 and we found both iron and brass. So, like the mineralogical diversity. I'm actually really curious to know if it's, what we see as an outlier instead of breaking down supplies of tin at the end, or if, like in other words, supply-driven innovation, or if there's just a lot more variants of what they're making and we don't know. I mean, like, that's soon we have a diagram of the video around it. Yeah. Can I tell you a story about that? Of course. I, about 2014, I was with a German scientist who was going to measure that. And we talked about it. He said, I think it's made out of iron. So I told him, here's the cal, I built the calibration form, we measured it, we got the nickel and the cobalt. And then I'm like, we need to, can we publish this? He's like, yeah, I'm going to do a little more work and we're going to do it. So we keep doing it, keep doing it, and that's about a year and a half. I'm like, we need to do this. Yeah. And then he says, just turn on CNN. And the team beat us. Oh, I did. I broke my heart. I'm like, that was the first time I'm like, ooh. Yeah. And what's worse is it was a competitor, the broker. But I think I told, I gave this example of media work, I am too. I'm worried that I gave away the photos. And what happened is that they didn't get the permit to measure it. They had an on-contact instrument. So they just walked by and visited without the permit. So just poor German scientists. I think they're pulling away with that. Yeah. I mean, what are they going to do with that? Well, the Egyptian government is pregnant. They might be, this guy might be sanctioned from working again. He works for a company. He isn't. Yeah. Anyway, that broke my heart. Anyway, but let me see. What was I doing? I need to open up broker instrument tools. Firstly, to do it with Windows. But yeah. Anyway, we're just going to do this by the book. What I'll do is I'm going to backtrack and discuss this principle a little more in terms of review where we're at for the purposes of the video. Yeah. Because I think this point is going to be pretty crucial. So let's go. We're discussing fundamental parameters. We're discussing fundamental parameters. So let's start. Which is actually a fundamental parameter is an equation. Yes. Okay. I mean, it's marketing, right? But anyway, let's look at here. So here's the use that's right. Here's geocentric. Which of these, again, is the best predictor for where the planets are moving? Both. Both. They're equally good, right? Equally good. Which one is an accurate representation of reality? This one. And then which one of these is generalizable? If I go to a new solar system in Andromeda, which one of these models can I use to predict those planets? This one. Right? Exactly. So to the point. The point is that prediction is not the standard for accuracy. In other words, like physical observations. And by the way, as an aside, could I quantitatively determine which of these is the best, most accurate? Or do I need qualitative data? Like observing the planets. I need qualitative data. So here's the key thing, right? We had archaeology worship quantitative a bit too much, I think, and we forget how important qualitative is. Do you know what is the difference in quantitative and qualitative analysis? Like what is the fundamental difference? Well, one you can create predictions on, and one is just the description. That's an excellent one. I'm thinking of it even more. So we're going to go this way. What is the color of this unit? Right here. What color is that? Right here. Very closely. Did you see it turn red or orange? What is the uncertainty of green? Qualitative doesn't have that. Now, if I had a colorometer, you can download it after your phone will do this. Can I measure and then change the light source and see flickers of green? Yes, that would be a quantitative measure. If we break down what Lucas is saying even more simply, it's this. The difference between qualitative and quantitative is the ability to demonstrate uncertainty. The qualitative is an absolute measure. The quantitative has uncertainty. With uncertainty, I can make a prediction. I can even put a confidence band on that prediction. With qualitative, it stays a description because without that uncertainty, I don't have additional information to build a model on. With the universe, you had people arguing this. People arguing this. It was Galileo's ability to observe rotation around Jupiter with those three moves that demonstrated the principles of the heliocentric thing. He didn't have to quantitatively model the movement of those moves to disprove this. He could demonstrate that the principle of large object in the center and everything moving around it was observed. Once it was qualitatively observed, this guy worked. Galileo paid a price for that. But in general, that is the crux of it. So, while we celebrate Galileo, there's a triumph of empiricism over authoritarianism. And I don't mean authoritarianism in the like, you know, Risen Brutawon type style governing the Turkey. I mean authoritarianism in the so and so said this, this is the law. Right? This is how it works. An example of this, I was in West Virginia. We had a geology professor come in, one of those older guys, a previous generation thing. And I'm measuring, he gives me a sample to measure. And I measure it with the tracer. And I get this, like it says this is 9% aluminum. And he goes, that, that can't be. And I said, why? He's like, this is red. Granite doesn't have that much aluminum. Like, and I'm like, well, I mean, here's a spectrum peak right here. Like, you can see that. And he's like, something's wrong with your instrument calibrations. And I said, I'm gonna be giving you a calibration. You can see the peak. And he said, I sent this to the most authoritative lab in geology. And I've sent everything. And they've never been wrong. And they said it's only 2% aluminum. And I was like, I don't know what to tell you again. Like, this is what it is. After he leaves, he's kind of angry. So my friend from Florida says, it's a piece of felt spot. Like, that happens all the time. But anyway, but the core point is, there's always argument from a point of empiricism. I can see, I can measure, whether it's a qualitative or quantitative measure, I can see how much aluminum is here. And his argument was from authority, right? This is this class of material that should not have this. That's the difference, right? What Galileo did is demonstrate empiricism. The reason I bring this up is Galileo demonstrated empiricism, not with quantitative data, but with qualitative data. So in archeology, we conflate empiricism with quantitative research. It is combination of the two. So in any case, the idea behind vibrational parameters is that using precise physics, such as here, I can model what is in that sample. So if I know the entry angle of the sample, if I know the fluorescence efficiency, the relative measurement depth, if I understand the detector chip's thickness and how that can affect its emulation, if I understand every physical variable that influences the spectrum, then I can back-calculate and estimate what percent is in here. Now, there's multiple algorithms, and by the way, just FYI, if you see multiple algorithms, there's some time to queue and things are not working out as they should, and people are trying to improve it. But that's the idea here. Is everyone's trying to cover it? I like how this guy calls it the fundamental fundamental parameter software, but that is the key. They're trying to figure out that aspect. So to see what's happening on the unit itself, let's see if we can't connect here, especially in our videos, really. Up there it is. So I'm going to use what is actually a fundamental parameter calibration of this unit, because again, this is originally an industrial unit. So you said it was a piece of felt starts because you happen to be pointing at it. Well, no, we've got to point it out. Goat spires and granite all the time. So you've got these... Yeah, so you said if you ground it, and you said a small portion of your rock off to the lab to take care of it, and that man didn't have granite, and you pulverized the whole thing, and that had no spark, that would explain it, right? Like, there's a simple reason for why he would have been wrong. Because there's a lot of force. Exactly, exactly. The key is, this reasoning was that, you know, what didn't account for, didn't allow for any empirical data to influence, right? So anyway, what I'm going to use here is I'm going to use precious metals, which is a fundamental parameter's calibration. Now, when Bricker makes one of these, it's typically a mixture of empirical and fundamental. In other words, they use real standards, but the majority, but those are designed to send disciplined fundamental parameters. So I'm going to come over here, and we're going to see what's in this gap. What is that? There we go. So this thing, it is 75% copper, 24% zinc with a little bit of lead. Does that make sense? What would this object be, if that's true? Just a brass. This is a perfectly normal alloy. It doesn't have a match in the library, because it can cross-reference, but how much did these add up to 100? Just about. This is probably an accurate result. And it's using fundamental parameters, right? It's using that geocentric model, right? So, not too bad. Where's that obsidian, by the way? Oh, I'm actually curious. This is going to be disciplined right now. Let's see. By the way, the results are here. It's just the Wi-Fi that's above. So, 25% iron, 8% zirconium, 16% ruthenium, and 15% rhodium. Is that accurate? Could this be some obsidian? So, can anyone tell me what's happening? Why does it have a fundamental parameter's algorithm? And if I was in this gap, what's up? Well, what's most... So let's think about it this way. This guy, do you guys remember what this was made of? Mostly copper. Some zinc. Now, does copper fluoresce in the spectrum? Yeah. Does zinc fluoresce in the spectrum? And does lead? So, everything I have to work with fundamental parameters shows up. What is most of this made of? Silica. Oxygen. Oxygen. Does oxygen show up in the spectrum? Yeah. So, how... It's missing the most common element. So, what does the fundamental parameter do? It tries to add up to 100% with all the stuff it can see. And it blows up. That's dangerous. Oh, yeah! It's terrible! The soils method would also be way off if you have some... I mean, so what is the thing? Broker can take soils. It can tell it to fill in the oxides. Right? You've seen this before. If you do this kind of... The soils method? I'm actually really curious. Do you want to press the... The soil? Let's try it. Let's see. I've never used this before. It still shows precious metals on top. Don't worry. It's still in my mind. It's the Wi-Fi refresh rate. I'm actually curious if this is going to do multiple days. Have you already run that? Holding that thing in there as a plan? I've got my fingers on either side of it. Not directly behind it. This is just trace elements, so... The results actually... That rebellion was probably pretty close, right? 140? Yeah, it's only 148. Yeah, yeah. But yeah, in any case... Although the soils isn't empirical. 147? 147. What's your money? I'm supposed to be 128. What's your money? I'm supposed to be 222. Yeah. Not bad. Not terrible. But the soil isn't empirical, unfortunately. So they have a soil mode and a soil count. The soil count in it is soilless. So it wouldn't surprise me if the empirical soil works. I don't think I've got the fundamental parameter. Geo-care. Oh yeah. There we are. That's good stuff. Alright, let's try that. This one is a mixture of fundamental parameters. And it should switch between light and heavy. With filtering? Yeah. We'll move the filter a little bit through. I think this is a dual phase. Can you hear it? Yeah, we will hear it. And you should see the spectrum change, too. So that's the trace spectrum there. I hope this has it. There it is. Do you hear that? Yeah, it does. It switches to the light element filter to do populate. So is that the green and the yellow? Or another filter? It'll be a no filter. There's only two phases here. There. Alright. So 79.9% silicon dioxide. Aluminum 16%. Potassium 4.1%. Calcium oxide 1.2%. I don't know. Do those match up? Yeah, oxides. Let's give it a fair shake. What's the aluminum in this guy? Do you have the oxides on there? 7.31%. Is that elemental or oxide? It's elemental. Here's the oxide. 14. Alright. What's the calcium oxide? 1.2%. Dead on it? That's pretty good. What's potassium? Potassium. 4.35. 4.35. And our iron? No, that's iron. Iron. And then the silicon? What's the SiO2? 73.4%. 79. So, not terrible. But what's it doing? How's it figuring this out? How's it beating the oxygen problem? It's capturing its overalls. Exactly right. So it knows, alright, the aluminum is probably this. The silicon is probably this. The potassium is probably this. That's what it's doing. Can it measure that oxygen directly? No, but it can guess if this is that oxidization stage and this has to measure up to 100, this is what it should be. What probably happened is our silicon dioxide skewed up because of that 100, forcing it to measure up to 100%. Now, if I have calcium carbonate in my sample, what will happen? What's the formula for calcium carbonate? CaCO3. So if I have CaCO3, will this be accurate? Does it know that that's calcium? Can it tell that everything can calcium carbonate, calcium oxide? No. Can it tell the difference between Fe2O3 and FeO? If I have a kaolinite clay, will it be able to tell the difference in Al2O3 and Al connected to potassium and a whole bunch of other elements? So it has to make a molecular assumption and force it to fit. And that's where things start to go awry. That's the key. So you get kind of close when you've got a pure form. By the way, the reason why I can work for obsidian is what is the chance that there's calcium carbonate in this guy? Zero is inspiring temperature when it outgaps all that CO2 anyway. So because it was high pressure fire, you're okay. So anyway, but that's the key. Okay. Let's take a look here at the results. So today, do you know why there's not light in the room? I have no idea. It should be there. It's a habit of the spectrum too. You just don't find it useful. I disagree. I think it's bad to exclude things because who is the judge that's useful to the scientists, right? Yeah, that frustrates me in no way. I mean, it's not useful to you. It's like a self-affiliate prophecy. Yes, exactly. It's an excellent point. It's an excellent point. But in any case, but yeah, that's the kicker. It's kind of unsettling how good the phenolomal parameters can do when it makes these oxide assumptions. So let's try a new sample. I just got a Washington post and every time I see that, I'm like, what now? R&C says it was hacked. It's hard to see that as an evidence. Couldn't have. So we're going to try this on the time zone. Not mutating it. Not mutating it. I'm actually curious how this is going to work. A weird sample. But this is still the soil. This is the geocamp. So this was designed for miners. The geocamp is the calibration. Yes. So here, I get something that looks a lot like obsidian. Do you see the potassium? What is the potassium in the obsidian? 3.61% What's in my iPhone? It's somewhat obsidian. Gorilla glass is reinventing obsidian, by the way. I don't know if you do that. But that was the core idea. So this adds up to 26%, 71%. Looks like it's going to be close to 100%. But I can tell you there's a huge problem here. This is also a soda glass. It's got about 6% sodium. Can this guy see sodium without a vacuum? Does it include sodium when it tries to add up to 100%? Nope. So then, I can guarantee you this is terribly off. Because when they scale to 500%, they make these results look great, right? If I don't know what I'm doing, if I don't know anything about the sample, I will get 100%. But here, because I can't measure a key element, NaO, it's gone. It's just being excluded. We just need to stand silicon to fill that gap, because that's a safe bet. That's also what happened with the obsidian, too, by the way, right? Remember how the silicon was at 71%. Does this have NaO? Obsidian. Do they report the sodium oxide content? 4%. So let's go back actually. So let's pop back one measurement to the obsidian. So this was the obsidian measurement, right? Do we see NaO on this list? But silicon is how much above? 100%. So what's the real value of silicon? Dioxide. 73.4. So all it's doing is it's using silicon to fill the gap. Because silicon will always be close enough. That's the trick. That's how you commit to someone that it's working. That's fundamental parameters in the nutshell. It has to add up to 100%. Otherwise, the entire premise is gone. And that's why, with geologic samples, it can cause all these problems. Because you can actually delete the possibility that that element is present. This is what happened to mining companies. Broker would be able to do good enough. Or Knighton, everyone who's been on one of the parameters would do good enough to make a couple segments. And people would be happy. And then the mining company would find out, oh, how much do we really have in that? Would it be over-invested in this mine? Because we didn't know we had this element. Wait, there's sodium in there. And the sodium makes it more expensive to digest to get the gold out. All those factors start to factor in. And that's where things start to go out by. And the problem is, is because we have an idea that everything should add up to 100%. Yeah. So this is maybe a very basic question so I can all on that. What's the difference between a filter, a computer, what's the difference between a filter, a calibration, and the fundamental parameters? Like what do each of them do, how do each of them work? So a filter is a manipulation of the white beam. Show me some elements. So a filter in front of the beam. Exactly. So I can optimize for this or that element. Fundamental parameters is a type of calibration. A calibration is an algorithm that converts spectral intensities to something else. Most typically, that is something else. It's weight percent, PPM, something along those lines. The goal of a calibration for a company is to sell units. So if I make my geochem mode and I can sell it to minors, that's money in the bank. Done and done, plenty of sale. If I can sell to an archeologist, that's money in the bank. That's the goal of a calibration for a company. What is the goal of a calibration for a researcher? That's key. Reproducible, right? What does reproducible mean in science? I can do analysis that replicates what he did. Exactly. If you guys are two separate people not working together at all and he comes up with findings, you can replicate those findings, right? That's the crux of it. Can I replicate this algorithm? You can replicate the algorithm. Sure. So you can go this way. Who made the algorithm? Did I make the algorithm? Brooker did. Does Brooker have a proprietary interest in publishing how its algorithm works? Why wouldn't it? What would be the reason Brooker wouldn't want to publish how its algorithms work? Because it will show the follow-up system. And more importantly, competitors will know what they're doing and up against. The competitor can figure out what sample will trick their algorithm and send it over. Some people try to reproduce results from this algorithm, can they? No, it's not important. So this actually gets to the heart. Even if you decide fundamental parameters is good enough for what I'm doing, the core problem you still run into is your work as a reproducible. So let's talk about the fram shaft of the day. It seems like it has its utility when you know nothing about the first pass. Yeah. It can certainly be useful for that. If you don't know what you're looking at, you can make it a first pass to see what is here, roughly. But the problem is the spectrum can do the exact same thing. If you don't know how to read the spectrum, that can be a nice entry point for a lot of people. No debating that. Don't worry about how you might miss things. That could be important. If you don't know my germanium chart, my germanium might not show up because the algorithm might not know what to look for. But that's just as true with calibration for obsidian. If I