Emeishan Traps

Transcription

[APPLAUSE] FRANCIS A MACDONALD: All right. Trudging up here in my snow boots, I'm really excited to introduce Mark Richards, who is the Professor of Earth and Planetary Sciences at University of California Berkeley. Professor Richards's research is focused on understanding really large-scale dynamic processes of the interior of the Earth and other planets and how those processes really affect geological phenomena that are observable at the Earth's surface, so that is he combines physical theory with geological observations to test hypotheses about the nature of the Earth and how the Earth works. And Mark-- he's best known for his work on how thermal anomalies deep in the interior of the Earth, that they result in mantle plumes, which come up to the surface and create these large igneous provinces and hot spots. You can think about Hawaii, or the Galapagos, or the topic of today's lecture, which is a Paleo example, which are the Deccan Traps, a large igneous province in India. Professor Richards, he earned his BS in engineering science from University of Texas in 1977 and his PhD in geophysics from Caltech in 1986-- my alma mater. He joined the University of Oregon faculty in 1987. And 1989, moved to UC Berkeley where he's held many, many leadership roles. And I'll bring out a couple of these because these are impressive considering the science that he's currently doing. From '97 to '99 he served as the Chair of their Earth and Planetary Science Department. And then from 2002 to 2014, he was the Dean of the Mathematical and Physical Sciences. And from 2006 to 2014, he was the Executive Dean for the College of Letters and Sciences. Now during that time, he tripled the amount of fundraising for the colleges, and he spearheaded integration of different disciplines to create an innovative, "big ideas" courses for undergraduates and made significant contributions to the increase in minority representation in STEM fields. Now keep this "big ideas" concept in mind. Despite all these duties that he was doing at UC Berkeley, Mark has managed to continue his research and come up with his own big idea. And particularly, his big idea that he's going to talk about today is a new idea on how and why the dinosaurs went extinct 65 million years ago. And this is such a big idea, that The New York Times decided to cover it in Sunday's Opinion. So instead of me going into it, I'm going to let Mark go into it. And I'm happy to introduce Mark. Let's welcome him. [APPLAUSE] MARK RICHARDS: OK. So we start here. The subject of the Cretaceous-Tertiary mass extinction is the subject of an infinite debate and interest. And now this is not working, so I'll use this. There we go. And this is-- of course, we lost T. rex. This is one of the largest mass extinction events of the Phanerozoic era, the last 600 million years or so. Wiped out about 70% of species-- marine and terrestrial-- and killed the non-avian dinosaurs. Of course, the birds are dinosaurs, and they managed to fly through this unscathed. And most interestingly and probably the reason for fascination, is that it led to this. This is a diagram. I'm not a paleontologist, so disclaimer here. But this is a diagram of the explosion indicating the explosion of mammalian species at the Cretaceous-Paleogene boundary. The formal boundary is known as Cretaceous-Paleogene, not actually Cretaceous-Tertiary. So here's the impact. Here is the boundary. And these little critters, furry things called mammals that were crawling around under rocks and stones, suddenly blossomed when you got rid of the dinosaurs. And that's why we're here. And so the ecological niches were opened up, and intelligent life evolved. And if not for whatever happened 66 million years ago, intelligent life probably would still be waiting to bloom on this planet. So it's an interesting event. And 66 million years is the wink of an eye in the 4.5-billion-year geological history of the planet, so it was actually quite recent. I mean, these critters that died off had arms, and legs, and brains, and eyes, and stuff like that. They weren't all that genetically different from us. In 1977, Walter Alvarez, a young geologist at Berkeley, brought this rock-- literally this rock-- to his father, the famous Nobel Laureate Luis Alvarez, at Berkeley. And he had a question. This is in a hand sample that's much smaller than this picture, of course, the Cretaceous-Tertiary boundary. These are the Cretaceous limestones here from Gubbio, Italy. These are Cretaceous limestones that are pink. And if you look closely at this, you can see these little dots. These are foraminifera, little critters that live in the sea. And these are large enough to be seen by the naked eye. Whereas, the foraminifera that survived the Cretaceous-Tertiary boundary are too small to be seen because the big ones all died off. In fact, most of them died off and replaced by other species. This is a very abrupt boundary, seemingly so, in the paleontological record. What Walter wanted to know-- there's this boundary. It was called the boundary clay. That's a few centimeters thick at this location. How much time was represented by this boundary clay? Now this turns out to be a profound question geologically speaking. And especially 37 years ago, when plate tectonics was relatively young and what are called the uniformitarian ideas of Charles Lyell's great treatise in geology from the 19th century were holding great sway. And the idea of catastrophes in the geological record were very taboo. And it's very interesting to be saying this in the land the Stephen Gould, of course, who modified that view considerably as well. But there was, needless to say, the idea to be tested was how much time elapsed here. Was this a day? Thousands of years? Or even millions of years? And Walter thought that his father Luis might have some whiz bang physical method for figuring out how much time there was. And he did. He decided to look at the platinum group elements because they come in with cosmic dust at a relatively regular rate and thought that if they could figure out how much the abundances the platinum group elements were, then they might be able to figure out how much time had elapsed here. Now you all know that-- most of you know the punchline. What was discovered was that one of those elements they examined by neutron activation analysis turned out to be an enormous spike of iridium that could not be accounted for by any terrestrial cause. And so they invoked the famous Alvarez hypothesis of 1980, which is probably the most cited paper in the entire geological literature. And of course, if you ask any of your children what killed the dinosaurs, they will say a big rock fell from the sky, a meteor, and killed the dinosaurs because meteorites bring iridium and other rare elements that we don't normally find. The fly in this ointment is that at about the same time, 66 million years ago, the Earth was experiencing one of the largest sequences of volcanic eruptions in the Phanerozoic era, the last 600 million years. And that's called the Deccan Traps in western and northwestern India. Now when I say large, let me be specific. The amount of lava that was erupted over a period of approximately a million years was enough to cover the state of California two kilometers deep or the lower 48 states of the United States about 200 meters deep. And you can imagine that along with that lava, as illustrated by these photos from this unpronounceable volcano in Iceland recently, comes a lot of