionally, the meteorite is large enough to dig out a crater, such as Meteor Crater in Arizona. During the early years of the Solar System, major impacts were frequent and dramatic, as we can see by looking at the Moon and at the other inner planets, but on a geologically active planet such as Earth the evidence will long since have eroded away. However, there are still plenty of small asteroids whose orbits intersect the Earth's. What if one of these happened to crash into the Earth, and trigger catastrophe?

L: Meteor Crater, Arizona; image from Visitors Center site.

Such an idea was proposed in 1973 by Harold Urey, the leading geochemist of his time (now most widely remembered for the Urey-Miller synthesis of the building blocks of life).

The search for a smoking gun

But would such a collision have left any trace, still detectable 66 million years on? In a word, yes. Worldwide, there is a layer of mudstone separating the rocks of the Cretaceous period from those of its successor, the Palaeogene. Up to this K/Pg (Cretaceous-Palaeogen)³ boundary, you find dinosaurs, ammonites, and the other creatures I mentioned. Above it, none; but then, within a few million years, a dramatic proliferation of mammals as they evolved to fill the niches left vacant by this mass extinction. Yet there is no difference in the nature and chemical composition of the rocks directly above and below the boundary layer, showing that the boundary event was a single catastrophic occurrence, rather than part of a long-term trend. And embedded in the boundary mud are glassy spherules, like those formed when molten rock ejected by volcanoes cools in the upper atmosphere. There are also traces of soot, microdiamonds, and shocked quartz, a substance formed from normal quartz when subjected to enormous sudden forces.

Below, R: Glass spherules from Wittenoom impact layer, Hamersley, Precambrian. Image by Traveling Geologist. Click to enlarge

A crucial question here is how long it took to form this layer, and the father and son team of Louis and Walter Alvarez thought they had an ingenious way of finding out (Louis, the father, was a Nobel Prize nuclear physicist; Walter, his son, a geologist studying the boundary layer where it lies exposed in Italy's Appenine mountains). They would test the layer for the element iridium. Iridium is rare in the solar system, as we can tell from analysis of meteorites, but it is even rarer in the Earth's crust. That is because iridium is very soluble in iron, so that when the Earth's iron core formed, it took most of the iridium down to the centre with it. (The same, incidentally, is true of gold.) Yet some iridium rains in on Earth continuously from meteors, at a rate that can be estimated, so the total amount of iridium in the layer would tell them how long it took to accumulate.

Below: Louis and Walt Alvarez at the exposure of the K/Pg boundary layer in Gubbio, Italy. US Govt image

To their surprise, the boundary layer was enriched twenty-fold in iridium, relative to the rocks above and below, and other geologists soon confirmed high amounts of iridium in that layer wherever it was exposed. Where could it all have come from? Nowhere on Earth; comparison with natural waters showed that no amount of concentration could have generated such amounts. The boundary layer also anomalously high amounts of other metals, including nickel, cobalt, and zinc. (As I have said before, scientists where possible seek multiple lines of evidence.) And within three years, the iridium spike in the boundary layer had been confirmed at 50 other sites around the globe.

So we are looking at the debris thrown up by the impact of an iridium-containing extraterrestrial body, that is to say a small asteroid. Moreover, the K/Pg boundary corresponds precisely to the general mass extinction. Not only of the dinosaurs, but of ammonites, and even many single-celled marine organisms that flourished up to the very boundary but are completely absent beyond it. Cause and effect?

Rush to judgment?

The general public readily accepted this idea. The scientific community was more circumspect, and rightly. As I already mentioned, there were a lot of other things that might have been involved. The sharpness of the extinction event was for some time a matter of controversy, obscured by statistical fluctuations and by local die-backs in the last million or so years before the impact, while checking that the impact and the extinction happened at the same time required precision dating by the then still relatively novel argon/argon method. The history of the controversy is detailed in a full-length book (Night Comes to the Cretaceous, critically reviewed here), while the Wikipedia entry on the timeline of the research has over 130 references. As late as 2010 the presitigious journal Science published (here and here) papers arguing over single versus multiple causation. It was not until 2013 that we had sufficiently precise datings for the impact and the extinctions to be sure that they really coincided, and even then debate continued as to whether we should consider the impact as a single dominant cause, or merely as the last straw for a biosphere already under stress.

First find the gun

The next step, then, is to find the crater and see what it tells us.

We can roughly locate the impact by the thickness of the boundary layer. This is measured in centimeters away from the Americas, rising to metres in northern Mexico and tens of metres further south, pointing to a location in the Caribbean. More than one group seems to have discovered the exact location. The area was explored from the 1950s onwards by Pemex, Mexico's state-owned petroleum company, who understandably regarded their findings as commercial secrets, but eventually released drill samples showing shocked quartz. Further information came from studies of gravitational anomalies (places where the Earth's gravitational pull is slightly stronger or weaker than average), and aerial photography. Stronger gravity signifies pileup of material, as at the rim of a crater, while aerial photography shows sinkholes, places where the rock has been weakened. In the figure below, the US Geological Survey has combined both of these in a single map of the impact area.