try my obsidian calibration on chart and just publish those results, will my germanium ever show up? No, the problem is is any model made by a human being has a chance, not a guarantee, but a chance of missing what could be the most important thing in that object, right? So David, when I was talking about those sediments earlier this morning when I looked at those rather than defaulting to running things through a weight percent calibration or a partial alien calibration I used that R-Tact program that they had yesterday and it was operating through and just through all the spectrum in together and see how well things overlap or see where spikes were or not. So you weren't using any calibration just looking at the spectrum to see how each sample behaves before any kind of computing went into it. In that instance, when you used the filter you're essentially highlighting parts of the spectrum and getting more specific information. Right, so there were three different analyses done on that same sample. So there were three different groups of spectrum. And then you can see where things maybe overlap and go, okay, there's that, this and that and you can figure out how those elements actually go into a standard values unit and then you can compare the standard values units together of weight percent, this and that. But always defaulting to looking at that line first. Rather than just what's being spit out of the calibration. Because it's being selective on what it's doing, because it's made by us. Right, exactly. So when you start, like, you work for a mining company what is the... what is the input in your algorithm? Are you looking? You start with, like, a sort of family of rock or things that you would expect to find or what? So first, I've got to say, full disclosure I'm in need of a particular mining company and the video is rolling, so I can't give you all the stuff I know. What I can tell you is, from the perspective of a company they make the calibration so that it can kind of cover the most general use cases. And they try their darkness to sell on point and walk away. Because that's their responsibility to get the results they need. That's typically how all companies do it. And that's a lot of times how a lot of people start running into problems too, because they can't trust those results. Big mining companies might use it as a first pass, for if something looks weird, what could be here. But that's about it. They still have to send their stuff off for lab analysis. In theory, if I have a specific mining site, I could hyper-focus on that using Lucas2 equations or whatever. But most mining companies don't do that because that would be too much work for mine. And they don't have the sort of human resources to do that. So a lot of times they have not been able to take full advantage of the handheld XRF Revolution because this problem where I can get results that kind of look good initially. But what I'm starting to produce is let's say, well let's go back to this algorithm. Let's say I've got an previously bad algorithm. This is the one, right? So I measured the city with a bad calibration. I knew it was bad going in and I got 16% ruthenium. If I'm a mining company, is that good news? It's a great news, right? So I can plow in money, focus on this, pull out a couple tons of this magical ruthenium rock. Ruthenium costs maybe $700 per pound and I can sell it for profit, right? But what happens to the calibration problem? Not only do I lose a lot of money in all the investment I've gained. My reputation is for it too when I try to pass it off to something, right? Like, that's pretty much it. By the way, can anyone tell me the technical reason why ruthenium and rooting are coming up here in this calibration? If I look at the spectrum, what's this? That's Rodeo. So the calibration is saying Rodeo in an estimating response, right? Even more to the point. What's most of this rock made of? Oxygen. Oxygen. What's the oxygen doing to the Rodeo? Blowing it up. It's causing it to reflect more and more. So if the calibration doesn't account for that, I just read as Rodeo. The content is where ruthenium would be. It thinks the content is ruthenium. That's what's happening. So what the calibration is doing is its assumption of 100% belongs up in its face whether it's a variable to not include which is oxygen. How could it not know that it's in Rodeo 2? It's not a variable. But I mean, Bruecker... Bruecker programmed the Rodeum correction. They might have programmed the Rodeum correction for things that don't have oxygen. Because Rodeo is a precious metals calibration on something with oxygen, right? So it didn't include that as an assumption. The problem is, we could go back to the drawing board to correct this for Rodeo, but I would still have problems out here, too. There's a lot of other elements it's picking up that aren't there and that's causing the problem. But I want to focus on a really, really important thing. What I've just given you is not hypothetical. What I've given you is the basis of a massive scandal in the x-ray industry with wet toys back in 2013. What happened is someone did this with a toy that came from China. A company that will go unnamed. How do you get it that they gave to a customer and they made the basic fundamental parameters calibration. They measured the toy from China and it came up as 30% lead. And so they published an indictment of the company saying they have lead in their toys that they're selling to children. Company name and bold letters. Company scrambles to handle the scandal. Actually, it was bankrupt because stockholders moved out. And then they found out there was no lead in it at all. It was just this background radiation from carbon. Carbon does the same thing oxygen does. Company went out of business because someone did these and they got calibration appropriately. People lost their jobs, lost their savings because of the bad calibration. So this rhodium example I just gave you substitute rhodium with lead. Same physical mechanism. Light elements that weren't incorporated in calibration drove a company out of business unfairly. But if you looked at the spectrum, you'd see the lead. You'd be fine. You'd know there wasn't lead exactly right. So what's the lead? I did that. I had some Mexican maracas. Very nice, colorful peanut. And I used our XRF on it. Did you look at the spectrum and then have the spectrum? At least. Well, that was our Nikon. I didn't even mean it. I didn't give away the choice. Always check the spectrum. So this is the answer to why I'm a little hesitant. So if you say geocad, could we just use it as a first pass to take a quick look, you can. But if there's anything in there that's going to break the algorithm, I mean that's the problem, right? The only guarantee you have is the spectrum. I brought up Plato's allegory of the cave early because we're all chained to this cave. I don't know, you weren't here for this. The allegory of the cave is we're all chained together and our backs are facing the center of the cave and we're looking at a cave wall and there's a fire burning behind us and there's something moving in the back of it. We can always see the shadow on the cave wall and so we try to introduce the true forms of and that's Plato's allegory of the cave. The calibrations are methods of turning those shadows into something we understand. If we look in detail at the fundamental parameters algorithms that are here, these are the fundamental parameters algorithms each of them try to relate the weight percent so the concentration of the analyte into something we understand. Right? C i is the universal and external material science for concentration of that analyte. This can be, so for example in the calibration that backfired where the extra f thinks there's what, 16% rhodium in this obsidian, rhodium is here and the intensity i where rhodium is and it sees that it's super high and it says, oh yeah, we've definitely got a signal for rhodium. Now why would a precious metals calibration have to include rhodium at all? Because rhodium can show up in precious metals, right? So that's what's happening here and then all these physical parameters if I didn't have oxygen present would have worked probably just fine. They worked great for this right here but what's the difference between this and this for the fundamental parameters algorithm? Oxygen, right? Can it measure oxygen? No, but it could measure every element in this. If this had, for whatever reason, rhodium in it in these concentrations, this precious metals algorithm probably would have gotten close to what's actually there because it would have recognized it's outside the balance for what normal performance is on a level. It's the oxygen that throws it up. It's the under variant. So that's the reason why I would recommend learning read to spectral and learning read to spectrum if you're going to be working to that level because a quick pass you might miss something especially if you're working in a research context that's something, typically speaking when you have something important that's unusual. It's not going to show up in the actual fundamental parameters, right? Because it's the algorithm it won't even show up in your empirical calibration that's not showing up. So my best recommendation is learning read to spectrum above all. Going back to Plato's allegory of decay is this algorithm finding a true form? Is CI, concentration of the analyte something that actually exists in this? Yeah. No. What's your case that it exists? That it's a true form? Uh, I mean that has concentrations of... What's your case that it's not? I thought we were talking about rhodium We can say it's... Yeah, the algorithm is basically telling you to play a ton of ideal for that form what your ideal should look like and doesn't account for the fact that it's not always going to hit on the ideal form so like it's not... I'm just guessing. What it is is let's elaborate a little further the concentration of the analyte is weight percent an absolute or a relative measure? Relative. Can a relative be a true form? Sure. In what context? How would relative be if you know what you have is a hundred percent? If there is... If I'm at a hundred percent then I've converged on the absolute but here's the thing it's a preconception of what a hundred percent is In the object let's say humans were back in the age of genosaurus and a piece of obsidian is produced what intellectual ideas we have about terminology, percent science don't exist What is in the obsidian in raw natural state? Is percent something that exists in the obsidian as a product of nature? What is the product of nature? What is it made of? It's made of all these it's atoms. It's just bunches of atoms, right? So atoms in this case would be the true form atoms are what exist in nature percent is a human construction for understanding that reality PPMs. They're great understandings of it but my core point here is that the goal of the fundamental parameters is to take information about the atoms to create what we call a synthetic unit so Aaron Ragnowski and Anna Steffen wrote this article a while back and one of those like tubes that no one reads but it had this really important distinction You need machines in archaeology Yeah, so in that volume it's fantastic to read I'm actually thrilled that someone read it to me it's like I always hate those edited volumes because I feel like they just disappear especially when they hit like the 20-30 year mark Richard Hughes as a chapter He does he's got the creative city in chapter where he distinguishes from geochemical and geospatial sources but the argument of Anna Steffen and Anna Ragnowski in this is I feel bad because I just dissed their work and said don't read it but I guess I did but my apologies but the key behind their idea in that volume was that there was an analytical and synthetic unit synthetic units are the product analytical units are the raw output from any instrument it could be XRF, it could be FTNR, it could be GPR and it's this raw and filtered data the synthetic unit incorporates assumptions about reality right so like in this case what would the raw unit for let's say Zirconian in the subsidy and B from the analytical interpretation Photons? Photons, suspecting what is the synthetic unit that incorporates some of my assumptions that's the key so my core point here is where fundamental parameters most often fail is in the assumption that CI they don't recognize that CI is a synthetic they don't recognize that assumption so the reason the there's nothing wrong with using physics to predict properties of the atoms of the spectrum there's something wrong with trying to predict the concentrations in the spectrum because the atoms are something that exists it's a real true form that exists in the object always has, always will our idea of percent and PPM incorporates assumptions about its composition in the case of this object everything adds up to 100% is correct everything in that spectrum the copper, the zinc, the lead adds up to 100% for your key does not factor in for here because we can't measure the oxygen so we're missing that variable so the assumption goes awry and thus the synthetic unit does not properly reflect the analytical product of this does that make sense? so I'm missing the oxygen just because we can't measure it so in fundamental parameters fundamental parameters, algorithm just for obsidian could it be good enough to do most of my work? yes, but there will always be the chance I get surprised by something and that throws things up and that's the key thing to understand about fundamental parameters so for me having worked in the industry and having been a published scientist so I've been on both sides of this the problem we run into is that in science, when I put my name on that paper I'm accountable for those results the company once it sells the unit is no longer accountable for those results because you bought it their incentive is not the same as yours and it doesn't mean that they're evil or bad it just means that they don't have the same incentive structure as you do which is why empirical calibrations work better because that way the interests align and you end up with something that works out very well now let's put that in mind let's have over here to the Lucas-Toothed Algorithm the Lucas-Toothed Algorithm this is what is going on this is the opposite this is the empirical so the difference in empirical and fundamental parameters isn't in what they're trying to do they're trying to predict CI the difference is fundamental parameters the difference is in physics within the instrument and its parameters to hard calculate what the CI should be empirical gives up on that and says we're just going to use some standards and figure out what it works best so empirical is just it's been criticized for being just brute force statistics I'm going to use the slope the intercept I'm going to basically model it out statistically and figure out what it is and my answer is only as good as the standards empirical so this would be y equals mx plus b right here and then I add in the influence of other elements and that's basically the idea behind empirical so in empirical my assumption is that my standards are good and final parameters my assumption is that all these physical parameters can correctly estimate any unknown sample that comes up to it and in the case of the Eucidian standards they're analyzed using ICPMS activation and then XRF was calibrated to match other results whereas when I'm on a parameter you don't know you don't have the same basis of knowledge but one thing I often like to comment on is this quote from Winston Churchill that I've always loved it was when I started working for Brooker I got really into World War II because everywhere I went in the world I remember when I was on the small island in Taroa in the Gilbert Islands in the Pacific seeing Japanese soldiers being on Earth after the like it was so crazy we were there trying to recover American skeletons and I remember they found this one bunker and there were three Japanese soldiers still fully dressed bullets on their chest lying as they had died from the initial bombardment and it was crazy to see that finding but this was on a small island where of course you can see lots of effects of World War II I went to China where I saw other effects of World War II like it was such a major it's sort of like World War II was Lord of the Rings and we're just kind of after that not realizing how profound the world was transformed after that and so I really got into Winston Churchill because he strikes me such a heroic leader during that though he made mistakes but he had this quote about the rise of fascism that's always resonated with me that democracy is the worst form of government except for all the others and that was his defense of it and his point was is that it has all these failings it's terrible but when you get married to all the alternatives it works out pretty good like it tends to get the better result every time the same can be said empirical calibrations empirical calibrations are the worst kind of calibrations I've only as good as my standards can I find a piece of the city that represents that it's absolutely possible I'm very aware of those standards is it a pain to measure and build the calibration for every instrument every time absolutely it's the worst way to calibrate XRM instruments except for all the other ways because they have all these crazy problems that can really come back to bite you the lead example is only one of them let's take a look at and I don't know if you're going to use XRM in your work but it is very important to understand because this debate comes to the epistemology of these scientific instrumentation and archaeology and that is Nico and Lucas have probably heard this to death but that is the debate between Ellery Fram Ellery Fram and that's Robert Speakman Robert Speakman and Steve Shepp I forgot Steve's first name so it's about these guys and the background of this is he had a niton unit and the niton unit is basically designed to do fundamental parameters and figure out what's quick and easy and Ellery Fram is really attracted by it because it's cheap so in other words it removes an economic hurdle for people to do research because in nitons I collect half the cost of one of the worker systems typically so it's half the cost because they're not going to have time to calibrate individually and they don't have all these basic fundamental parameters algorithms and by the way fundamental parameters algorithms are just written and they're just pasted on each unit separately they'll usually do a little bit of standardization to customize and paste the tubes a little off but by and large the algorithm exists independent of how hard bear it's on kind of like how the macOS operating system is on all these different computers where your phone system is so anyway what Ellery liked about it was that I can just get a result quickly and this is great for developing countries where they don't have an unlimited amount of money to spend on these things and it also means I can do everything out in the field as quickly as possible so I can start to get a giant set of data like his viewpoint is very understandable but what he wanted to do was assess its validity if we go back to that same obscure volume that was published by the 1980s or something but that, oh no, 1980 1990 because that's when Third Eye Blind's album came out anyway I remember on Twitter someone asked like what's the thing from your childhood that you weren't looking forward to but don't look forward to anymore and I was like Third Eye Blind's next album but in any case in that volume which accused article was that there's validity and reliability in understanding these systems reliability kind of comes down to precision but validity is does it correspond to what's actually there at the heart of it and we're actually valid for understanding what I'm looking at and Fram wanted to say do these off the shelf calibrations work and so he tested the fundamental parameters algorithm and he tested the versus of MERS results for the same week's standard so as you can see he measured a ton of samples here it is, it is he did a lot of work he compared Serena with a team in here he used a laboratory system of