effusion of noxious gases into the atmosphere, particularly carbon dioxide and sulfur, which gets aerosolized into sulfates in the atmosphere and does all sorts of nasty things to the environment. So the problem that has been with us for so long is what, of course, really killed the dinosaurs. Now the most accepted theory is the impact theory. There are a lot of reasons for that, which I will not go into. But there has been a contingent, a very vocal contingent, contending that volcanism is the primary culprit. Now I want to put a disclaimer out right now for this full-packed audience. And you can all go home if you're disappointed. I don't know what really killed the dinosaurs. And the fact that I can say that and be very best friends with Walter Alvarez at Berkeley should tell you something, which is this is a very, very rich story that has a long way to go. And it's moving very, very fast right now. And what you're going to see is a work in progress that we are very excited about. In fact, I'm going to tell you up front that there are a couple of-- especially one extremely technical slide that I'm going to put in front of you. And the reason for that is that it's just now being published, and there has been no time to develop the Reader's Digest condensed artist's conception version of some of these materials. And I'll do my best to explain them you in a fashion that doesn't put you to sleep. All of these co-authors, including Walter Alvarez, are very distinguished scientists in their own realm-- Steve Self being perhaps most distinguished volcanologists I know, Paul Renne, perhaps the best geochronologist on the planet, and Jan Smit, who is really the co-discoverer of the Cretaceous-Tertiary Impact published in the same year. It should really be known as the Alvarez-Smit hypothesis properly. And Sally Gibson, who is a distinguished geochemist at Cambridge. Every one of these co-authors are people who have tried to prove the outrageous idea I'm going to tell you about wrong. And in every case, the evidence that they brought to bear, which was considerable, ended up either being consistent with or supporting the hypothesis that I'm going to suggest to you. So when you're a scientist, the more outrageous your idea and the more outrageous your hypothesis, the harder you are obligated to try to show that it's wrong. This is something that separates science, perhaps, from religion that people don't understand so well some time is that in science, when we have an idea, or a doctrine, or a theory, or something that becomes accepted, the best thing you can do in science, often, is show that the other fellow is wrong. And the best thing you can do when you have a new idea is to-- that you should do-- is try to find everything that's wrong with it. We are trained skeptics. That doesn't mean we don't have some optimism in our bones. But this is really a story, in many ways-- although I'm going to flash through an enormous amount of territory technically without explaining in detail. An awful lot of this story is trying to find evidence that either supports or, better yet, refutes an outrageous idea. So then, I'm going to show you the two slides of this entire presentation where I'm actually an expert on the subject. And this is the first one. In 1989, the year I came to Berkeley and met Walter Alvarez, I published this paper, which is my most cited paper, as a theory for how, as Frances mentioned, large igneous provinces in hotspot tracks. So like Deccan Traps is a large igneous province. Hawaiian-Emperor chain is a hotspot track. I published a theory about how that works. And these are the older observations, but this is from the original paper. And germane to this study is that here we have the Kerguelen hotspot, which is currently active. And there's a track going back in time along this trail of volcanoes that get older and older-- excuse me, I'm on the wrong track. There's the Reunion hotspot track. If we go back along this trail, along this trail, back to 66 million years ago, when the Indian subcontinent was actually sitting above this hotspot, and you had this enormous outburst of lava called the Deccan Traps that covers this area. There are many other such pairs, the Galapagos with the Caribbean Plateau, the Yellowstone hotspot and the Columbia River Basalts of the northwestern United States, et cetera. And the explanation for that that is still widely accepted but might still be wrong is that these result from hot thermal plumes in the mantle in this following fashion. This is literally a shadowgraph shining a projector light through the side of a vat of corn syrup that's being heated from below. And corn syrup has a strongly temperatured event at viscosity. It gets more runny when you heat it up. And the form of upwellings, plumes, like in a bowl of soup, you get from that are these large, bulbous heads that are followed by these conduits that feed them from this hot boundary layer down here where you're heating. And so the theory is that the heads form these big splotches of large igneous provinces that come out very fast. And the tails form these nice hotspot tracks like here and along the Kerguelen Ninetyeast Ridge and the Tristan de Cunha hotspot track in Hawaii, which isn't on here for reasons I won't go into. Now here's where it gets really thick. Get ready to suspend disbelief. This is a stock diagram showing what are called the big five mass extinctions of the Phanerozoic, not including the one that we're inducing right now, in the Holocene. These are, for example, here, at the Cretaceous-Tertiary boundary, about 50% of the [INAUDIBLE] families disappeared as indicated by this little thing diminishing. There is an extension at the Triassic-Jurassic boundary about 200 million years ago. The largest extinction of all is the Permian-Triassic extinction at 250. And then the end Ordovician and the Devonian, which I won't talk about today. Here's the problem. Of the three mass extinctions that have occurred since 250 million years ago, they are all very precisely associated with large flood basalt volcanism, large igneous provinces, like the Deccan Traps for K-T boundary, the Central Atlantic magmatic province at the Triassic-Jurassic boundary, and the Siberian Traps at the Permo-Triasic boundary. So they outlier is the Chicxulub Crater or the K-T impact at K-T time. There is no evidence that is widely subscribed to for impact at any of these other boundaries. It gets worse. The Permian-Triassic boundary was recently shown, understood to be double boundary with a peak here at the end of the Middle Permian or Guadalupian period, and then the bigger peak at the end of the Permian. And the Emeishan Traps eruptions in China were 260 million years ago at this little peak. And then the Siberian Traps are here. So actually, the last four mass extinction events for which we have the best record are all associated with massive volcanism on the planet. And only this guy, the Cretaceous-Tertiary boundary, is associated with mass extinction. You can kind of see Charles Lyell rolling in his grave here because this is really confounding. How could this possibly happen? You could not make up a greater geological quandary than this. And it's basically been with us for 3 and 1/2 decades. And a lot of people had just plain given up. Well, we didn't give up. I've had lunch almost weekly with Walter Alvarez for 25 years. And whatever else we're discussing, it always comes back around to this. God dammit, we're going to have to figure out what happened sometime