The map shows the northern coastline (white line) of the Yucatán peninsula at the southern tip of the North American Plate, and the sea area directly to the North of it. The entire region is a limestone plateau, formed (and, where submerged, still forming) by deposition from the shallow waters of the continental shelf. The gravitational map clearly shows two concentric rings, known collectively as the Chicxulub crater, after the Mayan name of the village nearest the centre. The centre itself is a few miles offshore. The outer ring is 120 miles across, the inner ring some 55 miles. The white dots are sinkholes, such as commonly occur in limestone landscapes, in an arc along the outer ring, corresponding to weakness in the limestone. What does this mean?

Most of us have seen slow motion images of drops hitting water, forming a crater with an inner mound (see image, R), before the water collapsed back down again. But recently I saw one where the inner mound was so large that it was unstable under its own weight, and collapsed to form an inner ring within the crater. Similar effects are predicted when a sufficiently large impact occurs at the surface of a moon or planet. While the shock wave of the impact is travelling through the rock, the rock moves almost like a liquid, with a column of material bouncing upwards, and these features are preserved when the shock has passed and the rock once more sets solid. So what we are normally left with is a crater with an outer rim and a central impact mound, and we see such features routinely on the Moon. But with the very fiercest impacts, as with the largest water drops, the central mound it so high that it collapses outwards under its own weight, and instead of a central peak we are left with what is called an impact ring. This is what we are looking at here; an outer crater rim, and an inner impact ring. The Chicxulub crater is the only one on Earth known to show this kind of structure, although it has been observed in craters on the Moon and on Venus.

L, below: Schroedinger Crater on moon, showing impact ring concentric with crater rim

If we want to sample the rocks most severely affected when the crater was formed, that impact ring is the right place to look. And so, last year, a team involving 38 scientists from 34 separate institutions in 11 different countries, under the auspices of the International Ocean Discovery Project, drilled (see here and here) through overlying younger sediments into the impact ring. The drilled cores are now in a repository in Bremen, Germany, where they are stored at low temperature to prevent the growth of bacteria, and can be examined at leisure. We will be learning more about them in detail over the next year or so, but an overall picture is already emerging.

The findings tell us just how drastically the Earth's crust was deformed by the impact, and finally give us the answer to those two long-standing questions; was it really this impact that killed off the dinosaurs, and, if so, how exactly. The investigation was the subject of an excellent recent BBC programme, The Day the Dinosaurs Died, on which the rest of this post draws heavily, and which I hope will be made widely available for educational viewing.

Here's what the drilling found:

As expected, 600m thickness of marine deposited sediments, dating from recent (top) through Palaeocene (bottom), as shown by examination of index fossils.⁴ Confirmation that at the time of the impact this area was, and has been ever since, part of a shallow sea at the southern tip of the North American continent, lying between the Pacific to the West and the newly opened mid-Atlantic to the East. The sediments, mainly limestone (calcium carbonate) and gypsum (calcium sulphate), characteristic of such an environment. This will turn out to be extremely important.

At the base of this section, a 12m thick layer of sand, coarser at the bottom. This is ascribed to the tsunami produced by the impact, with the coarser grains having fallen out from the water more quickly. Beneath this, we have the rocks disturbed by the actual impact. There is a layer of rock that had been molten, with different texture and colour, followed by a jumble of granite derived from deep beneath the impact site, with fragments of the pre-impact sedimentary rock that had lain above it. There is shocked quartz, and the granite fragments also show shock damage, and are so greatly weakened that they can be snapped by hand. And significantly, there is a shortage of the expected fragments of gypsum in the entrapped sedimentary debris.

The asteroid itself would no doubt have been volatilised by the energy of the impact. So large an object could not have been appreciably slowed down by air resistance, and would have landed at a speed of some 40,000 miles an hour into the shallow sea. As for its size, we can estimate this in a rather surprising and indirect way. Nuclear testing in Nevada released known amounts of energy, and left craters containing shocked quartz. By comparing what is observed there with the Chicxulub crater system, we can estimate the total energy of the impact. It works out at the equivalent of 10 billion Hiroshimas. From the energy and the speed, we can work out the total mass, and hence the size. About nine miles across, a little bit larger than the Alvarez estimate, but still I think remarkably close. Since the most reliable Alvarez estimate was based on the amount of iridium, I suspect that their shortfall may be due to part of the impactor vapour having escaped earth gravity altogether.

Then survey the carnage

We can now estimate what the effects would have been at different distances, using what we know about the behaviour of materials under extreme forces, and compare these estimates with what we observe.

The initial impact would dig a hole 20 miles deep, and 120 miles across. Deep rock would rise as high as the Himalayas, and then collapse to form the observed impact ring, all within 10 minutes. The impact fireball would kill everything within a 600 mile radius. A shock wave would spread out through the solid earth, and a monstrous tsunami through the oceans. The atmosphere would have seen hurricane force winds, and rock condensing in the upper atmosphere would produce a rain of tiny glass spheres, such as those still preserved in the mudstone boundary layer. The glowing hot rockdust would generate heat several times greater than that coming from the Sun, roasting to death all surface animal life up to a distance greater than a thousand miles, and starting firestorms in the vegetation.