Mer and then compared it to what he called the HHPXRF for the handheld portable XRF so is it a miton 3, 4 it's a gold or whatever by the way you can tell with Mer that they were going for that name right but anyway, so this is the lab result and then this is the result from his handheld factory calibration what do you think, is that good or bad it's not great not great what would we use which of these numbers would we use to assess the performance of his handheld versus the lab R squared R squared miton it's a residual between yes, so the idea here is so this is the known on the X axis here it should be flip definite but this would be our authoritative states so to speak from the lab and by the way they destructively overestimate so one of the ways labs give accurate results is they take this, grind it to a fine powder press it to a pallet, put it in the machine and the machine is precisely calibrated for exactly that matrix so it gives a much more reliable result as handheld we just kind of measure it as it is it works and all so anyway if we flip it around and the other parameters change the R squared would mean 83% of what the lab saw was predicted by the handheld so that's pretty good I would actually argue the R squared is not the best measure for this instrument the best measure is the slope how do we interpret the slope so slope to me if I'm going to compare two different instruments the slope is more important and like the R squared value it should be as close to 1 as possible a slope of 1 means for every ppm to the lab saw the x or x on that ppm yeah what's the slope here so how do I interpret that slope that means for every 1 ppm to the lab saw how many ppm to the the unit C half the ppm 50% off is true value so in other words the R squared measures precision reliability the slope measures accuracy it's kind of hard to wrap your head around it but when you're working with a known and you do know what it is the R squared is only telling us how close the line the points are to a given line the line itself this is the core problem the line itself the data is not a product of the know that's by the way a shameless plot of laying cloud cowl whatever it is cloud cowl you see this dotted line that dotted line is the slope of 1 so that way I can see does my regression line have the same slope as the dotted line in the case here it does if I use the bad cowl let's do total counts trying to shake it off that line so what are the problems when I coded my calibration algorithm I actually forced it to come with the closest answer to the line so it's really hard for me to break it let's see I don't think I can break it here but in any case the idea here is in this algorithm even though this calibration is terrible it's going to ask for a point 2 the slope is still 1 in other words it's still according to those basic predictions let's do rainforest again when I do my rainforest algorithm I should take a second here I might break so in your red cloud cowl in every case you have the counts on the lock in the estimate percent people like to like to do it the opposite way but I far, far, far prefer to see the known on the Y XORF material scientist calibration terminology exactly backwards how the rest of the statistics does it they typically want to see the counts as they know exactly that's exactly the right drives me nuts every time I like to be as strict as possible but everyone's got the preferences I might get too much blowback I might switch cloud calisthenics and technically Ellery Fram is doing it right in the material science sense but in any case I designed this so that you can compare the actual line from a regression to what it should be with the slope of 1 if on Ellery Fram's paper I plot the 1 to 1 it's going to course up here right so that would tell me immediately visually what I know from reading the slope which is he's systematically off every one of those predictions from the handheld is going to be 50% off because it's not getting oxygen or I don't know why on the instrument it's not I think for titanium it could be the oxidization state it could be it's not factoring the influence of iron on titanium there's all kinds of reasons why it could be so far off the point is is that it's always going to be off now what Fram argues is that he can post-process this he says look I can add in and correct that but then what he's basically doing is he's making it here for calibration anyway instead of using the photons he's using a derived method from someone else and then fixing it it's what we call type standardization in the x-ray field which is I'm matrix correcting but like if you have to do that anyway why don't you just do it here from the beginning isn't even going to result so that's titanium and it's not a black box exactly so let's look here this is a neutron activation analysis versus the handheld x-ray how does this calibration look? better, why is it better? it looks like it's supposed to 45 at least what's the slope of value? so it's still 38% off it's actually going to overestimate the manganese, right? so if my manganese is here it's going to systematically overestimate my 40% these units are different it tricks you into thinking it's a 45 degree angle that this drives me nuts I'm curious I don't know like I like to say journal archeological science is like the Daily Mail of Archeology like you get some really good articles in there and some you get are just terrible but yeah, no that right there he forces it there basically with the axis I don't think he intentionally did this I know the software he's using, it's Mach numbers he just plopped it and Mach numbers does that but I think it fooled him as well as the reviewers but in any case, yeah, it's still 38% so let me put it this way just to highlight why finding out parameters causes a problem in mining if I'm a mining company, let's say I've got manganese money and I need to have 1 weight percent of manganese in the rock to make a profit extracting it because it costs money to pulverize it to adjust an asset and extract the manganese to this calibration to make money because the problem is it's 38% over if the calibration tells me I have 1 weight percent manganese I might actually have 0.6 weight percent and I'm in the rep you can see my mining companies got bitten pretty hard by these fundamental parameters algorithms right, because systematically if they measure the misstanders the misstander comes up great, that's wonderful they probably use the misstander to guide the model but when it actually comes time to make money I'll bet a million dollars on this and I've got that kind of an error I'm out so I won't last long so anyway, that's the key there how about this one, zinc good or bad? 78 so it's closer it's only 22% off now it's going to underestimate 22% of the time you can kind of see here where I think the core problem is anyway which is you're systematically gone now I'm doing the obsidian sourcing so this is a rubidium rubidium is a core, how accurate do I need to be with rubidium to be able to accurately source what would you say your acceptable error is in obsidian rubidium to source you want like 95% well let's say what about absolute terms, right each source has an absolute definition of how much rubidium is in it or is acceptable in ppm so 5 or 10 ppm 5 or 10, does that sound reasonable it depends on the region but what region would rubidium be essential it's important in the Andes in the Andes how many ppm off from a value do you have to be to say this isn't from that source to do a negative test what would you guess for alka I mean it's the ratio that's fair, that's fair from North-East California where there's a lot of variability we wouldn't want it more than 15 ppm 15 ppm off if it was like if everything was off the same way that would be workable you wouldn't want just rubidium off so if I've got let's say I've got this calibration which is 0.66 so not great but this wouldn't work that wouldn't work but let's say it was 0.9 let's say if I have 250 ppm from the handheld what's the actual concentration of rubidium in that rock if my estimate is under well 85 that'd be 30 ppm off in addition to the inaccuracy that's the crux of it so that to me this has a failing instrumentally right then but here's the thing Frant makes an argument where he says look these still cluster I still know, even though these results are inaccurate that this is a cluster and this is a separate source I can still make my source assignments this way so what he uses and this drives me now he quotes what I told Adam Nazaroff back in 2007 he quotes us and he says even though the calibration is off the instrument itself is still internally consistent and I remember the bar that I said that to Adam that we put it in our paper internally consistent so I can still make the source attribution as long as the instrument is internally consistent so his argument from his data we published it we did publish it in Jats in 2010 actually so he's using our research in this case which still drives me nuts but I remember talking to Adam taking through the results because what we did back in 2010 is we used Steve Schatt's machine here as the control for the broker tracer, we did the test I remember it because I had worked on the Obama campaign and it got done and I was like I really need to do something like a launch floor so I jumped on that project which is so funny this is a case where something like Bronze Age never got me a job but this like a third author position on this project launched my career but anyway, it's funny how the world works but I remember when I looked at that data let me see if I can bring up our paper it's just funny because I'm like where that lab is oh by the way my favorite part of this is we document I shouldn't have used the shadow of the figure again, Jester viewers but this is the diagram of HughesNet's 98 paper so that way you can see what he means by validity so in any case if we use this if we use this precision would be the replicability of the measurement accuracy would be comparison of the nine sources validity is the ability to distinguish the two components of the source and then validity is the ability to address archeological inquiry that's HughesNet to 98 so in this case let's judge the Necton unit making replicable measurements with Fran's data on the blood of Necton right, let's give it a no comparisons of known sources, does it work? no, does it accurately distinguish two chemical sources? it distinguishes the two that's a yes, we're going to do that but accurately it is the problem we'll put it aside we're going to take these in isolation I'm going to give this as a yes to Fran and then does it address archeological inquiry? yes so that's by the way why Fran called it validity because it meets Hughes's definition he is totally side stepping reliability and precision and accuracy he's saying is it valid and so in that case that's what he's thinking so yeah but anyway what we did is we used cluster analysis to compare Berkeley's unit with our tracer so that's why I'm really proud of this, I came up with the bullseye idea because I didn't know how to visualize how two instruments performed so I used K-Means cluster analysis which means I tell it how many clusters there are in the data and it automatically draws the bullseye the bullseye had to calculate the core of it is is a trance to find the center of the data and admit that for the tracer you can see the offset it's a miss it's systematically off here I can see the I can see the systematic drift one unit to the other by about 10 ppm so the bullseye is the initial equation now? the bullseye right here that is drawn that is Steve Scheffer's history so these are the letters there along the tracer we had that we just got and then we measured the same same obsidian pieces here at Berkeley as Steve Scheffer used to do so the dots are from the tracer the bullseye here is from the lab unit here because again visualizing the two is really hard so the bullseye is designed if you eyeball it you can see the drift between the two and I am coming in about 20 ppm, 2 lb zirconium and 10 ppm, 2 lb zirconium I can quantify the difference between the two it's not like a standard deviation or something, the outer and inner circle exactly right, this is standard deviation number one standard deviation number two on a single on all these measurements these are 1, 2, 3, 4, 5, 6, 7, 8 samples of Mr. Pecquet this is the deviation of the k-meat cluster so the deviation here by the way is not based on the data it's based on the k-meat clustering algorithm it's a little complicated that's why it's not in the list that's exactly right, not in the list because the idea of k-meat cluster is I've measured a center point of the data and then I have the variation in that center point as varying standard deviations so that's why it isn't in the list although I would argue that if I did draw it once in this lecture it was already right here anyway but the idea here is is that theorist is meant to be off but when I told Adam in the bar he was like well what do we do I don't want to publish this paper until Bruce is wrong and I might be absolutely do that what's more but we will point out we'll give this bomb to everyone that the device itself is internally consistent once I understand my offset then I can directly calibrate it so our point in this paper was it's a solvable problem Brooker just isn't there yet I got so many angry emails from Brooker when we published this and then they hired me on a PhD but this was the crux of what that was so anyway but that's the key right so we argued this and the point of the paper that we wrote was that this is solvable so I argued this is my mistake I argued that the Berkeley instrument here by Steve Shackley was accurate the authority of Steve Shackley meant this is a known value and the instrument is off I went wrong that was right so the calibrations this is from a conversation with Bruce I've been videotaped if I misrepresent you with Bruce or Steve Shackley I apologize what was explained to me was that there were many centers of obsidian who wanted to do obsidian work that was Mer and Steve Shackley I don't want to name any names one group put obsidian on epoxy that we were talking about yesterday so their backscatter was influenced by epoxy another group normalized to all the background so instead of normalizing the component they normalized everything from 20 to 40 keb so either of those two things either having the obsidian on the epoxy or that would cause this type of error that type of systematic error so what we're seeing here is not that the Berkeley instrument is correct and that the tracer is off that same goes about 5 ppm around each I would say we're seeing a difference in how the data of the spectrum is processed because if we mount the obsidian on that epoxy we're violating the assumption that it's 5 ppm that's the crux of it so this is actually what the student yard was back then this was as good as any archaeologist was going to get with any study so are these calibrated there at this point and what kind of calibration was it it was the Mer version 1 obsidian calibration but like I said I'm not going to name any names because I don't want to name but there are two problems with the existing calibrations back then one was that they were mounted to epoxy so this backscatter was inaccurate and two they were using that normalization protocols and so it was a consequence but those two in place by the way if you normalize from 20 to 40 kev can you tell me what the biggest thing that's going to predict the value of rubidium how thick the obsidian is because I'm normalizing all that backscatter I'm exposed to how thick the obsidian blade is so blades will show up over here and this happened to some like well back not in the night I think it was in the 80s or 90s something like that and they ended up completely supplying all the obsidian sources they made an argument that all the blades were from here but all the cores were from here so obviously they were importing obsidian from one place and then exporting to another that was the economic relationship the backscatter was causing the underestimated rubidium and the straw team and the misassociated to another source you can totally see how that could happen though right yeah anyway so here I was being unfairly critical I think that's just a critical out of that Steve but either way, if you measure and later I think Kurt Ruddmaker did something similar in Olga and he actually found the differences between neutronation analysis wavelength dispersive and hemilextrorept actually come out to about 10 ppm they did ICP in the mass as well and then they saw that time level of variation right and then there was that report that from my class colleague who was Kaiser where they did all three methods exactly and then they came up with this is sort of the 2012 exactly what frustrates me is that they never published that officially that should have been a jazz article I think even looking through for citations too I was surprised no one did that but anyway this is the case with Graham's article he used our argument what I thought we were arguing for good calibrations he took it as as long as the instrument is internally consistent it is valid in Richard Hughes sense and we don't have to worry about reliability we don't have to worry about accuracy or precision so in other words this right here is beside the point we don't need to worry about that enter speakman and shackley who take a very different case they come coin this turn that's going to be hard to read they coin this turn it's silo science and what they mean is that as promulgated by Graham and others I think I'm the others that it is perfectly fine to provide results that are only internally consistent and do not conform to established international standards and data then we are entering a time of silo science where each researcher's data is self-contained independent and cannot be verified we find this is not only unacceptable but another aspect of the social science that is definitely the science aspect plays scientists with portable XRF technology burn mic drop right so the core point they have here is that Fran arguing that internal consistency is the only metric for the performance of these XRFs means that each XRF is its own self-contained world this XRF could I get it could I get it to tell me what's another South American G. Viad, etc yeah I can do all that that's good in archaeology these guys are like who can test you who can come in and test your results and find a problem with them if it's fundamental parameters and it's a locked box that it's not reproducible and then every individual worker is working independently we just have to take their arguments on authority we can empirically test them that is the argument of speakman and chaplain so to point out the reason we use weight percent in standards and empirical calibrations is to make everything reproducible so that way the work you guys can do can be picked up by someone else and it's reproducible, it's invalid and it's reliable in the full sense which are used back in 98 that's the idea behind this world about that and that's really what the fundamental parameters on empirical calibration is about it's not about whether physics is a standard in terms of estimated concentration it's about whether or not you're doing science or not because if you can't reproduce it it's not science, it's just authority we haven't moved that much far past the allureo if that's the case that's the crux of the argument so in the report this is the report from I think I've got a copy I can bring that up too so in conclusion he says do you guys feel like using this method for the best results with so any best elements for NAA analysis are sodium, aluminum, chloride, potassium S, C B, E, S, C V, N, S, B C, S sodium, sodium wait the double is sodium yeah yeah and then that's what is that the microwave so I I don't know what the M.A. is ICP, M.S. is a couple plasma mass spectrometry so anyway the ones we care about are strontium, petroleum, aluminum better from exactly exactly and then what they do with their calibration is they use the values we use for strontium the values we use for rubidium come from NAA exactly so to understand the difference between these two techniques