before we retire because it's just too good a problem to let go. Well, let's see. This isn't advance. OK. In 1991, Alan Hildebrand was doing his PhD thesis and rediscovered what Panex geologists in Mexico had known about since the '50s, which is there is a large impact structure here spanning the north coast of Yucatan, which has subsequently become known as the Chicxulub Crater. It's about 200 kilometers in diameter. And it was known to be at about Cretaceous-Tertiary time, about 65 million years ago, but not exactly. And there are all sorts of tsunami deposits around the Gulf of Mexico that actually led Alan Hildebrand to start looking for this to begin with. In 1991, Walter Alvarez, here, who had since become good friends with Jan Smit, the co-discoverer of the evidence for the impact, visited a locality near Mimbral, Mexico and found an enormous tsunami deposit that was right at the Cretaceous-Tertiary boundary in this exposed section. And that really was convincing evidence enough for many people, and subsequently, very high precision radiometric dating using the argon-argon method has actually shown that the ejecta blanket around the globe from the Cretaceous-Tertiary impact, from the Chicxulub impact, matches within 30,000 years precision the actual biostratagraphically defined extension boundary, which is really an astonishing accomplishment. And so Walter published this book which if you haven't read it, you should. Because it's one of the finest pieces of science writing you will ever find-- T. Rex and the Crater of Doom. And one might have thought that this would be the end of the story in 1991. But for the reasons I've just shown you, obviously, it's really not. And unlike things such as the question of global warming, this really constitutes a genuine scientific controversy even today. This is a quote from paper published in 2010 by Schulte and 39 co-authors. The end of this says, "lead us to conclude that the Chicxulub impact triggered the mass extinction." And this is a quote from Keller et al. in the same issue of Science with five co-authors saying, "Deccan database indicates a long-term, multi-causal scenario and is inconsistent with the model proposed by Schulte et al." Now in science, we don't count bodies or witnesses because we don't vote on whether things are right or not. We go out and try to figure out what actually happens. And in this case, in what I'm going to offer you today is a possible reconciliation of these views that two years ago I would have laughed at if somebody had suggested it to me. So flash forward into how I got back into this business. How I got back into this business was I took my family on a diving excursion to Cozumel, Mexico. And in that process-- this is my son here having a good time. I'm sure he would be really unhappy if knew I was showing you this slide, but I'm 3,000 miles away. I can get away with that. We took a bus excursion to Chichen Itza, the great Mayan civilization, great city in Yucatan. And they built these magnificent pyramids. This is actually one of the smaller ones. This is one of the bigger ones. And the reason they were able to construct this civilization where there is marvelous limestone sinkholes-- the impact, by the way, hit limestone, which is another part of this story. But these limestones sinkholes-- these are steps here. So you get an idea of the scale. These gorgeous, freshwater sources allowed people to establish civilization in this area. Now here's an interesting map. It's a little blurry here, but the white dots-- this is Chichen Itza here. This is the north coast of Yucatan. And the white dots here are the locations of these cenotes These sinkholes are called cenotes. And Walter had showed me this diagram just before I left on vacation because, you know, I was going to Yucatan. I had to go talk to Walter about Yucatan. And he pointed out that these cenotes are excluded by this ring here, which is the edge of the gravitational anomaly of the Chicxulub impact in Yucatan. So I was impressed with the fact that the shattered rock in here, for the most part, can't support these sinkholes. And I came back to my hotel room. And I had about as close to what you would call an epiphany as you get in this business. And so I want to tell you I literally sat up bolt straight in my bed at 3:00 in the morning, got out my computer while my family was asleep, and started searching the literature for something that I will show you in a minute. So the previous summer before, I had gone on a rafting trip with Leif Karlstrom, who was a graduate student of mine and Michael Manga's at UC Berkeley. Leif happens to be one of the finest bluegrass fiddle players on the West Coast and a brilliant graduate student. And he'd be telling me about something like that he and Michael Manga and others have been doing a lot of work on the observed triggering of volcanoes by earthquakes, which is a widespread phenomenon and very well documented. And here's the diagram. Here's the first of one those technical diagrams, which I will try to take the time to explain to you here. This is a result of that work. On the bottom axis here, is earthquake magnitude, so this is a modest sized earthquake, a 5 or a 6. These are very large earthquakes, magnitude 8 and 9. And on this diagram is a log plot, so we have 10 kilometers, 100 kilometers, 1,000 kilometers, 10,000 kilometers distance away from earthquakes. And these blue and red dots represent triggered eruptions, volcanic eruptions, by earthquakes. So the very largest earthquakes, for example, magnitude 9 earthquakes like the one in Tohoku, Japan, are known to trigger volcanoes for distances up to a number of 100 of kilometers, 700 or 800 kilometers possibly. But actually, the distance is a little bit smaller for reasons I won't go into. But there's an energy threshold. And these contours gives an energy threshold for the triggering of volcanoes by earthquakes as a function of magnitude and epicentral distance. And these contours-- this is 0.1 joules-- that's a measure of energy-- per cubic meter. So this is an energy density of 0.112 and 100. And the basic threshold is something like 0.1 to 1 joules per cubic meter required for earthquakes to cause incipient volcanoes, that is volcanoes that might be sitting around gaining enough umph to erupt to actually go off. And this is now a well-documented phenomenon. Chicxulub, the Chicxulub impact was approximately a magnitude 11 earthquake, which is kind of like having a magnitude 9 earthquake everywhere on the planet, if you can get the idea, which is an event unprecedented, of course, in human history and may be unprecedented other than this for the entire Phanerozoic era. Because this is the largest-- the Chicxulub impact is the largest impact we have in the last billion years of Earth's history that we know about. If you scale this plot-- and I admit that that is a stretch because the physics here is unknown. If you scale the Chicxulub impact at magnitude 11, you find that the energy densities are sufficient to trigger volcanoes out to 10,000 kilometers, which means effectively globally, that the entire volcanic system, including the mid-ocean ridges, arc volcanoes, and the Deccan Traps were susceptible to being thrown into an unusual phase of eruption by the Chicxulub impact. Now I told you something that maybe you've forgotten. But the Deccan Traps volcanism was already underway for at least a million years before the impact occurred. We know that very well from the geochronology. So the impact could