Below: hadrosaur family tree, in the 30 million years up to obliteration by impact. Debivort via Wikipedia. Note scale. Click to enlarge

We can see the effects of the tsunami at Edelman Fossil Park in New Jersey, 1700 miles from the impact site. There we find the expected range of Cretaceous remains up to the boundary layer, with a clear cut-off above it. The boundary layer itself is around four inches thick, but full of a jumbled assortment of fossils, and fragments of wood. There is a surprising mixture of species, bird next to crocodile next to giant turtle, land creatures next to shallow water turtles next to deep sea turtles. No tooth marks, no scattering of limbs; this was sudden indiscriminate killing.

L: Nothofagus (southern beech tree). Genus dates to before the impact. Still with us.

At the tip of South America, in Patagonia, 4000 miles from the impact, abundant fossils show that at the time of the impact the area was home to herds of hadrosaurs, duck-billed dinosaurs thirty foot long. It was also home to trees not too different from the southern beech trees still flourishing in the area. So the trees survived, but the dinosaurs didn't. Why not?

At last, the gunsmoke

The drilled cores give us the answer. As I mentioned earlier, they are surprisingly deficient in gypsum, although gypsum is abundant both in the more recent sediments, and in nearby sediments laid down before the impact. So the missing gypsum must have been destroyed by the heat of the impact, sending sulphur dioxide gas up into the atmosphere.⁵ We already know from studies of volcanic eruptions that over months or years this sulphur dioxide would slowly react in the atmosphere to form a haze of sulphuric acid droplets, scattering sunlight and cooling the Earth. That haze, rather than simply dust, is how the Mt Pinatubo eruption affected climate.

Extrapolating from Mt Pinatubo and other recent eruptions, and taking into account the enormous volume of gypsum sediments in the impact zone, we can estimate that the haze produced by the Chicxulub impact would have blocked enough sunlight to reduce temperatures worldwide by more than 10^oC (18^oF), while acid rain would also have contributed to the death of much marine life. On land, trees would have shed leaves and shut down as if for what would turn out to be an unusually long winter. The dinosaurs, hugely diverse and successful as they had been through some 180 million years, had no such way of adapting. The herbivores, and the carnivores that fed on them, perished.

Where did those ten minutes come from? From the rotation of the Earth. The asteroid is falling towards Earth on a fixed trajectory, but the Earth itself is spinning beneath it, one revolution every 24 hours. This corresponds to around 1,000 miles an hour in the region of interest. So arriving ten minutes earlier or later would have placed the impact some 150 miles further to the East or West. And if this had happened, the asteroid would have missed the shallow gypsum-rich continental shelf, and encountered only the oceans on either side. No gypsum in the impact zone, no sulphuric acid haze, no long deep winter. While things might have been pretty rough for anything living within a couple of thousand miles or so, the rest of the world would hardly have noticed.

R: Ancestral primate (~3 mY post-impact), artist's reconstruction

The fossil record tells us that over the following months or years, ferns, and then other plants, resumed their cycle, and whatever animals survived would be free to adapt to new niches. Who survived? Small creatures able to feed on insects and seeds, not dependent on the products of day-by-day photosynthesis. And larger creatures if they can survive for long periods between meals, and become torpid in cold weather. Snakes and crocodiles, birds, and those small mammals that had managed to survive in the shadow of the dinosaurs. Among them, our ancestors.

I thank Mark Boslough (Sandia), Phillipe Claeys (Free University Brussels), and Paul Marshall (University of North Texas) for helpful discussions. Drop image, blue dye into milk; note that it is milk material that forms the inner column. T Blackwell via Wikimedia Lunar crater image NASA, enhanced James Stuby, via Wikimedia. Primate image (Purgatorius Unio), drawing by Nobu Tamura via Wikimedia

1] More correctly, non-avian dinosaurs. Birds fit in among the dinosaurs in much the same way that humans fit in among the great apes.

2] The exceptions are those ecosystems based on the reactions of minerals, such as the rich communities surrounding hydrothermal vents.

3] Formerly known as the Cretaceous-Tertiary (K/T) boundary

4] The method is based on the fact that the exact forms of many marine microorganisms changed over time, so that sediments with similar fossils must be similar in age. The method is 200 years old, and gives reliable relative dates, now calibrated against absolute dates determined radiometrically.

5] This is not a new idea; see here and here.

Equations: On heating, CaSO4 –> CaO + SO3 and 2SO3 –> 2SO2 + O2 (CaO will dissolve in the melt, or will react with atmospheric CO2 to make calcium carbonate)

Slowly, in the upper atmosphere, 2SO2 + O2 + H2O –> H2SO4; this sulphuric acid will form tiny droplets, scattering sunlight, and the haze will take months or even years to disperse.