neutron activation this is really important for you to understand because you're going to have to cite a lot of this out neutron activation analysis is I take my sample I put it in the nuclear reactor I close the door I close another lid door I go over to the next building and I put turn it on and the nuclear reactor activates everything in this guy like in other words we're irradiating every single element here and those elements start decaying we're just filling them up with neutrons and now they start breaking down and they produce a spectrum spectrum that shows the decay of those elements and elements like rubidium decay into other elements and produce a spectral signal so they can measure by the atom what is being produced and they can use that to infer how much rubidium was there based on how much rubidium was activated and decayed when they're done they have a machine that takes the sample puts it in a nuclear containment thing because that sample is now radioactive for the next 400 years and now it's radioactive waste but we now know a lot about what was in that sample what makes neutron activation analysis good and we can do partpabilia and we can do a lot of elements so there's an argument by Patty Crono from 1983 she's the wife of my disfacer advisor and she's our nuclear surrounding source and this is the best method XRF can't compete even though XRF is attractive it doesn't give you the range of elements you need to know you'll need to know that because you'll have to cite that if you find good results with this or if you find bad results with this that's good you didn't say what is the difference between long and short irradiation you know I don't think that long irradiation, long something I think it calls smart is he talking about the wavelength in that context or is he talking about the duration do you know I think it's duration if it's duration it should be the same basic energy the longer you run it for the more expensive it's going to be because it's a nuclear factor but the chance of getting smaller quantities of exotic elements I think is what's key there so that will change the detection limit I think that's probably what he's referring to but I don't know for sure but if I had to guess that is what it would be but that's for your navigation analysis now the thing is not all elements break down in the same way so that's why we see that some elements are great for M.A. and ICP M.S. that's what we're using in the next couple of hours of time what we do there is we fire a laser by the way the ICP, this determines I'm going to be digested in the acid into the solution or you do it on the physical sample in this case they put it in a four fold acid dilution which is crazy dangerous they're using stuff like hydrochloric acid that is just fantastic stuff but it dissolves everything and decouples from every other element and they take that solution and they fire lasers at it and the lasers break down the outer electron shells and that creates characteristic spectra and that can be used to quantify things I think M.A. is a typo down there it's called M.D. it's microwave digestion that makes sense so one thing I want to highlight here FE is better started by new generation of analysis FE is not better studied by ICP M.S. brooker has made a handheld ICP system it's using a laser and physically works a laser evaluation it actually works great with ceramics by the way let me show you that it is it's not exactly on a ceramic you usually don't notice it but it is absolutely destructive this is what it looks like here and I wrote a fingerprinting out it's similar to what you get from ICP M.S. so here it's better it kind of looks like XORF with two different lines this is what it would look like so here I've got oxygen, calcium, silicon, phosphorus but there's multiple sodium is actually super easy to measure with this technique it shoots a laser on the sample it basically creates a ball of plasma and the ball of plasma spits out photons from the outer electron shell of what was just in it elementally it's still the same thing you're basically creating this cloud of electrons and then you're able to measure from that light what you're seeing in any case when you produce that the fluorescence is changed so how many peaks does iron have in XORF? two two, right? can you guess how many peaks iron has in ICP M.S. or laser induced inflation? 18,000 so so that's why they use the neutralization analysis for iron because it's just impossible to do with ICP M.S. it's a hot mess they typically take one of those peaks and hope it works so anyway, your spectral interpretation issues you can do qualitative XORF spectrum work it's really almost impossible to do qualitative ICP Lids work because there's just too much but anyway, that's the key there so that's just one example of why one envelope could be better than another because they have different properties on a rich method but they combine both leaves and they produce that and the point here is using this approach with this set of obsidian if Niko does work and if Lucas and Kapter do work it's equivalent they can compare each other's results if Niko goes, Lucas is crazy, this is wrong jumps in with this unit, he can test Lucas directly and get the same result and that's reproducible and now it's science, right? the arguments you make are substantiated by that data set in any case, but that's the core of what they could hear Fran, unhelpfully comes in with a new argument wrote in his papers here so this is where Fran responds to Steve Shackley or Robert speaks to Steve Shackley and I'm just going to read you the title is it a city in sourcing about geochemistry or archaeology what would your answer be? both, both, yeah why would it be both? because you can't do both without the other so Fran is trying to divorce the idea that archaeological questions require good geochemical data whereas I think the argument from Steve Shackley and Robert Speakman would probably be thinking Jeff he does prefer Jeff, I know it's just the published words it's sort of like when I was a kid I pronounced items items because I would say a freak inside in a box of cereal there was also 19 oz so in any case the argument that's the fundamental debate can be divorced, archaeological inquiry from the base of geochemistry the group who uses empirical data standards of data says no, you can't the group who likes to buy cheap instruments and point and shoots says it's good enough for government work so I wanted to make sure that we're staying on the point of ceramics because the obsidian has kind of been monopolizing some of this conversation so it's really easy to source obsidian and have enough ceramics when you have a site that somebody worked out 10 years ago and they end the tracer and they created chemical groups based on their analysis have you been have you seen people going back to those same ceramic assemblages and analyzing them again the same way we would do with obsidian are people doing that are there issues let's hold that thought the answer is yes there are people looking at it but there's a lot of mixers because you're not necessarily sourcing the ceramics you're just creating geochemical groups that is even more complex than doing much more obsidian you don't have sources you just have chemical groups powder comes with clay clay sources ceramics are complicated I think a lot of the work is obsidian sourcing is fundamentally predictive ceramic analysis is mostly descriptive we're describing the different groups we have as best as we can I think in most cases true sourcing is geochemically impossible whether or not this is extra for NAN because the reality is in spite of a volcano there's a dot on a landscape that produces rhyolite every once in a while boom, right there, that's my source there's this defined region in space and time that produces that a geologic formation can extend across half the continent and have outcrops in different areas and geochemically I might be able to find a Scandinavian ratio pinpoints this formation that formation is huge the nature of this variation is fundamentally different than it is with by the way, I cut myself with the obsidian good job making it I'm good for now we were talking about earlier you can't really necessarily source ceramics because you have different ingredients and different residences we're talking about how cookware in the mycelium world gets tempered with salt and other various things obviously you can't source each of those to one location in argument, some people use micro probe so they'll use an SEM and what the SEM does is it fires electrons which gives you a nice surface topology but I can target a clay species and ceramic I can shoot electrons at it and I can get a spectrum just like XRF it's a little different because I'm using electrons I'm also doing it in pure vacuum so I can see things like carbon too but it's also all going to be the same thickness you guys have one? I'm just going to leave the question to you guys you will want to use that if some of you are working, absolutely the negative by the way of that is it's spectrum is typically confined by about 25-30 kV so electrons like the medium strontium don't show up well at all XRF is way better for those but for light elements compositionally it can be very very useful for your work too but it isn't necessarily helping us to change it it's like 7 7 cm we've got an older model the composition per size is larger they've got the same amount because they're miniaturizing the same way everyone's miniaturizing people are using it as a circuit board oh yeah you can see the carbon but anyway, yeah, so that's the deal so in any case if you are sure it's small enough you will get the EDS spectrum and by the way the rules of EDS are a lot different than XRF so in XRF we're talking about the resonance where titanium fluoresces and activates another thing, not quite as true in energy disperses in electrons because the electrons don't do that now in the product light they still do that but it's not working with the same depth of penetration so those effects are much reduced so you can use that for some of your work if the struts are smaller you just have to bring it back to the US you seem to bring the EDS with you put it back to the US maybe there's no but anyway, that's the key thing but it works with some of the properties so on the note of ceramics I want to segue, actually I need to use the rest of it first when I come back I'd like to segue but I'd like to do, since we've got a very qualified crowd here, I'm going to do a really deep dive into ceramics and not just ceramics but the environments that form them so when you see elements you don't have to interpret them it doesn't have to be relevant because if you keep the stratigraphy and archaeology these same rules apply they're just not on a different scale of time but all of the chemical interactions can be very simple so I'm going to hold that thought and I can like 10 seconds and we'll continue so have you started doing any kind of analysis with this? no, no, no for me this is very yeah it's just like it's this, you know the things I do about x-ray control coming in are that it's non-destructive and it's purple so for someone who works in Greece where you're not really allowed to do any part yeah, it's portable it's incredibly portable wow, so this is like straightening this is like out of the front you painted it in the fire exactly that's why you have to go portable so tell me what a filter it is cause I don't quite get that even for me I'm like if I go to the bathroom to go back I'm going to be lost I have to go and I have to make it as fast as I can really possible this is all really good because I'm starting to understand yeah, right well you can make a lot of different deals just keep going back for a lot of what I would this is not something you can do for my dissertation but the research I want to do after that basically having something like this and I can go to Greece and I can point to XRF at the Connery another artifact and be able to say it's just not Greek is yeah, it's helpful because there's debates on that I mean you maybe are you a fan of the video that I'll approach at the third plaques at my CD? no, I've not been to my CD yet but yeah that's one of the things I look at at my CD you have sort of a series of these finance plaques they're still in some debate about whether they're even made in Greece or Egypt I do think they're probably made in Egypt but they're still yellow do you want to source that another plaques to see what it might be? yeah, potentially that's one of the things my exhibition that I worked on requested the XRF to bring to other stuff and shoot the one that you have because there's only one that's been found outside the palace that's in the first site shoot that potentially compare to that with other finance and sort of be able to prove I would just like give this a tip what is finance exactly? I can go over that later but I heard the term come around I haven't really looked into it but it's helped more than all that it's solid it's a silicate essentially I don't know a lot of the science that I'm used to looking at let's just google it but I think it's essentially it's ceramic adjacent the glaze on it it's a style so what I would say it's not ideal as you can remember what I would say is this we're going to go into this a little bit this is like the finest watch you're ever giving to I love paleoformatic reconstruction a lot of deep time I really shouldn't have done archaeology I should have been a geologist but the thing is you want to know how Egypt Egypt's geologic history and Greece's geologic history and I'll make it simple Greece is the product of the uplift of coastal sediments from the movement of Africa north so in other words the Europe and Africa are squeezing together that's confined to Mediterranean it used to be bigger than the Pacific it's the Tethys Sea and it's shrunk down to nothing both Egypt and Greece are the product of that uplift so they're covering the geology of the Tethys Sea which is good because different marine basins have different complex geochemistries that happen because that salt and water makes things possible and aren't possible in fresh water in terrestrial areas so you end up with so what you end up with is these geochemistries and that can be how you answer questions related to this there's that kind of that kind of deep time so in any case that's the idea so I would argue that the most important thing you need to look at is the geology which is not a welcome news I think because you're much more interested but the more you understand about the geology you will understand how to interpret what lived it up in a clay and what that means about the environment it formed in and then the probability that it formed in the Greeks or the Egyptians or something along those lines no one in archaeology looks at it to that depth typically speaking it's almost always geologists who ask themselves those questions but to me understanding that formation process is going to be the absolute clue to what's going on there which means there's a calibration if you get the five up and that is the mud rock calibration because it is calibrated for exactly that I get a lot of flak from this from archaeologists because archaeologists and I'm going to talk about that when Kathy comes back here too but like Brooker didn't like it either Brooker's like this is not just calibration it's built for but you have to understand where these things come from and we'll kind of go into that in a little bit so mud rock we can't create it ourselves you need these standards and the reason we have 169 yeah that's fine you can certainly add that to the calibration if you'd like to too after the fact the calibration that Brooker has now is an updated version of mud rock that's based on 40 42 standards I think not the original 26 but I'll talk about why I think it's crucial to these kind of ceramic sources it's not because ceramics and clays from the earth are two different things because they're not really like obviously we have tempered salts and it starts to change the amount but remember the XRF is measuring everything, two of the XRF a compressed clay from a geologic formation and a ceramic are the same thing clouds of oxygen with other things mixed in like silicon, aluminum and all that so quantitatively you get close to the same weight percent what's important about the mud rock side is it's trained on the elements that can key you into paleo environments and it's the paleo environments that can help you source within a geologic formation let alone an entire way and that's why I think it's essential to add a lot of clays from different areas it also helps you understand the verbs not just the nouns because like obsidian we don't need to consider the fact that obsidian forms from different environments because by definition it's always from the same thing a rhyolite producing volcanic eruption almost like a tech type let's say a tech type or about the salt or let's say it's compressed ash or something like that once they start to develop those different kinds the elements aren't just paths of descriptors of source, they're active descriptors of what happened to producing that rock and that's the key, the more you understand about those formation processes the better around you're going to be in conferring from the water so with that let's take a look at the mud rock calibration they did what I think my last lecture which is publish the cow in journals so this is in chemical geology now what they did there, I can send you this paper too if you'd like I think this is my favorite empirical calibration paper because they've published, by the way which element which line there's slow corrections, there's background corrections and the standard cabinet so it's the most open calibration I've ever seen and let me tell you something even more the people who funded this this to me is the most open science approach to abstract analysis, they published everything about their calibration everything exactly, and they did it at the behest of the evil oil companies so the background of this is this calibration was not built for archaeologists it was not even built for scientific research it was built because oil companies owned lots of money so the background of this is the fracking revolution so in the United States, fracking as a technology has completely transformed our economy, right? America back in Bush years, we talked about energy independence all the time how do we get independence in countries like Saudi Arabia and all that because the things they do now, we don't talk about energy in the heads anymore because we are energy dependent most of the oil used in America is produced by all of America now we outproduced Saudi Arabia now, which is crazy, right? it's all because of fracking now, we say in Saudi Arabia, not because we need the oil but just because we're okay with killing journalists but but anyway, the key thing here is what transformed that well, it has to do with the horizontal movement of pipes so in an oil company, do you guys remember that you guys remember watching the Simpsons episodes who killed Mr. Burns? do you remember how the school found oil underneath and then Mr. Burns built well, it tilted sideways that's what fracking is what fracking does is the oil they drilled pipe here and then it goes here not for a few meters but for a few kilometers and then they can just go through the entire geologic deposit all along its side and yank everything out before, these weren't economically to extract because of geologic formation, if I just drilled down and I've got oil, the oil's kind of captured under the salt dump this big pool and I can suck it out the fracking exists across the entire geologic formation what they do is they drill the pipe in they go this way and then they pump the water and they just break that geologic formation apart and then they gap, shake solution floats to the top and then the longer they can break it apart the more gas they can get this was not possible 20 years ago it's possible 10 years ago and that led to a resurgence of the oil industry in the United States they have to know those geologic formations they have to be able to predict where it is because if you drive your pipe into the wrong formation, you