not have caused the volcanism as a fundamental cause, but I'm going to get to it in a minute a model for how this might have happened for a volcanic system that was already under way. Here's another key piece of information that imminent volcanologist Stephen Self brought to my attention a little over a year ago. And he pointed out that there's the Deccan Traps and there's the Deccan Traps. This is a map of the Deccan Traps volcanic area and in northwestern India. This is the city of Mumbai here. For most of the history of the Deccan Traps, the lava flows were restricted to an area about this size. This is one of the lava flow formations called the Thakurvadi, which actually didn't constitute a major flood basalt event at all. And then within this sequence of eruptions and, it turns out, probably very suddenly, there were these enormous eruptions that I will refer to as the Ambenali and Mahabaleshwar flows, or alternatively as the Wai subgroup flows, that covered an area about the size of France or Texas in probably only 100 or 200,000 years. Some of those flows were so large that they went down the Krishna paleovalley, and flowed into the Bay of Bengal on the east of India, flowing at least 1,000 kilometers and maybe 1,500 kilometers from their source. If you can imagine single lava flows 10,000 cubic kilometers in volume and flowing from here to wherever 1,000 kilometers from here-- I don't know, Chicago or whatever. That's an impressive current. Now I'm going to show you another technical slide, but this one is comprehensible, so stay with me. This is called a stratigraphic section in geology, where we put time on the bottom. Time goes from bottom to top. This is old down here and young up here. And the Deccan basalt group of formations, which is about 3 and 1/2 kilometers thick in aggregate of lava flows, is divided up into the Kalsubai, Lonavala, and the Wai subgroups, which are subdivided into these formations, which I will not pronounce to you or it would take most of the rest of my lecture. Now these lava flows are indicated by their thickness in situ, where they're mapped stratigraphically. So this amount of time here represents approximately a million years. But it turns out that those immense lava flows that I was talking about, these huge flows, were restricted to just mainly these two formations and a little bit of this, so really, Poladpur, Ambenali, and Mahabaleshwar with the big act being Ambenali. And these are called the Wai subgroup flows. And the volumes-- the shading is a little faded by the lighting here-- but if you scale the volume of flows by the area times the thickness of the formations, everything that happened up through the Bushe Formation in the Lonavala subgroup was relatively small in volume. And the total volume here is only a couple hundred thousand cubic kilometers, which is small compared to large flood basalt events. 70% of the volume of these eruptions occurred within approximately these three formations. Now here's the kicker. A group of paleontologists, including [INAUDIBLE] and others, showed in 2011 that if you go and look at drill cores-- you people can't see this. Over here, in the Bay of Bengal, where these lava flows cap Cretaceous sediments, that those lava flows come in exactly at the Cretaceous-Tertiary boundary. So we've gone from a small problem to a huge problem. The small problem is-- well, number one, why does the K-T boundary have an impact and a flood basalt? Given that the flood basalts typically-- they're main phase of eruptions only last 1 to 2 million years, what are the chances that the largest Phanerozoic crater we have happened during that period of time? But now it's a lot worse because the amount of time separating the K-T boundary and the onset of these three huge lava sequences is probably less than 100,000 years. And the chances of that happening at random, geologically speaking, are about 0, or actually, 1 in 100. But, you know, we can round off to 0. And we know that the Chicxulub impact did not cause the volcanism because it was well underway here as a general event long before the impact. And of course we know that the volcanism didn't cause the impact. That would be really quite unthinkable. But the hypothesis is as follows-- this is an image of one of those big, hot blobs of corn syrup but scaled to the thermal convection in the Earth's mantle of hot stuff from the core mantle boundary or the lower mantle region coming up and making a big pancake underneath the earth's crust and lithosphere. And this stuff cooks a while. It forms into veins and channels. It sits at the crust mantle boundary here we call Moho and drops out some heavy minerals as it cools. And then it erupts the Deccan Traps or a large igneous province. This is the canonical-- or this is the paradigm for how large igneous provinces form according to this plume head and tail theory. Now imagine this. You've got this region here. It's about 50 to 100 kilometers thick and about 1,000 kilometers wide. It is saturated with partial melt, that is if you add a small amount of partial melt or you make that melt become a little bit more interconnected, it's going to come squirting out. And that's basically what causes periodic eruptions. You hit the Earth with a magnitude 11 or larger earthquake. You pass Rayleigh waves, or surface waves, through this region that have meters of displacement, and you get some kind of physical effect probably that may be something like liquefaction of sand when you have earthquakes, which we do, in the San Francisco Bay Area. It's not difficult to imagine that perturbing a system this massive with this large a disturbance might result in something anomalous. And so our idea is that these extraordinary lava flows within the Deccan Traps were triggered by this event. Now to give you a good analogy for how to think about this, this isn't that unusual in nature. This is a photograph from the San Francisco earthquake in 1906. And anybody who knows much about the San Francisco earthquake knows that it was only about a 7.8 magnitude earthquake. It wasn't that huge. It was very destructive. It destroyed most of the city of San Francisco. But the main reason it destroyed the city of San Francisco was because of the fires and not from the strong ground motion. So that's what I would call a major secondary effect. And of course, for the Deccan Traps, the secondary affect-- the question you want to ask then, was it the impact and the fallout from the impact that killed the dinosaurs or was it the volcanism that caused the extinction boundary? Because we know that volcanism is associated with the other three major extinction boundaries since 250 million years ago. Now I'm going to warn you before I put up the next slide. This is a horrendously complicated slide. It represents a huge amount of geological information all at once, and I don't want you to get dizzy when you see it. But the reason I'm going to show it to you is that I'm going to come back to it at the end of the talk with some new data that will be much more comprehensible. But I want to give you a flavor for what it is that we've been up to. There are people in this room that know more about some of this than I do, so I will excuse myself. But here again, we see this stack of Deccan formations. You'll recognize these names. OK. And there are three different data sets represented here. And I'm going to tell you what they say without explaining them to you. This data set here says that at the Bushe-Poladpur boundary or the onset of the