can break it there's an even bigger danger so a while back Harry wrote he was with Bruce Kaiser and they had a grad student measuring a core like this, every 10 centimeters and he's talking to Bruce they're talking about, oh yeah oh yeah, it's in his little chillies there's a heater in here what? yes, there was wine see, by the way I got a good comment I used to be super activist-y when I was in college during these and all that we always do plans for production but we never do nice consumption what we just did there is why we have to drill all that stuff, right? but the key thing is here that's the process behind it all so what drove all this change was that you don't want a frack in it but you open yourself to vulnerability so an example is Harry Rowe had a student he was talking to Bruce and the student is scanning every 10 centimeters like this with the tracer it was the same tracer you guys have right now and then the student takes Harry Rowe aside and says there's a gypsum from about 1 meter to 1.5 meters so Harry Rowe interrupts the conversation gets on the phone, calls him and says, I need to add a cement cap and the guy says, ok, Harry Rowe goes over to Bruce and says I just sent him $100 million what happens is what's gypsum made of? Does anyone know? calcium sulfate the sulfur if you drill a hole the gypsum is crumbling if it falls in to the deposit of oil and natural gas that sulfur will dissolve in and contaminate it and that sulfur is not prohibited by regulation because sulfur is the main pollutant for acid rain in the air so that higher deposit is worth $0 so $100 million worth of oil is now zero because of that contamination that's why they built this calibration one pipe would it go through? if I get a rock of gypsum this big to drop into a pot of oil like that's 38% sulfur all those sulfur atoms percolate out and if that sulfur goes to 0.1 yeah, that's a pollutant hazard that's going to create acid rain in damaged environments so where's the cement chopped in? it just blocks the gypsum you just coat it, right? I put like a little blocker in so that way it cannot fall in it's just like a retaining wall so as they drill they're pulling you down? exactly, then the gypsum can't come in so the oil companies funded open science no oil companies stepped in and said they're the only ones who know how to do this they do know, they pay Harry a lot of money now that they know how important this work is but they made this open for everyone to use because everyone had the same problem no one is invested in someone else burning the natural resource it's like the oil fields in Iraq by the fire all it does is add pollution with no economic effect so that's why they made this an open calibration so what they did was they drilled these cores in the formation iron, almost, right? like the entire set of history of life on earth is captured in one of these cores and what they did in each of these cores, you can guess they're a lot of money, a lot of money goes into Germany they take slices of each core horizon and they turn them into standards because each geologic event tells me something about the history of the earth and if I've got built the calibration for mud rock I've built an earth calibration for all the sediments now the reasons of our archeologists are ceramics made of the same place but what the point is, is they're exposed to more variation than we are so their calibration publishes a bigger elevation and this is a free calibration essentially because it's been published? I mean, you still have to understand it somehow you can talk to Bruce, by the way if you talk to Bruce and you ask and bag and actually what you can do is is you get your tracer, you fly up to Bruce and he will meet you when you calibrate on at his house and that's something he does as a favorite scientist because he likes people who are interested in having this incidentally we have access to a copper standard set that's great excellent wonderful yeah, you guys have this calibration in your units this is used for that way you can use this for ceramics I use it for geochemistry but we have a green, yellow did you do it through Bruce? yeah, good call Bruce will help you out if you ask very nicely I just really put in a good work for you because I love my Palestinian Greeks but yeah, that's the key thing is that calibration gives you the ability to do that so that I can share with you how we scan this so if you have a sample you can treat it the same way and run it through exactly the calibration but to get the calibrations it's not just like you enter the data from here you have to actually because you have to calibrate because you're training your tube to see that much oh, it's a tree okay, that makes sense so these are like cores, then? why does it actually be... so each of these dot points is a standard from a geologic horizon from one core that's what they are so if I have my core and I have a geologic horizon it's this big, it's a light gray something like that have you seen pictures of these cores? I've seen mining cores here, let me see if I can do this right and so do they just take a whole bunch of them from an area? yes and kind of average amount kind of thing let's see if this works guys I can show you an unbelievable map of what these cores from data that Harry's donated before he was bought if I kind of oil it we really have to lose this video, I'm talking too much but it's still recording I haven't talked to Harry in a couple of years because he's gotten a really crazy contract but it's got a very strong confidentiality requirement so I'm going to upload a spreadsheet drop box or no, desktop projects so while you're doing this as a clarification we basically there are different types of calculation we've been talking about the one that is the parameters is essentially it's not an object like elemental calibrations you basically are uploading a formula that you're telling how to interpret the data whereas these calibrations are training your XR, you're shooting at things and teaching it to recognize it in the future you've got it okay, that's it so yeah, you got it so like NAA those are constructed methods and so yeah, I've really dialed in and we can use XRF non-destructively so like if you're for example looking for a source of gold without the calibration you can shoot the gold or a diamond or a city and you can get all of the information out of it but you can't source it because you haven't sort of taught the XRF how to look for that whereas you can then go and be like here is the gold mine in say California, here's one in Mexico and then the difference is you're not training the XRF the XRF is only telling you what's in it you are training yourself in that case this is what this character is the silver gold ratio in this mine matches this object there you go, I can make an argument the problem with gold is that they re-milk it because it's valuable but yeah, in that case you got it this is an example of a core by the way this is a map from a tracer by the way they took a tracer and they measured that force, it's an elemental map of the core so this is what the XRF kind of see so you can imagine pretend for example this is a gypsum layer well you can see why tapping right there can save you all that money maybe this is the clay I don't want maybe this is the formation that has the oil and I can target it that way so that's how they use this for that it's basically that idea that's why this calibration was made it was made to interpret the history of the earth and then I can grab a different element that's iron there and then I can see with iron I've got these flecks of iron maybe the iron pyroite can hurt my equipment so I need to change my process and chop up the core so I don't damage my pyroite there's all kinds of ways to get this information and then I made it so you can flip your map and see what it is you can also add one of my cloud killed up files and it will calibrate this so I can see the weight percent of each pixel this is my favorite app no one uses it so what does it look like is it a raster no this is a vector it's like points well sorry this is a raster I raster the elemental data so it's coming in exactly exactly exactly you've got it anyway and I've written this for the tracer you can upload data to the tracer you can measure the grid in a specific way it can infer from the filename the coordinates of that so for example if you want to do a whole palace a mycenaean palace you could go over to mycenae and measure every quarter meter or whatever increment you want you create a map of that floor we did this in shallow oil for eating hotter we worked really well because we could actually measure we did it there so we could measure contamination that was damaging things you can also do it for living surfaces so we had the opportunity to meet with Bruce for that mudrock calibration but brooker provides one that's apart from that one or it's the same thing so the one you do with brook with Bruce has the 40 standards full the one brooker had at the time you purchased the unit was the one that's in this paper which is 26 standards that said brooker has bought the same one Bruce has it's the we call it mudrock 2 and so it doubles the number and it actually vastly increases the number of operation curves from a brook and then there's another paper 2015 husband speakman where they provided their team of brookers mudrock yes they do yeah they absolutely do I've got strong opinions about that paper and their argument but that's okay people can disagree so this is another one that you will probably want to look at just as background you can just type in speakman mudrock they're these kind of papers aren't really getting a lot of them are available you're going to have to have a lot of their signed in what is mudrock it's not made that's what it sounds like it's like a compressed wet mud that's in a geologic formation they dehydrate it before they make the pop so it's still a drain pop and they're still in those little discs so it's compressed pellet discs but it's compressed mudrock and the mudrock is how geologists refer to it there's sand, there's silt style it refers to some of the water properties of the stones so the ones that first had it were these big pucks that all stick together and these big little things and they're brown, red, yellow they're all different kinds of colors so let me go to desktop projects tracers like silk stone or different like what would geologists call it mud rock that's like the slang term it's not mud rock it's sort of like when you're talking about something it's a mud rock that's what it's named after but they're really funny but here's the calibration curve from that mud rock set so here's iron, cobalt nickel, copper zinc, gallium selenium, arsenic selenium so the reason why I like it for ceramics these are elements you can source with with ceramics, selenium and gallium so it captures those better than the soil because the soil calc doesn't have enough of these to make a calibration curve so because the mud rocks have all this they're selected for error length of variability it makes a good general calibration for silicates in my opinion if I make a soil compressive mud rock I still get very good results and they're reproducible terrible results as long as your sample is a compressed silicate that's made of mostly oxygen mud rock will do a good job of capturing its elementary ability that's the rationale behind it so that's why I like it the best people disagree with me but if you're trying to capture a lot of exotic elements that show up in these that can be sourceable this to me gives you the best platform you can paint worker to do this it'll cost you I think $4,000 for the two phase whereas I think we're planting to sell in the city in a couple days can also work too but that can be a way to say if I have a choice between really begging Bruce to measure his standards and getting the 5G or getting the 5I with the worker supplied mud rock I would probably go with the 5G plus asking Bruce really nicely that said if you ever need to get your machine repaired the worker will recalibrate it for you for the first time whereas you have to ask Bruce especially again I'm turning the excuse I got it two birds with one stone I would be reliable at the same time the mud rock looking at it to help you improve your video it's actually hitting a city level of good that's where we are by the way notice here look how high the Australian goes it goes up to 3% because I've actually seen this in ceramics in the sourcing project I'm working with University of Chicago Accommodated Persian tablets from SUSE have up to 1% straw table then so you actually have to have that much variation in these whereas a normal ceramics model wouldn't miss that but that right there is super sourceable right? like it's geologically weird that's the problem when you do find something sourceable the problem I see people have is that their calibration is too narrowly defined in that case they miss the exceptional ones that are sourceable so that's why you want to make sure you do it this way zirconium can I show you something too by the way I've learned this when I build my calibration curves do you see how these two are weird that's because the standard's got switched this one goes here that one goes there so in my calibration plots you can actually see when someone gets confused who puts one standard on before the other you can create this T shape with the validation slope you can draw a little square around it whenever you can do that between points the most likely explanation is that the standard's got swapped you'll see it on multiple calibration curves so I can see it here too that's the calibration curve so I wrote the algorithm to not do a linear regression but to do a loss curve if the loss curve is straight I know I've done very good slope and intercept corrections if it gets wavy I know I need a lot more work that's the reason I need to work on the calibration more although in this case I think it's because ICPMS is not great for niobium which is one of the reasons why they use NEA for that the problem I brought is they didn't do NEA they only did ICPMS so any elements that are bad with ICPMS like niobium aren't going to be as strong so what we're going to work on are going to be solved there's an issue it's the it's on the iterator and we use the yellow filter and this is with the yellow filter so the brimstone is contributing a lot more to it if you use the green filter it looks a lot better the gray and arched is that like a two standard deviation? it's a this is a 95% confidence band based on the e-regression model this is based on lows, this is based on linear regression but yeah it's a 95% confidence band I'm using did you plot step smooth method LN here and then here I just default to lows so yeah but anyway the reason I do that is because what I learned quickly in building cows is when you look at the cow if the calibration curve is straight you've got a really good cow if it's slopey like this that means I need to take my spectrum with a different parameter so that's something nice so I did that as a visual cue for me to know what I'm looking at but in any case the key point here is that when R2 is really good you should use it for these kinds of things and Harry Rowe did a phenomenal job of validating it so what they did to validate it is they took cores and they measured the cores using with a dispersive extract destructively with a black line and the WDXRF is the red look at that agreement with the handheld in the laboratory and so they did this because then they can scan it pretty quickly as they come out so anyway long story short mudrock is very well validated for quick horizons now what I'd like to do is we might do a late lunch today is that okay? I'd like to do a deep dive into what the elements mean because I think this will be useful to any of you who do anything other than do you guys anticipate using Qtracer for more than just obsidian I don't know if it's a what's that in this? let me give you an example of how that can work I think it is do you guys approve of this? I have too many things open so I'm taking that critical mass I lose everything you may use it in Mexico on anti-virgin surface layers as opposed to just geological surface layers on anti-virgin surface layers awesome this is from chocolate candy in New Mexico I'm published I really need to get out of this but here we took a profile of the Arroyo Bank by the way this right here is what we think is the drought during the end of the Civil War period that led to the abandonment of chocolate candy because we find the classic base surroundings right down here and they disappear and then you see the kind of flow lines here so here's the droplet you see all the lines dropped there yeah so what we looked at was calcium ratio this is a proxy for evaporation so in other words what happens is calcium and strontium behave similarly right? they both have two valence electrons because calcium is the right amount of strontium on the periodic table so two outer valence electrons are binding so they behave similarly and they're about the same size chemically which is why we take up so much strontium in our bones you've heard of strontium isotope sourcing in your intellectual work that's because our bones take up strontium because when our body digests something it doesn't say I want calcium it says I want an atom with two outer valence electrons and an ionic radius and what and you grab that strontium as well as calcium typically speaking if I'm doing ocean studies the ratio of calcium and strontium is easy even calcium is a proxy for temperature the warmer the temperature the less costly it is to incorporate strontium so if you've heard of C2 C3 and C4 plants or you know about carbon let's do carbon 14 carbon 14 dating is central in archaeology we all know how it works why does the plant take up carbon 14? well it's just carbon it reacts the same way but does it take up carbon 14 carbon 12? no which one weighs more? 14 14 which one costs more metabolically and moves through the system? 