Wai subgroup, the chemistry as reflected in isotopes of neodymium and strontium tells you that this stuff is coming right out of the mantle with no time to interact with the crust. These diagrams over here on the right show that at the time these eruptions were initiated, the stress system governing the fractures in the crust that were feeding these eruptions became randomly oriented, which means that they probably resulted from huge magma overpressure in the crust rather than regional processes of extension that may have been going on when these formations were happening down here. All of this indicates that effectively someone turned on the hose from the mantle very, very quickly with this partial melt flood literally coming out from an ongoing set of eruptions that had otherwise been percolating along at a rather modest rate. In the middle column-- I'm going to ask you to strain your eyes-- what we have here is time on this axis up here from 64, say to 70 million years. This dash line falls along the K-T boundary. We know that's at 66 million years. And these are the age ranges as determined by previously published radiometric ages on these formations. The thing you should take away from this is this is horrible. These ages had huge error bars, and they're worthless. They tell you almost nothing with precision of what we need to know about what happened in this lava stack relative to the K-T boundary. Because we know, in fact, all these lavas came out within a million years of this boundary, certainly between 65 and 67. And these are all over the map. So the essential scientific problem and challenge of trying to nail down what actually happened at the K-T boundary, what killed the dinosaurs, what's the relationship between the impact and the volcanism comes down to primarily uranium-lead and argon-argon geochronology with mass spectrometers. So time for a road trip. The sampling of the Deccan Traps have been very haphazard in the past. And so this is Mumbai Harbor. My colleagues, Steve Self, Paul Renne, geochronologist, myself, and Paul's student, Courtney Sprain took a trip to India last March. And we had a lot of fun. Our first visit was to-- anybody ever been to Elephanta Island in Mumbai Harbor? It's a national park. And the reason it's a national park is that the medieval Hindus carved these lovely stone temples into the lava flows of the Deccan Traps. And here is Steve Self, the volcanologist. This tourist is looking at Vishnu over here. Steve is looking at the lava flow crust tops. These are actually lava flow tops. And there's a lava flow top right here that goes through Vishnu's mouth and comes right through here. And you can see the structure of these individual lava flows just beautifully. Now that wasn't the main show. The main show was the Deccan Traps, the large sequences that you see, that dominate this entire landscape in northeastern India. They're called the Western Ghats. Ghats is a Hindu term referring to steps leading up to the Hindu temples of medieval times. And you can see that these lava flows-- these are individual lava flows that are, of course, colossal in extent and thickness. They lead to a step-like topography. Has anybody ever traveled in this region of India? Now you know what it was that you were looking at. Yeah. So it's very entertaining to travel on these extremely steep mountain roads with every conceivable kind of vehicle coming at you. This is a very hazardous form of field geology. And I just never cease to be amazed at the ingenuity of Indians and the various vehicles that they manage to navigate the various roads with. And we always had the friendly combination temple and chai stop to keep us awake during our field travels. These are the famous strawberry arrangements of the town of Mahabaleshwar for which the Mahabaleshwar Formation is named. And it's just a lovely, lovely part of India. And we had a wonderful time. These are the langur monkeys with the Deccan Traps behind. And perhaps these monkeys are saying, you know, if it weren't for these rocks behind me, you might be looking down the throat of a T. rex at this point. But getting down to business, we were there to look at the rocks and sample them. This is about 20 meters tall, I believe. And this is a single lava flow. Some of these lava flows reach thicknesses of 50 to 70 meters. This is in the Ambenali Formation, these huge sheets of lava that spread across the entire Indian subcontinent. And this is called sigmoidal joining. This forms as fractures after the lava is cooling. And of course, what we were really there for was to sample the rocks. This is Paul Renne, geochronologist swinging his very large rock hammer at these poor rocks and trying to get samples that were specifically tailored in nature to doing high precision geochronology. And this is Paul looking at another formation that turned out to be really run. And this is another shaggy dog story coming. This is called a red bull. It's a term that the Indians use to describe soil horizons, that is weathering horizons between lava flows. So this is an upper lava flow. And down here below, which you can't see, is another lava flow. So the time interval between these massive lava flows is probably substantial, perhaps of order 5,000 to 10,000 years separating each individual lava flow. And these weathering horizons are very interesting for a variety of reasons because they do give you some sense of time elapsed. So this is [INAUDIBLE], one of the people who went in the field with us as well, holding a much less macho hammer here next to one of these red bulls. It wasn't actually red. But this is still a weathering horizon between two major lava flows. Now this particular contact happens to be at the boundary that is the lower boundary of the Wai subgroup lava flows, the lava flows that come out and cover a good bit of India. Or it's called the boundary between the Bushe Formation and the Poladpur Formation. And I looked out across the horizon the other way, literally from where I was standing taking that picture, and I noticed something interesting. And that is that the entire landscape from this point is dominated by these horizontal bench structures, or terraces, that correspond to the same topographical level for as far as the eye can see. And this is what it looks like. Oh, this picture's a little fuzzy. I'm going to show you a better image in a minute. This is standing up on top of the Ambenali flows, the largest lava flows known on the planet, looking down at these terraces. There's one terrace horizon, another terrace horizon, and one back here that you can hardly see because the picture's little blurry. This is a very clear day in this part of India unfortunately. It turns out that this is a bit of a smoking gun. It's a bad joke, actually. But it turns out that these terraces are very, very interesting, more interesting than I would ever imagine. I came back to Berkeley after this field trip, and I showed these photos to Walter Alvarez saying, Walter, what the heck is this? He said, I don't have any idea. Let's get on Google Earth and have a look. Google Earth turns out to be, of course, a real research tool these days. And this is a view using satellite images and the tilting up perspective capability of Google Earth that Walter showed me. And this is effectively standing pretty much in the same spot. And again, you see these terraces. These are agricultural terraces built on the topographic terraces going as far as the eye can see, dominating this part of the landscape. About