14 14 so there's a kinetic cost to incorporating carbon 14 we can actually mathematically model this for every atom of carbon 12 you take in for every... let me get this right so what it is is if you take up 10 atoms carbon 12 you will take up only 8 of carbon 14 because there's a cost to taking in the carbon 14 the same is true for strontium isotopes are all over the calcium the strontium weighs more so it costs more in a bit it's like you have salmon go up a river up a waterfall skinny salmon do better than fat salmon because they have less mass to move in other words gravity forces kinetic fractionation it's kind of complicated but the reason we have isotope ratios is because not all isotopes behave equally there's a difference in abundance there's also a difference in how we take them up and that's called fractionation when you start losing it well which one weighs more calcium and strontium? so if I have water evaporating is calcium or strontium more likely to go into the air? no, into the air calcium exactly right because calcium is like so atom for atom whatever my starting ratio is it will start to tilt and that's why you see things that happen with the lines as I have a water event there's more evaporation of calcium relative to strontium so strontium increases relative to calcium the more water is exiting the system you can use this to calculate the water that had to be there if you know what the parent calcium strontium ratio is by the way what happens here we call this the strontium bowl is that evaporation stops and that's the drought if there's no water the calcium strontium ratio is what it was originally when water is added you start losing that calcium relative to it so that's an example of geochemistry we can calculate if I can prove that all these sediments have the same background chemistry which fortunately in a place like Choco it's pretty easy because we know the geological formations we can actually model what is being lost from the system with the process of evaporation this is what we call a proxy we can kind of estimate the movement in the system based on the effect it has how about like erosion of stuff into the system you can use rubidium rubidium strontium is what people typically use because the rubidium is highly variable in rocks and the strontium seems to be confined be kind of maintained by benthic systems in other words plankton in a water system you can measure it that way if it's a regular system let's say it's a terrestrial sands are mixing in there's a lot of proxies for that but let's go over these systematically and talk about what it is that happens in this process so let's see I want to open up something city topic this is my favorite presentation that we were given by the way this topic is so fun to talk about alright so we're going to talk about the analysis I call it core analysis but it's really geochemical analysis and it all begins with physics so we want to put on our physics hats on first and then think about what chemistry does with those physics so here's the periodic table right the key to understanding how to interpret data from either sedimentological sequences cores or ceramics is understanding why the periodic table is the way it is so if you look at the periodic table and I ask you why is strontium taken up by bones then why do we incorporate strontium in our bones looking at it can you tell me based on the periodic table why that happens we're strontium so it behaves the same way let me ask you another question why is arsenic poisonous it's simple as AS what's it below does our body need phosphorus what happens when arsenic enters our body what does it try to do so do you know what your energy it's adenium triphosphate what happens is the way your body makes energies you got energy so when you eat food one of the products of all that food is adenium triphosphate that goes into your mitochondria and then your mitochondria split a phosphorus off so it becomes adenium diphosphate that process releases energy and that's the energy that power your muscle movements your thoughts all that if you put arsenic in your body guess what starts to happen guess what system starts to break down it's not a system you want to break down it's sort of the equivalent of pouring water into your gasoline tank it's liquid your car's going to use it like it was gasoline but it's not going to behave like gasoline and your car won't work great and that's the reason why it can kill you why is lead bad for us where is pb of periodic table two what's at the top of that column carbon what does lead do well it can start making several binds your body needs carbon so now it's not like adenium triphosphate it doesn't go into a critical part of your body that specializes with phosphorus does it just kind of ends up in random places the problem is your body can't get rid of it it just stays there it's sort of like I have a well-oiled machine and I put in a gear that's way too heavy so it doesn't move at the same speed so then things start breaking up and that's one of the things you see with lead poisoning Legionnaires disease you typically see people start to have four cognitive functions they tend to commit more crimes there's an argument, have you heard about the argument for why crime dropped in the 1990s unledged gas before in the 19th century what it was in 1976 gasoline had lots of lead in it they unledged it by legislation and so there's a generation of children who weren't sprayed at the face by lead every time a car passed by and when those kids got to be in the 20s they committed far fewer crimes and so the crime rate dropped and plummeted in New York they were like, oh this is the success of stomp and frisk policies but it isn't happening in every city not all cities have the same policies you know, Freakonomics, they talk about this and they attribute it to Roe v. Wade yeah, I disagree with them entirely I disagree with them entirely because you don't actually see that the worst was happening before Roe v. Wade they just weren't happening legally I don't like their argument because it's a correlation argument the lead argument is cosmic we know what lead does to the brand we know that people who eat lead paint are more likely to commit crimes there's an established positive relationship to it so I'm much more persuaded by that data and it's systematic everyone is affected by it we can't go and prove that every person who robbed the store did it because their mother didn't want to have them in 1973 we do know that a kid who was born in 1973 was exposed to magnitudes in the word for lead so that the lead caught that cause causes changes in risk perception you're still using loaded gas in India in 2000 when I was there so you can actually test that yeah, absolutely you can well, yeah, does all of it in America, is that the only reason? in Libya I had it it was for fuel ejection oh yeah there's a reason people didn't want to do it it was not a cost-free change so, yeah, what was it it was like lead-trying and methylene or something I can't remember yeah, you didn't need to do it there was an argument this is, you can look this up there's a scientist who worked for the oil industry at the time at the let he also did something else that was super dangerous he's argued to be the worst scientist in history or the most harmful of scientists in history because everything he worked on led to public health crises in his career and I think he died because this blanket caught fire and he worked alive because of the environmental failure there's something funny about that look it up, I don't want to misrepresent it but it's a funny, it's an apocryphal story is, yeah, one industry scientist has so much damage because like lead is yeah, like lead is bad stuff because it comes to you up basically arsenic takes over a key body process and kills you lead goes up your body processes in terms of endometrial trauma it just, it has these long-term effects on your cognition and your impulsiveness and all that because it doesn't hit a critical body function not just cheeseburg oh like, it depends on whether cheeseburg or made it, yes, but yeah but in any case, but what we're talking about here is the core to understanding the creative table do you remember the story about how Dmitry with Memblev predicted the properties of an element that didn't exist in his time he could predict how much of lead what it's a chemical properties were that was Jamaican, it's between silicon and lead carbon he understood how it could bind and all these properties before bliss discovery because he knew how to organize and the way to organize is one valence electron two valence electrons three valence electrons, four valence electrons these guys get heavy and form a new geometry of those and actually if you see how it drops in here like this by two it's actually the same here the true periodic table if we come over here the true periodic table I forgot the border of all this because that works really hard to do this so you see how this drops in by two exclude these because these are a separate branch put these two and then these guys form one table this forms a sub table and then you only see this broken out this is what happens this is what it really looks like it's another drop that's the same way and if we find a new row it's going to drop again if we keep creating synthetic elements that's the periodicity of it and that's what the table looks like so what you do is is each of these blocks you can compare then with this block there's a set of behaviors and then with this block there's a set of behaviors I hope that makes sense but that's kind of the big picture but anyway let's focus in on each class to understand what's going on class one elements by the way what I want to do is I want to make a horoscope and instead of giving you like Aries or Leo you're super volatile reactive but you're ready to form Carmen Bond but in any case that's what happens these all behave with each other by the way typically speaking when I run my machine learning stuff I'll find that rubidium can be predicted by potassium if I don't have 40k eb I find that sodium can be predicted by potassium plus rubidium if I don't have light elements you can actually predict the relationships here because of long term geologic processes where you find one typically find others so that's class one elements they've got one element from binding that's by the way like sodium do you want to throw a brick of sodium in your water laps to cool it down but it happens because that sodium wants to combine with every oxygen molecule in there and it wants to push up hydrogen so that's class one elements super reactive what number are they trying to get to because they don't remember from chemistry 8 so class two elements these have two valence electrons they also tend to be basic why are they basic? well because they don't have one valence electron by the way what does ph what does ph mean? what does this mean? I only give you a simple mnemonic device ph percent hydrogen that's free and available for binding determines how acidic it is acidic is very reactive hydrogen is very reactive basic has a different number of electrons and so it behaves a little different but yeah that's what we have here these all behave each other when you have one you have all so for example if I have a lot of barium I will often have a lot of strontium if I have a lot of strontium I will certainly have calcium calcium I simply have magnesium and they tend to form in groups and you can use the ratios of these with each other to understand kinetic events the same that you can do with the class one elements because if these all start off now what happens is that calcium is the most common all of the magnesium, all of the strontium all of the barium if I know that there are 100 atoms of calcium for every one of the strontium and that's my starting system deviations from that tell me if something happened to take away calcium and that's how I can infer heat or water or some other property of that material so that's what these guys are doing so you can, so it's accurate to say that they occur together just because they behave similarly exactly exactly exactly yeah that's right especially calcium strontium they have similar ionic radiuses barium is a bigger ionic radius or I think it's got a smaller ionic radius don't quote me on that I need to look it up the more protons you have the more the electrons cluster in closer so that ionic radius struts down it behaves a little differently magnesium behaves a little differently too because it's bigger but yeah, that's the idea behind it all so in any case the other one that behaves kind of similarly is beryllium do you remember what your XRF detector windows made of in the tracer 5I in this guy beryllium it's a beryllium window so you have an idea that you notice the properties from that information by the way beryllium is fatally toxic to 15% of humans and fine for the other 85% and no one knows why so if you're beryllium window vaporizes for some reason not an attendee you're probably going to be fine but anyway so that's the class 2 elements where's graphene in here that would be carbon class 4, now 4 so if I have 4 bonds I could put a carbon in each of those bonds and if I can cap it out that forms a perfect crystal that's one atom thick and that's what carbon behaves the way it does alright transition metals these guys as they gain mass they tend to lose reactivity but the properties tend to be shared on the columns so typically speaking even the rules we have about the classes follow the same rules here well I'm going to give you an example what is the most valuable element on this list traditionally if I'm a likening Greek which one of these is the most valuable gold right what's the next most valuable and then after that what do you notice they're all lined up have you ever noticed that before they all look similar to each other the property gold has that is so attractive to people it doesn't do great well but why silver attractive there's some silver it's very readily right what is it that gold and silver have in common to make people value them they're valuable yeah they're valuable other things are valuable what sorry that was my original absolutely you are absolutely right they're shiny remember yesterday with primates when we talked about our cone cells how all of our perception of art ultimately boils down to seeing red fruits and green trees why would shiny be important to human ancestors well we can see a lot of other things too but I mean it stands out but what about it stands out to something else that would be a stressor for our we needed fruit and trees why would we be attracted to shiny water what does water look like on the horizon when you're thirsty that glimmer right this is pure speculation but that's why all those elements in that column that's the property they have with visible light the visible light that we shine down from the sun reflects off them in a very specific glimmer this is invented as of two years ago in a laboratory but can you guess some of the properties before he came he was ready to act by the way do you know who this is named after he's like the already discovered x-rays very good anyway so the name with the type of gold you can only see with x-rays after him so anyway yeah that's the core of that column some of that more copper is at least to an ancient society more economically useful than gold but it's still kind of a precious metal because it still has that glimmer to visible light if we lived on a different planet where infrared substituted for visible light we might find that zinc and cadmium are more valuable because they shine just right these properties are based on our perceptions of nature but they're not unique to them it's just that configuration of atoms has that configured reaction of light by the way the other thing there's one other way that gold differs and copper differs from every other element on here what is it what color are they to visible light so what color is titanium silver what color is rhodium silver it's kind of metallic what color is pure cobalt what the oxides but the metal they're all grates they're all boring grates all of them without exception except because they react to visible light especially so that's the key behind it there's another spectrum, ultraviolet maybe iridium it goes crazy but the point is our eyes revolve for this range of light and that range of light interact with those guys in a special way and that's the basis of our economy for the entirety of written history that is our economy it is this shiny stuff in visible light but the core point that I want you to understand here is not necessarily that but for example this right here is probably more insightful do our bodies need zinc so what becomes poisonous so tell me do our bodies need iron yes is this good for you yes because what would it mimic iron is niobium that for you do you need this new diet do you need a vitamin with benedium in it no because your body can't use it the key point here is toxic elements it's our fault we're the ones who are so sensitive to zinc that we confuse cadmium with zinc and create that so that's the key but the core thing to remember is the columns are essential next column class 3 elements these elements have low reactivity and they're very stable so that's the key thing about these in geochemistry by the way when you do ratios this should guide your ratios these stable elements are good things to ratio to the only two here you will be able to ratio to at least with current XRF technology are these guys aluminum and gallium in paleochromatic studies aluminum is the most frequent denominator for elemental ratios aluminum so if I want to understand so for example let's go back to your head to think a lot of what's coming in well my sediment is going to have aluminum and silicon the ratio of silicon to aluminum could tell me changes in sedimentation source because the chance that the ratio of the aluminum and silicon in two different geologic horizons is not very good so that ratio can tell me that difference but I would do that because I know aluminum is stable it's not going to react with anything locally there is no chemical process in temperature or climate it's going to change the aluminum that I have so it's inert it's like gold basically in terms of metallurgy so aluminum is a good denominator for that can you predict which one of these elements is in all of your pop cans and beer cans if you don't believe me we can try it out what am I saying here let's come over here let's open up Rtax device connect does anyone have an aluminum backed iPhone I guess my computer but I'll just use my computer I know this is a little bit of a cut up so I'm going to do 40 kph I'm going to do 10 microamps I'm going to do the titanium filter rated at this for only 10 seconds it'll automatically turn that extra beam off we can disable that but that's enough to work with gallium right there that big peak what is the purpose of gallium in here do you want to open it no it came with what so the ratio of gallium to aluminum could be a sorcerer and by the way just to highlight this you don't read about gallium at all in the archaeological literature but the mud rock 2 set it's in the obsidian calibration it is but it's not a good curve that's terrible this one is a much better gallium curve it still has some problems they're L lines right K lines it fluoresces right amazing by the way the other thing you need to do a correction for gallium is lead so that can cause problems too but anyway that's gallium so it comes along for the ride with aluminum almost without exception I don't think we had an aluminum production that's capable of aluminum would you want to it depends on what purity you need I'm curing a silicon chip on it to make a computer chip that gallium could be important so if I have more gallium its normal properties might cause a little ripple that causes the crystal crack so it actually can become important to have these things pure especially the level of which we produce things so coming back here let's come to the class 4 elements this is the important stuff so with 4 valence electrons these guys are the most flexible in forming chemical bonds they're sort of happy with any circumstance so like that person who doesn't get stressed out in life everything works out great buying all these other things so by the way notice for example on earth what is the basis what is the element that unites all life on earth what is the most common what is the element here that unites almost all geology what is the element here that unites all bronze age metal do you notice the thing these form the basis for that by the way now you understand my tenders so important in bronze age because it forms such stable connections that it hardens the metal to make it usable and lead is used as a flux to help make things work too so each of these is super important what does that tell you about germanium do you think germanium is valuable so here's the thing it was one of the most valuable elements on the planet you can get right now for two reasons it behaves like silicon it makes a semiconductor but it operates at different temperature threshold so if I'm making a flying robot it will operate more stable if I have germanium in that chip for satellites it can be super important too second it's strategic metal so in other words it's important to natural security and third almost all of it is in china so if you can find it in the US by that land like I said I know one lithic source that has a lot of germanium in it but I also know where it comes from and I won't publish it or buy it because I don't want to hurt the stuff there but yeah it is super valuable it is useful stuff here but Dmitry Mendelev knew that when he made his pair out of table this was the only thing that was missing before it was discovered interesting arsenic is right in the spirit because arsenic is the other bronze ingredient absolutely and Peru too you actually see a lot of it in the old world but I think the new world understood why arsenic was important I don't think the old world understood why it was important because they had 10 and they could get covered to the chase but yeah good insight and Timony can do something similar too in modern metals but yeah, class 5 elements these guys really changed their reactivity based on their weight but same deal they read the reactive minus here do you guys remember the paper from 2010 that people were convinced that arsenic existed for phosphorus in life NASA had a press release everyone thought it was going to be aliens but life had an arsenic and what they found was that life they thought it was replacing phosphorous it turned out to not be the case they could just handle a little bit of arsenic in a system and not die but the reason is because its chemistry changed how it treats phosphorus ours can't do that so we don't do so well with one arsenic in our system by the way, arsenic I have eaten so much arsenic it's not even funny before I did archaeology I used to eat dinosaurs do you know how you find a dinosaur bone? what do you do? 16 times so I found like 40 dinosaurs in my hometown from outless Wyoming I liked a lot of rocks one of the sites is actually called LLT for lead like this after I did all this work and measuring and finding all these dinosaurs I grew up learning how to use x-ray fluorescence and I thought hey what's in all those dinosaur bones I was looking for there's lead, there's arsenic, there's uranium and there's thorium so lots of nasty stuff yeah but anyway arsenic you body can accommodate a little bit if you're exposed to too much of it on mythrodates famously did that um so yeah so in any case these guys have those properties so so yeah alright then the last group of elements class 6 elements these guys again become more and more reactive oxygen as you know is super reactive sulfur is reactive by the way for ceramics we were saying these two are super important in geopemple environments I'll talk about them a little more later but in circuit geochemistry well let's just jump over it right now if I have oxygen so there's lots of oxygen in ocean waters right? one of the big chemical transformations that happens is that oxygen leaves oh by the way sand machine where's my bag? over here zero just to keep track of all this so with these guys I have oxygen in the system and oxygen leaves the system metals that are soluble in water with oxygen volumes become insoluble with sulfur bubbles why would sulfur replace oxygen? same reason everything else right? it has a similar number of valence electrons and binds in a similar way so it happens is have you heard the statistic that there's a ton of uranium in the ocean because uranium is soluble in water if there's oxygen if uranium combines with sulfur it precipitates out and so if I see a sudden spike in uranium in my ocean waters that tells me there's no oxygen so that if you have an anoxic basin that formed in Egypt 200 million years ago when it was in the pan-classic ocean and then you had an anoxic event then you would see a spike in uranium if grease didn't have that anoxic event you would just see leadstone and so that would tell you aha I know what I'm looking at you can open up a geologic map and see where people look for oil because that will tend to follow with the uranium you might know how fracking is radioactive you might have read this in the news it's because of uranium precipitating out when you have these vents so in any case that's all these guys here it's also one of the reasons why we love you Mr. Aganski because our bodies are incorporated in our tissues we ingest it so it's not just radioactive it hurts you but you'll get stuck in your tissues and then you're not in a good way yeah the russian agent 2007 that's how you know that that was meant to send a message they wanted that to be an eclipse because they used that specific voice if you really want to take someone out what you use is you use potassium chloride and you inject it in their bloodstream and that equalizes the verbenium sodium potassium channels and they just die silently without pain and it's very hard to trace anyway, we'll talk about it so point is that Putin wanted that but anything you read in the news Putin killed so-and-so they wanted it yeah, that one can be really nasty too but then the nerve agent they used you notice how they always use something different to get it in the news there's a reason for that yeah, exactly anyway, so class 7 elements the halogens these guys are super, super, super react they're like the bizarre versions of the class what evidence by the way, fluorine can catch anything on fire anything, if you spray liquid hot fluorine on the ice the ice will catch fire because fluorine will outcompete the oxygen for the hydrogen that's not what it can do chlorine by the way, as we moved down reactivity slows a ton because reactivity is relative to the atomic weight of the material as well so chlorine it's still very reactive but not nearly as best as fluorine fluorine, same deal iodine, it actually can be beneficial for our bodies I read somewhere correctly if I'm wrong but the more iodine you have in your diet and your child, your IQ can actually increase because of changes in the sodium potassium channels in your body there's actually a proposal to intentionally put iodine in water like fluorine that sets a lot of people when you start going to water, but it's an essential salt that your body needs it's an essential salt because of identifficient areas like practically honest they really noticed that but using salt as a vehicle they would just attend to it another interesting finding is that this is actually relative to the western states there's more lithium that occurs naturally in the waters on the east coast in the west coast based on geology and lithium is an antidepressant when you actually look at suicide rates suicide rates per capita are higher in the west of the United States than the east and they attribute that to lithium and so there's a proposal to add lithium to water from California and Colorado and all that it's interesting but anyway, but long story short the reasons for why those elements are useful is because of the reactive properties right, iodine readily binds and that's one of the four things our ancestors for whatever reason blogged on to it as opposed to say bromine which is probably just a little too reactive by the way, bromine if you look closely at your XRF spectrum you almost always see this element it's not in the calibration, is it? I don't think they've picked it up at all but I think this could be used in sourcing for the obsidian too you see it in volcanic ashes as well quite frequently if you measure, next time you go to a restaurant bring the XRF not even kidding, I've done this if you measure the shells you will find bromine on the outside of shrimp and clams, but not on the inside because they used the bromine to prevent barnacles from lodging on hmm so let's say I'm a shrimp can a barnacle fall on my back? yeah, is that bad for me? yeah, it's going to weigh me down it costs me more to move so they've evolved resistance by depositing bromine and why is there bromine in the ocean? same reason there's chlorine in the ocean because they pave similarly to what they dissolve on the water so anyway, the other thing you can use is a bromine-chlorine ratio because salt can be added as temperature to ceramics but I don't think it's as useful if you have a true clay source though this ratio can be very useful by the way, the tracer kind of jeep can measure fluorine so we can jump down one level it's absolutely no use to you as an archaeologist but if you're a medical device maker or you're a silicon manufacturer then you're going to have to work finally, the noble gases do you know what they're called? noble gases they don't interact with all those proletary elements because these guys, their outer shell is built so from a geochemical perspective you can almost ignore them almost there's only one exception the nuclear weapons the insect is added to the development of the first nuclear weapons is based on Krypton so Krypton doesn't do anything it's just a noble gas it's very rare to find it but there was a couple an aunt and her nephew and they were refugees from Germany both for Jewish and they were in the United States and they were taking a hike in upstate New York and the younger nephew was interested in uranium now at the time the only isotope of uranium that they knew of was uranium 238 do you know what the half of uranium 238 is? 4.4 billion years it's not reactive enough to sustain any kind of uncontrolled nuclear reaction a different isotope might have a different reactivity anyway, the nephew was interested in finding uranium for experiments and chemistry and he thought he sure would have uranium but when he opened it up there was only varying inside and it didn't make any sense because there was no strontium there was no calcium, it was just varying and so he said I didn't understand what was going on at all and his aunt, who was a nuclear physicist was thinking about it she says you have to have strontium and calcium with barium but then it occurred to her when they opened the rock it could have been something else if you add krypton to barium you get to 92 which is uranium and so she deduced that there was a different form of uranium creating the barium and krypton and she said, you know what this means there is a reactive form of uranium that can be used to produce a nuclear weapon she knew that Niels Bohr was working with the German government to look into this so she wrote to Einstein and said I think there's a possibility Einstein wrote to FDIL by the rest is history but krypton, the absence the presence of krypton is what clued in people that there was a different form of uranium and the reason they didn't find it is because what happens if I open the rock where the uranium can get to barium and krypton where does the krypton go? it's gone immediately, right? I've only measured krypton once it was a bag of the German facility any normal gas is going to be gone which of these will you see all the time Tracer by the way one of these argon it's 0.98% of the air you breathe you can actually use argon sometimes in ceramics as a proxy for porosity if I turn on the vacuum in here and I measure my ceramic if I get argon it's coming from the sample so if you have little air pockets in the surface of the sample you can get argon too if you're using a vacuum so in that case you can actually use it to kind of get a porosity not a lot of people do that because it's tough to set up just right but it is a possibility that we can talk about the lanthanites the rare earth elements these guys, they're actually not that rare I don't know if you knew this the glue is just really hard to extract in this obsidian we have all of these guys present in my hand right now but they're in the part per million or part per billion range and they're evenly spread everywhere you rarely find a solid deposit of them but they always you tend to have and the last group of elements are the lanthanites these are the radioactive ones and there are you can always tell that because uranium is the really nasty radioactive one that's early on in the list these guys are all lethal even at small doses so that's a rough tale to make by the way I find it a little better if you can think the guy who writes XKCD he wrote a book on it called what if and it was scientifically accurate answers to ridiculous questions and one of the questions he got was if I had a building block about the size of a laptop computer of every element and I stacked the periodic table the way it is portrayed here what would happen his answer, the top rail would just float away hydrogen and helium are just gases it's going to sit down but this one's going to cause a small localized fire if I add this row underneath this row the whole room blows up and I die if I add this row the whole building blows up if I add this row the whole building still blows up but now it's slightly radioactive if I add this row half of the city is annihilated and a radioactive nuclear blast that pulverizes everything and don't add the last row but in any case that's the basic pattern of the periodic table what happens is these guys get cross-reacted and that's basically the properties so that's our lovely periodic table and as I mentioned earlier it really looks like this we just can't get a piece of paper long enough in this way by the way we can do the same for the transition mounts they come out and then unify it it's however you want to display it one of the most creative things I've ever seen is a radial display in the periodic table there's a lot of ways to do it now let's talk about what do these elements mean if you're measuring a ceramic or you're looking at an archaeological site geochemically what does it do and so what I try to do is use musical groups that people know of to illustrate how this works so putting in the blowfish is going to form your most common elements iron, calcium, silicon so these are the most common elements they're annoying they always sound good but they're always just kind of on the radio these you'll see everywhere they crowd out all the useful signals and actually what I usually try to do is I try to ignore these as much as I can when I actually look at the data there's a lot of analysis in that context so this is calcium counts per second and here I can see calcium damaging and actually I don't think I can show you all that data so we'll just stick to these alright so these are your normalizing elements right these are the ones you can use as a denominator so if I'm making elemental ratios to source where sediments come from these are what I want to use as the denominator your video so first up aluminum, bono, is the absolute best it's the most blase, normal, predictable always sounds the same element you can find it is not reactive in any way in any geologic context as a consequence it is the best thing to ratio to see changes in reactivity so if I normalize phosphorus to aluminum iron to aluminum potassium to aluminum I recommend using aluminum when you can't use any in all cases that's your best preference and if you want to get any sediment change or anything that is the best and by the way for ceramic sourcing it's not going to be the abundance of given elements it's going to be the ratios to each other that can tell you to screen groups but aluminum is your absolute best denominator for surroundings absolute no questions asked but what's the problem with aluminum if I'm measuring with a filter, can I see aluminum it goes away so if I'm measuring with a filter the next best substitute for aluminum is Revitium while Revitium is a very reactive element it is also a very stable oxide nothing happens to Revitium oxide points it's made so provided you have Revitium oxide it is going to be highly stated and it will behave like aluminum as a denominator so the second safest denominator is Revitium so if you do a light scan aluminum is what you should use as your denominator if you're doing a trace scan normalize to Revitium that will be best by the way there's no coincidence remember Bruce's role for identifying obsidian and all that there's a reason he's using Revitium Revitium is something that is produced reliably by volcanoes and it tends to dilute down because there's no other source of Revitium so that is Revitium he does so let me clarify for determining what type of rock it is he uses Revitium for determining what source of volcano it is he uses zirconium zirconium is highly resistant to chemical changes but it also varies randomly but I want you to be correct if I look at variation on earth from zirconium it's not that much of a type of rock chert has a lot of zirconium I can even find zirconium in calcium carbonate limestones it's common in just about everything in granites and belt spars everywhere Revitium specifically is higher in igneous rocks relative to all the other kinds so Revitium is a good tracer for a broad big picture of planetary geology zirconium is better for fingerprinting a source does that make sense? so if I'm trying to understand process Revitium if I'm trying to understand location zirconium this is a complicated lecture but it's really important to understand and think these properties if you have any questions or anything that's not clear please let me know especially if you just happen to do it I know I just want to write those elements down so I bring these elements up because they're so big in the spectrum they distract us sometimes that's the only reason I bring these up it's not that they're not ever important they definitely are I only use this to highlight that they crowd out the signals when you take your spectrum sometimes you want to take your spectrum to minimize so for example an example application of this though in geology it's in metals I had someone come up to me and they wanted to study phosphorus in iron alloys from medieval England and they're like what do I do? the iron every time I take the spectrum the iron is the only thing I see so I told them measure at 6 KEP because then I can't excite iron and then the only thing I see are small elements on the surface so I can manipulate the X-ray beam to isolate these elements out and that's why I bring them up here logically they do all kinds of different things but they just kind of crowd out the signal and then we have you two here at this I bring up because they're so stable so these are your denominators this is what you can reach out to I should not have gone that fast two three four don't touch anything so again, so Zirconium is good for location the last one in here that is useful is Titanium Titanium is actually potentially really useful in archaeology especially in Greece but not for ceramics for marbles so if I have an ocean what type of geologic formation does an ocean a shallow ocean leave behind? yeah exactly so limestone really marble is just limestone what happened is that chemically again it just got deep enough the heat and pressure made it amorphous but here's the thing if I say titanium in limestone where do the titanium come from? it's not from the limestone it's from the shells not from the sea shells so if I'm in the mid-atlantic ocean and I have a deposit of limestone that has titanium let's say it's today let's say we get in our marine diver we go to the bottom of the Atlantic Ocean and we pull out our tracer and shortly before we die due to nitrogen we analyze the bottom of the ocean to see titanium where would that titanium come from? the Saharan Desert what does the Saharan Desert do? it produces lots of sand that glows the wind where does that sand get blown? it gets taken by the westerly and glows over the Amazon if you get a chance watch the new Earth documentaries on how the Saharan is the fertilization source for the Amazon rainforest sand from the Saharan Desert gets blown by prevailing global winds and moves across the Atlantic Ocean and is deposited in the rain shadow of the Andes and that produces a spike of potassium, iron and all of these nutrients that plants need to grow because the rainforest destroys its own nutrient base and a rainforest 90% of the nutrients are in the biosphere very low in the soil its only source of recharge is the Saharan winds blowing over the Atlantic Ocean but a lot of that goes into the Atlantic Ocean but look in the Atlantic Ocean I see more titanium that tells me prevailing winds were over here does that make sense? if I'm a marble and I see titanium that tells me something about the climate it tells me about the wind over that ocean different marbles from different places so that's how we can use titanium to source if this left arm matches this right arm from the statue so we can use that we're at the Acropolis Museum and we're using this to piece together marble statues it's a super elegant thing we can also use it for prepellant climatic work so for example here's an example where they can use these stable elements like titanium and zirconium to tell a titanium story does titanium look like zirconium here? notice how it does the same thing do you know why it does that? where's zirconium on the periodic table? what's above it? titanium so that tells you a proxy ratio by the way in your sedimentological sequences titanium and zirconium will vary with each other but each different geologic zone will have a different starting ratio so if you have a ratio of zirconium to titanium that can tell you different sedimentation at our geological site by the way what other element would you predict would occur if you have lots of zirconium? this halfium is only found with zirconium in nature exclusively we've never found it outside of zirconium so the more zirconium you have the more half you have