two days later-- and I was literally out in the backyard throwing footballs with my son, which is kind of a sacred thing to do for those of you who've done that. And Walter calls me up. Actually, my wife comes out and says, Mark, Walter Alvarez is on the phone. And I said, well, can I call him back later? I'm playing football with Noah. And Sarah goes back in. And then she comes back out and says, no, Walter says he needs to talk to you now. And so I go, and I talk to Walter. And Walter says, Mark, there's something I need to show you. And I said, well, can I come over tomorrow? He said, no, you need to come over right now. Now when you're working on problems that involve the K-T boundary and Walter Alvarez calls you up on a Sunday afternoon and says, you need to come over right now, you need to come over right now. And this is what he had to show me. He was looking at aerial photos again through Google Earth to the north several hundred kilometers from the site that I just showed you, seeing the same terraces dominating the landscape. And this has never been published or noted in the literature that we're aware of. And it's kind of amazing. But he said, Mark, look carefully at this. Look and see what you see. How good a geologist are you really? I said, Walter, I'm a geophysicist. You know? Give me a break. He said, look at these fractures. Now these are fractures that occur below the terrace level, and they're pervasive throughout the landscape until you hit this terrace level here. They stop. Now this is the next image he showed me. He said, look, on this high-res image from Bing Images, you can really see these fractures. And these terraces from above that are labeled PB-- that's for Poladpur-Bushe boundary-- do not have these fractures. The fractures stop stratigraphically at that level. And then he said, look at this fault over here. Here's a close up of that fault. That fault runs right up through all these other fractures, through these lower formations, and it stops dead at the Bushe-Poladpur boundary. Now what does this mean to a geologist? It means to geologist that there's missing time. Because stratigraphically, what had to have happened is that these fractures and faults had to have taken place and then the lava flows came in on top of them. Something really significant happened here. Well, that just adds to the mix. We know that the stress system changed at this point. We know that the geochemistry changed at this point. We know that the volume of eruption, or the rate of effusion, increased by, perhaps, an order of magnitude. And now we know that there was also a break in time, which suggests that the Deccan Traps were kind of yawning and saying, well, maybe we're done, and then something happened. We still don't know exactly what this means. But it means that there is a major break in the character of what happened in the Deccan Traps at this point. So Walter's wife, Millie, snapped this photo of me, which I will always treasure. And I have to say that it made me very happy that 25 years ago, when I came to Berkeley, I very decidedly did not take sides in this volcanic versus impact debate. Because Walter and I have remained great friends through all of this as a result. But back to the real business. These are the sample bags that Paul Renne and I were bringing back from the Deccan Traps and Kanchan Pande's lab at the Indian Institute of Technology. Now if you've ever tried to ship 600 pounds of rock from Mumbai to the United States in the post 9/11 era, it took a lot of creativity on Kanchan Pande's part to figure out how we were going to justify this. But we did get the rocks back. And the purpose of these rocks is to do very high resolution mass spectrometry geochronology looking at the potassium-argon decay scheme, which can give such high precision. And now, I'm going to show you the initial results of that work, which literally are only two weeks old. First of all, this is back to this horrendous figure that I promised you I was going to explain. And now I'm going to pare it down to size so your eyes are hopefully used to this stratigraphic column. Time starts here. It goes up this way. This is about a million years elapsed time. The big event starts here at the bottom of the Wai subgroup flows, the Poladpur, the Ambenali, and the Mahabaleshwar. Blair Schoene, and Gerta Keller, and Sam Bowring at MIT and Princeton just published a paper in Science with these two dots that are actually a little bit larger than the actual error bars of precision on the measurements of uranium-thorium dates on minerals called zircons that are in the red bulls and the soils between these lava flows for the Mahabaleshwar and the Ambenali formation. These are just discernibly older than the K-T boundary. Just after. Paul Renne's recent dates-- now, this green dot down here is probably problematic. And I'm not going to discuss that, but it's not to trusted at this point. And neither is this red one here that Paul Renne just produced. But these are very good ages that Paul Renne's lab just produced in the lower formations because these are the first two dates that have come out. Now the hypothesis that we've been suggesting is that this boundary between the Poladpur and the Bushe, that terrace-forming boundary where all hell breaks loose in the geochemistry, and the tectonic signal, and the amount of lava should fall right, of course, at the K-T boundary at 66 million years. And this is something I would never do with this scant of data in a published work, but since this is a public forum, you're tempted to draw a line through these. And if you draw a line through those data, it looks pretty darn good. And we just found this out a couple weeks ago. So what have we learned? I would call this a hypothesis of reconciliation. And again, I want to repeat what I said at the beginning. Everything I've told you may turn out to be wrong. And that would be OK with me because the way that science is supposed to work is that when you go out to test these ideas with a higher precision lens, you find out what really happened. And so far, since that hasn't been done, we're going to understand this a lot better no matter what the actual causative relationships might be. And that's a beautiful thing. But it's a hypotheses of reconciliation with no alternative. There is-- I know of no other suggested alternative explanation for why the very largest eruptions within the Deccan Traps by a long shot, the K-T extinction boundary, and the Chicxulub impact should all have occurred within 100,000 years of each other. So I'm going to go with it for now. If this coincidence is difficult to dismiss, if seismic waves from Chicxulub impact-- it is likely that they were large enough to trigger volcanism worldwide. And that leads to all sorts of other things we can go look at. For example, the mid-ocean ridge system should also have been triggered into extraordinary activity at the time. The signal from that may be difficult to detect, but I never cease to be amazed in this journey so far. And we haven't tried that yet. If, of course, the Chicxulub impact triggered the largest Deccan Traps eruptions, then clearly, the Deccan Traps may have contributed to the extinction itself in addition to the impact. Now my paleontologist friends tell me that no known taxa in the paleontological record disappeared after the iridium anomaly. So I'm very skeptical still. I remain very skeptical-- I want to be clear about this-- of the idea that anything other than the impact was the main causative agent of the K-T extinctions. But I'm a geophysicist, and if a magnitude 11 impact triggered the largest volcanic eruptions that we know of in the last 250 million years, that's a pretty cool geophysical event. And that's going to lead to a lot of work no matter what its relation might be to the extinction boundary. And the punchline here is that the best kind of hypothesis or theory is one that you can test. It turns out that this hypothesis is one that's probably a lot easier to prove wrong than it is to prove right. If we found another smoking gun like an extraordinary pulse of activity at the mid-ocean ridges or arc volcanoes or something else that we haven't thought of yet, fine. But we haven't thought of that yet. But if, for example, we showed that, in fact, this major phase of volcanism actually started before the impact, then that would through a lot of water on the idea. So far, the geochronology looks like it actually isn't perfectly in line with what we have predicted. So it's a little epilogue here just to kind of drive the point home at just how much fun you can have when you're a scientist. On our last day at the Indian Institute of Technology in Mumbai, Kanchan Pande went to his file cabinet and dragged out this manuscript that kind of had the appearance of finding the Dead Sea Scrolls although it was only published in 1986. And it's a well-known paper by a group from Beane et al. who did the chemical stratigraphy of the Deccan Traps. And he pointed out to me a simple sentence that you will not understand, but which I will translate for you. In this paper, it says, "On the southern end of the Western Ghats south of Mahad--" which is where we were working in Mahabaleshwar-- "analyzed samples of drill core from [INAUDIBLE]." The reference is an unpublished PhD thesis of John Mahoney in 1984. "Placed the base of the Bushe Formation at 450 meters below sea level." What this means is that there are drill cores that go through the Bushe-Poladpur boundary that are probably sitting around in warehouses in India. And in fact, we are in contact with one of the hydrologists who works with these cores. He's a volcanologist turned hydrologist at the University of Pune. A year ago, if somebody had told me that we might know within a meter or two where are the actual K-T impact iridium anomaly might be within the 3,500-meter stack of the Deccan Traps I would have laughed at that idea. But now, at that terrace level, within a few meters plus or minus, is our best bet for where that might be found. And so we are going to soon be on a hunt for the archives of the hydrological agencies in this region of India to try to find intact core samples that can be run for analysis and see if that would get me another smoking gun or at least a very strong indicator of exactly where the impact occurred in this massive stack of lavas. So this is a fun field to be involved in. If I have left you with the impression that we know the answer to this question, I have failed utterly. But I hope I've left you with the impression that this is a grand adventure, that a great deal of humility is called for, but that in next few years, we're going to make a lot of progress. Thank you very much. [APPLAUSE] So I think I'm allowed to answer questions. Yes, sir? AUDIENCE: [INAUDIBLE] Between five and three years ago, I started teaching a course and looked into this question of what killed off the dinosaurs. It occurred to me that the Deccan Traps are almost antipodal-- MARK RICHARDS: They're not. AUDIENCE: --to Chicxulub. Not quite, but close, and tracing back 65 million years ago wasn't exact either. But it occurred to me, still, that the seismic energy created by that impact would have gone around the globe and concentrated near the Deccan Traps and perhaps set off the Deccan Traps at that point. MARK RICHARDS: Well, that theory was actually published very shortly after the evidence-- AUDIENCE: I didn't finish my question. MARK RICHARDS: Oh, excuse me. AUDIENCE: Sorry to be so long-winded. MARK RICHARDS: OK. AUDIENCE: But anyway, I was intrigued by this, so I went to some geophysicist friends. And they said, oh, that was thought of long ago and it was worked out mathematically. There's no question that that couldn't have worked. So I was wondering-- because I didn't follow up-- who wrote that paper? And what holes there might be. MARK RICHARDS: Yeah, yeah. Well, no. It's a very good question. And it helps me clarify some of the subtlety of this because there is a history to this problem. Shortly after the evidence for the impact was discovered in 1980, it was proposed by a number of people that perhaps the antipodal-- when you have a seismic wave, an earthquake on one side of the planet, the waves travel around the planet, the surface waves and the body waves, and they coalesce at the antipode 180 degrees away. And you get a maximum amplitude in the far field from that. So because we knew that the Deccan Traps were going off roughly at the same time of the impact, people said, oh, Deccan Traps. Maybe they were antipodal to the impact site. Because we didn't know where the impact site was. When the impact site was discovered, it was discovered 130 degrees away from the reconstruction for the Deccan, which is well within the error bars. It is clear that it was not at the antipode. So that idea went away. And furthermore, the people who do cratering studies and looked at the energy said, that is not nearly enough energy even if you did have antipodal focusing to melt the mantle to produce that amount of lava. The difference between our hypothesis is therefore two-fold. One is you don't have to be at the antipode. And secondly, you don't need that much energy. [INTERPOSING VOICES] Because all you have to do is trigger the melt that's already there. And to change, effectively, the permeability of the system not that much to create instabilities and channelization of melt to get this outpouring. And that's what-- it's odd to me. Other people may have thought of this idea, but it hasn't been published before. When I came up with this thought, the first thing I did is say, OK, somebody else has done this. I scramble, look through the literature. But this seems to escaped people as a causative mechanism. So it's a very, very good question. Yeah? AUDIENCE: As a result of all this, what made the dinosaurs [INAUDIBLE] starvation or darkness around the Earth with no sun? MARK RICHARDS: Well. This is where I get repeat again that I'm not a paleontologist and there's a lot of controversy about this. Unfortunately, the impact scenario and volcanism scenario do some of the same things. The Chicxulub impact hit a carbonate shelf, which releases a lot of carbon dioxide and sulfur into the atmosphere. The killer agent with the Siberian Traps is thought to be because that mantle plume volcanism came through a carbonate and evaporite stack of 1,000 meters or so of sediments that were very carbonate and sulfide laden. And so launching sulfur into the atmosphere create sulfate aerosols, the acid rain, global cooling event. But it gets out of the atmosphere pretty quickly even if you get it into the stratosphere, which is hard enough. Carbon dioxide, on the other hand, has several hundred year cycle, maybe 1,000 year cycle, for getting it out. And so you could see a one-two punch with global cooling followed by a longer period of global warming. There's has been endless speculation on the-- people obviously pay a lot of attention to the pattern of species extinctions. For example, those foraminifera that I showed you in that original rock from Walter Alvarez, those are planktic foraminifera. They live in the shallow water. The benthic foraminifera that lived at the base of the ocean hardly noticed the K-T boundary. So there is a lot of work on this. And I could try to repeat some of the arguments I've heard. And I would simply be creating the opportunity to lead you astray. There's a vast literature, and it's unsolved. It's unresolved problem. Yeah? AUDIENCE: In the stratigraphic diagram you showed, the line [INAUDIBLE] points, is the interesting part that it's almost vertical? MARK RICHARDS: No, the interesting part is it goes through that blue star, which is where we would predict-- AUDIENCE: I'm sorry. I don't understand the point to it. MARK RICHARDS: Yeah. OK. Good. I'm sure that other people didn't understand that point as well. The hypothesis is that at the time that this Wai subgroup, this outburst of lava, came as a result of the K-T impact at 66 million years. So the intersection of this contact boundary and K-T time is this blue star. So the prediction is that if you were to get-- when we get the dates for all these other formations, that they will presumably fall on a line roughly through this point, so that this stuff is younger, and this stuff is just a little younger, and this stuff is just a little bit older than the K-T boundary. So I am almost sorry I drew that line on there. Like I said, I would never publish something like that. But I wanted, basically, to try to illustrate what, obviously, I didn't explain very well. So thank you for asking the question. Yeah? AUDIENCE: I have two examples of earthquakes being caused by perhaps non-antipodal explosive events. One recently occuring [INAUDIBLE], which they say was a result of fracking, [INAUDIBLE]. That in the fracking process, they do these explosions to [INAUDIBLE] correlation between [INAUDIBLE] picking up several hundreds miles away, not [INAUDIBLE]. And then the other [INAUDIBLE] is bombing in Baghdad,. There was a relatively minor earthquake that occurred along the Appalachian Trail that started on [INAUDIBLE] and went all the way down the Appalachian Trail, about 3.2 magnitude. [INAUDIBLE] backed up and hit Canada and the [INAUDIBLE] MARK RICHARDS: OK. I will make a very strong statement, which is the energy of the bombing of Baghdad could not have induced earthquakes around the globe. There are many, many other events larger and closer that would have done it instead. So that's probably not your-- AUDIENCE: Well, to me, there's a correlation. MARK RICHARDS: Well, correlations are things we have to be very careful with in geology because we tend to find things we're looking for. Somebody else? Yeah? AUDIENCE: I read a few years ago about [INAUDIBLE] craters around the Permian-Triassic extinction. Any comment or any-- MARK RICHARDS: I don't know of any large craters at that time. AUDIENCE: OK. I think there was information published in Science. [INAUDIBLE] but they were dated to roughly that time period. MARK RICHARDS: There have been some published suggestions of what are called microtektites in the Permian-Triassic section. They have been hotly contested. I don't know of any large craters, certainly not anything like as large as Chicxulub associated with Permian-Triassic time. Yes, sir? AUDIENCE: Can you say a little more why you're dismissing the 2.2 [INAUDIBLE]? MARK RICHARDS: Yeah, Paul doesn't like this date very much. And I'm not sure exactly why, but I tend to trust him. This date is on what is called a melt segregation feature. And it's hard to understand how it also contains zircons that have Precambrian ages in them as well. It's hard to understand how you drag Precambrian zircons up in a large lava flow without dissolving them and preserve them. And so this may be a sill or an intrusion that comes after the fact. And so this age maybe a little young. But I think Sam Bowring and Blair Schoene may argue with me on that point. And so I'm just a little bit-- I'm not discounting this age. It may actually be right. And if it's right, it doesn't disprove anything because there's a lot more to go here. But that's why I'm a little cautious about that one. Yeah? [INTERPOSING VOICES] Go ahead. AUDIENCE: Vincent Courtillot, I noticed, was the co-author with you in 1989. MARK RICHARDS: Yes. AUDIENCE: And I know that through a number of years now, he has been pushing the idea that the Deccan Traps were the real cause of the extinction. MARK RICHARDS: Yes. AUDIENCE: And as I recall the argument, one of the things he had to establish in order to make that a reasonable hypothesis, because he had the gases released doing the killing, that is from the Deccan Traps, that it has to be released in a short enough period of time. And the difficulty was in making estimates of the time period of those eruptions with a small enough error bar, like tens of thousands of years. MARK RICHARDS: Yes, yes. AUDIENCE: And he wasn't anywhere near that. MARK RICHARDS: That's correct. AUDIENCE: What is his status now, and how does he react to your [INAUDIBLE]. MARK RICHARDS: Well, Vincent Courtillot was a good friend. And I disagree with him vehemently about the evidence that he cites for-- you know, he's trying to suggest, for example, that these entire formations might have been erupted in a few thousand years. And the fact that there are well-developed soil horizons between individual flows pretty much makes a nonstarter. What he is correct about is that from Mahabaleshwar time to Jawhar time represents probably an order of 600,000 or 700,000 years. As you say, that's not nearly quick enough to do the environmental damage that he's after. Yes, sir? AUDIENCE: Are zircons ubiquitous in these formations, so you'll be able to gets dates [INAUDIBLE] MARK RICHARDS: I'm the wrong person to ask. You generally don't find good ones in lava flows, which is why the argon-argon method-- if you actually want to know how old the lava flow is, it's good to use argon-argon methods. The zircons from these two dots are zircons that are deposited in the soil horizons between lava flows. And there's an inference. In fact, there's a whole spectrum of ages. And you just take whatever the youngest one is, is the maximum possible age for that formation. So it's good to have both. It will be good to have both of these data sets. And we have three to five samples for each one of these formations. They're going to be a lot of red dots on this diagram. And the width of this dot right here is the two-sigma precision of the method. And so it's going to improve vastly in the next six months. And we might be right. We might be wrong. Yeah? AUDIENCE: Is there no correlation between known meteorite impact [INAUDIBLE] recorded history and increased volcanism? MARK RICHARDS: Oh, people have looked-- the question is, is there a correlation between impact crater occurrence and large-scale volcanism. And that has been looked for extensively. And there's no correlation that I know of. And that is from the people who would like to find one. AUDIENCE: So for this hypothesis to work, it would have to be something-- just because it's unusally large-- MARK RICHARDS: This is a one off event. AUDIENCE: Right. MARK RICHARDS: Yeah. AUDIENCE: So like the one in Arizona-- MARK RICHARDS: There's no evidence for impact at any other major flood basalt event that we know of. AUDIENCE: OK. PRESENTER: Well, thank you all for coming. We could ask him questions all night, but I want to take him off to get some food. [APPLAUSE]