and if you can see the half even the extra signature which you can if there's enough zirconium that can also be used as a tracer by the way do you know why they call it a tracer? why? because you're teasing up these signals alright so that is titanium sorry I've got some plots in here that are like that are confined so I can't do it so heavy battle these are a diagenetic elements these are the elements that take a sound and distort it after the fact so whenever you see these elements in a geologic context this tells you about diagenesis does anyone here know what diagenesis means? in Greek is second birth it's chemical processes that happen after deposition so once I've deposited all my stuff if these elements come up they're the evidence of secondary chemical processes after the initial formation so a fossil is produced by diagenesis all these new things come and criminalize the bone after the fact it's the second birth of the bone diagenesis can be a huge pain for archeologists if I talk about carbon secondary carbon contamination that is a kind of diagenesis or second birth of that material because I've changed its elemental construction dolomite, which is also mine with marble is calcium carbonate limestone that has had a magnesium replace the calcium you know what that happens now because they have those two electrons that would be a diagenetic preventive so diagenesis is whenever we have secondary transfer we often say ceramics are permanently unalatable because the act of vitrifying the clays prevents diagenesis so the equivalent of ceramics is putting a content filter on your kid's smart phone so they can't listen to heavy metal would be basically the analogy that's what makes ceramics so nice because they're frozen in time by the vitrification process so copper we don't understand everything about copper it's now a question but when we have chemical things happen after a geological deposit happens copper tends to come in so when we see a copper ore in cypress that the micrones are using they're exploiting a diagenetically altered geologic area another element that is diagenmated is iron iron is often in clays but we see it diagenetically form most often in the form of pyrite which is actually surprising in the common we often see it as ghoul's goat when it's in our hands but if I'm just looking at a ceramic there's actually a surprising amount of iron pyrite that can exist there so one of the ways you can tease out pyrite is to do iron by sulfur because it's iron sulfide so you'll see changes in the iron over sulfur can tell you pyrite composition of that material the last one is arsenic a lot of arsenic shows up diagenetically can you guess specifically what arsenic is replacing phosphorus so whenever you have a lot of phosphorus some material arsenic comes why was there so much arsenic in the dinosaur bones I looked when I was a kid sorry you missed the story there's a whole thing behind it because what's in a bone what is bone chemically is arsenic phosphate so when arsenic comes up it replaces phosphorus and thus my arsenic is supposed to be a child so anyway backstory is when I was a kid I grew up with what dinosaurs were so I'd go outside and lick all these rocks they stuck to my tongue that's a dinosaur but that's arsenic's role in diagenetics so anytime we see that replacement happening it's a diagenetic process petrified wood would be a diagenetic process so now we're on the topic let me show you this research using instrument 1878 which Lucas now has and it is here it is so here's an example of diagenesis we found there's a site in Sweden where they have a mesolithic occupation and then a medieval occupation right up top and we found the bones are all commingles so there's no stratigraphic separation and then radiocarpentate the bones then tell which animals were in which habitat and set a pretty massive mess and then I extrareft one of the bones and I noticed it had a lot of uranium and itria and when we compared the radiocarpentates to the itrium and the uranium we found they split almost perfectly the more itrium you have the older the bone is and then we did a micro extrareft of the bones and you can see the uranium on the inside of these bones do you notice that the bones that are smaller from the mesolithic and the big bones don't have the itrium? those are cow bones domesticated cows and the mesolithic they didn't have those yet so here we can see at this archeological site they can use the extrareft the bone has itrium, mesolithic it doesn't medieval this weight appears to be a constant in nature by the way and cave deposits itrium corresponds to age too so you can use for bones you can use the rare earth content of the bones, via diagenesis second birth is basically what you're seeing here by the way, that's uranium these ones are radioactive the age of the bone increases the likelihood that it's gone through diagenesis and thus I can predict its age that way so you can use, this is a cheap way to do some sorting if you've got comable deposits so anyway, that is diagenesis copper and arsenic there are other diagenetic elements like uranium and itrium but it's a big field of diagenesis typically speaking, diagenesis is any replacement of elemental content in the material I'm showing the common indicators because these are rare but you will find others more often an example of how we can use that is here so in the Mediterranean at one point it's salt content it's hyper specific I should wait for our friend just a second any questions about this so far does any of this seem more useful than other parts? there's a ton of stuff I know it's about your specialty no it's all good we're going to look at the sediment again that Craig got with some of these ratios because I didn't do any ratios we're going to look at ratios we do get ceramic in a great basin there's no reason we want to analyze it typically we tell it apart visually it's all good to have one in the back it's non-destructive right we have our clouds on both the machines it doesn't stop on this other than our own perfect so I want to show you here the Mediterranean about five million years ago the Mediterranean dries up completely Europe and Africa are the same continent for sure at great times there's no Mediterranean ocean it's called the Messinian Solenity Crisis which is like the most I don't know why they chose that name it's so weird but you basically have a vast salt plain separated in Europe and Africa and animals migrated across it we actually have footprints that have migrated up to Greece the denial would have extended as far as cyclists because it would have been land it's a super weird period of time about 5.9 and 5.1 million years ago Messinian Solenity Crisis don't ask me to spell that but it's super interesting super bizarre and it leads to a lot of transfer of animals in wildlife the fact that there's continent footprints on Crete suggests that the practical area might examine comfortably into the Mediterranean but, diogenetically we can tell this with elements because when it happens by the way, we call it a sapropel when you see these events, these hypersolidity events but here you see it the copper goes off the charts because diogenetically the chemistry is weird and stressing the replacement what do you notice the normalizing to why would you do that in an ocean environment because it can't be that the only source of titanium in an ocean is sand glowing it the only way, there's no other way for it to get into it organisms do not take titanium to form bones or shells so if I've got a limestone titanium is my key and then copper tells me something funky happened and this pulsed in from the second source and that's diogenesis in a nutshell and here's another example of it where you can see individual diogenetic events if you're trying to study a site the diogenesis, what the copper tells you is don't read too much into it because something has happened after that position right, so like if I'm looking at this statement and I say something really interesting has happened here and then I look and I go oh shoot my copper to rubidium is skyrocketing here that tells me something's happened after the fact so I might not be reading anything about the formation of this environment so it tells you not to read in too much now some geologists who have different questions might be more interested in that but typically speaking, a spy from copper or arsenic will tell you something else has happened here after the fact we don't actually know much about its formation because there's too much alteration right, it's the same reason why we can't just so you know how we do C3, C4 isotope ratios in mammals, like humans you see diet, you'll never hear about that with phosphates and the reason is because diagenesis changes the carbon right, if I say diagenesis I can't infer I don't know if I'm measuring the diet or something that's happened after the fact by the way, the other reason when it happens and it's super interesting is it's absolutely useless for dinosaurs there were no C4 plants in the dinosaurs C4 plants evolved at the same time humans evolved they're very recently 7 million years old the two are connected they think the opening of C4 grass lines and once grasses learn to do C4 C4 photosynthesis they begin to outcompete forests in certain climates and that creates grassy expanses to the trees and that incentivizes bipedalism between forest pouches so when you do the C3, C4 ratio there's actually a deep link between why the C4 plants exist and why humans evolved to exploit those habitats anyway, which kind of puts agriculture amazing context in the big picture anyway can't look at that there we go alright, so common evidence rubidium and zirconium these guys can be a useful ratio for grain size which can identify past kinetic activity so in other words, what cinematologists use is the ratio of zirconium to rubidium to see how big the grains are the idea is to figure the grain of zirconium you tend to see so here's an example of that where they use, where is it this is the zirconium rubidium ratio and they're using this let's see if I remember correctly blue is the grain size measured by an orbital disorder and red is the ratio of zirconium over rubidium so this corresponds very closely with grain size in these seven columns so if you have a chemostratographic section and you run down it that will tell you grain size why is grain size important? it takes a faster flow faster the water flows the bigger the particle size that is transported so it tells you something about changes in the flow rate if you're looking at a river system again, these are very specific things but this is the kind of thing you can use by the way, this would not be useful in ceramics because what's happening in ceramics? they're petrified so if it's a high temperature we've changed the structure completely you get what the grain size used to be but it wouldn't tell you right now what it is hang on one second, I'm going to write down the title what's up there this relates directly I think to this less you're absolutely right they're going in your rubidium that was one of the questions that our the GI asked me was how can we get that grain size without sending it in to get sorted I didn't know the answer to it this is a good cue I think though, this is typically used relative are you measuring them as little packets of samples or are you measuring them in situ to see the chain packets of samples? can you measure in situ? sure in situ is going to be the most useful because then it would take a trend and measure each piece I saw I mean if they're doing the stratigraphic control each one's coming through a different layer you can still infer but it's not grinding them up though don't grind them I have ungrounded samples do you know why you don't want to grind them? do you know what happens? if I annotate the sample the smaller particle sizes fall to the bottom the bigger stays up so it starts to sort so if I measure an extra ref on the bottom of a sample cut the smaller particles relative to the big ones that's why in situ is better because during the original context there's no sorting so you can get a representative to measure the original system the less you do the better it is for this particular complexity so I would strongly recommend if you can try to measure in situ for this see how much I tell them can you pay for us to come out pay the bill, don't do it for free but this analysis will always be better when done in situ than with the bare samples so my strong recommendation we ran into this problem all the time at Brooker because people would take the sample they'd mill it and then put it in and when you mill it you agitate it and the small stuff falls to the bottom and so you over-represent the clays relative to the quartz because the quartz doesn't mill as well now can we get a milling system that will stop that? possibly it is really expensive and I haven't seen it if it exists it might be it's not my field of specialty I've had that all the time people would mill and then complain to us that the aluminum numbers would be high why? because the aluminum was in a clay and the clay filtered to the bottom and then we see that spike shoot up so in any case, less destructive is better for this for generally speaking if you can't do it it's not homogenizing properly well I mean they're homogenizing properly by the book but relative to the minerals they're looking at they're causing sorting so in other words, if I went to the field and measured I would have a representative distribution of where the lowest is whereas if I agitated that lowest is going to filter between the poor spaces to the bottom so it would help you to sub-sample your less based on grain size could so if you had a charatographic unit and you did this and you had three peaks and you figure out charatographically this one, this one, and this one then you can target those and then you can do your elemental racism see where that goes so what you're talking about there is chemo stratigraphy so in other words, we're doing stratigraphy so regularly stratigraphic analysis I use visible light to see differences in the layers chemo stratigraphy I use x-ray light to do the exact same thing and I might get a different result there might be differences in the x-rays that are invisible to you but it comes to particle size absolutely so would you have to have a volcanic based rock or soil to do this because it has this could be technological and it will have no circling or rubidium in it to if you don't believe me, take your trace that's why I just didn't associate those elements with volcanic rock you will notice that that is characteristic of volcanic so if you measure ceramic and you see that seesaw you've got ash and it should be tall if it's shorter and it's sort of like strontium is bigger and zirconium is bigger, that's a regular clay so you can actually, once you do this enough you will be able to eyeball the volcanic ash because their relationship to each other is unique but each element with the exception of yttrium is almost always in every spectrum of the ceramic or clay or soil there are those elements are into a great view exactly so yeah, no, this is a good I would strongly recommend this is why you keep your bulk walls clean people have your straight clean bulk wall and just zip up it with the XRF take some time you get tired, flies come out in the end try to get some shade but this data is the best it's so much better than collecting a small sample and that way you can choose your increment if you want to be really fine about it you can move 0.1 or 1mm up the column you have to do it by hand, which sucks how will you see where you are what I usually do is I mark on the nose of the tracer and then I move up by those increments we'll have a measuring tape also so we can validate oh you just press it flush you press it flush against the wall so I would go here go here and still go here yeah, and it's very this one doesn't have a camera but the tracer is by my doubts, yes and so that way you can visibly you can even put a scale bar in the camera on the window and use that too if you want to be super exact so yeah, absolutely yeah, but in any case but that's the point in your video it's super useful for grain size and it's pretty well substantiated in the sentiment of watchable studies it's a pretty reliable ratio let me make sure I'm not going to crack it like that yeah next ratio is Mike so is this Craig Skinner? no, no, Craig Brown okay I did 1MRG or gotcha anyway, so that's Craig's favorite this is my favorite I actually don't like to go back that much but Mike, this happened was pretty good it's not as good as what was the guy Pete Sessions, is that the early that was a little more cool though but anyway, calcium and strata long story short these two occur with each other the heavier calcium is lighter calcium tends to evaporate out strata does not so this is a pretty direct signature for water and terrestrial systems actually in marine too you'll see the same effect it's more of a clay base than a limestone base you'll see the same effect more strata invariably means hotter more strata let me clarify this there's actually two different ways to use this if I have a clay system where there's water calcium strontium ratio tells me evaporation if I have seashells so just seashells raw biotic life hasn't been changed that tells me water temperature the hotter it is the more strontium if I have a seashell from Norway and a seashell from the same species or a similar one from Colombia I can tell you which one comes from the whole water with the XRF based on the strontium and it's not a small difference it's a gigantic difference in the spectrum it's like a 10xL so calcium strontium in a terrestrial system tells you evaporation and a biotic system tells you temperature it can sometimes tell you diet and wellness the more that someone eats fruits the lower the strontium is in their bones where they're in alien the more that they eat leaves the more strontium they have because fruit has less strontium than leaves relative to the calcium so this can tell you frugivore this is polugivore if you're trying to do ecological reconstruction sometimes it can be used for human diet but it's not quite that exact yet but that is that ratio and here's an example of these guys they use strontium calcium and then this is oxygen isotopes and they say the same story which is amazing to me so you can almost use strontium calcium from the tracer as a cheap proxy for delta 18L if you have a clear condition where evaporation is the dominant feature so this right here is strontium calcium ratio that's the blue line the green line is delta 18L the oxygen isotope we use is a temperature for the proxy for temperature so you see how it changes the strontium calcium behaves identically to oxygen isotopes now here's the figure the same blue line for strontium calcium with the temperature in black isn't that beautiful so strontium calcium is a very good pelliclimactic signature and so in this case they're using this for kinetic reconstruction over millions of years but it's a very solid one if you've got a biotic system it's a temperature proxy if you have a terrestrial system it's an evaporation proxy so you guys, the evaporation aspect of this is the most important to you but if you do isolate out your shells your gastropod shells you could potentially control them chronologically they could potentially tell you temperature as oxygen isotopes are so if you need to do oxygen isotope analysis but you're on a budget this can work too alright, next titanium and iron I click too fast let's go back titanium, iron so these guys in marine systems now I say marine systems this is still important to you if you're looking at ceramics because where do those clays come from marine systems right, marine geology is important in understanding ceramics these guys can indicate terogeneous, in other words from Earth so like I said in the Sahara desert flows its dust over the Atlantic Ocean that is primarily a signature of titanium and iron and let me show you one of the most famous pili-clinic studies that uses this this is from the karyakul basin just off the coast in Venezuela here they've got barbed sediments what barbed means is that they've got annual deposition of calcium carbonate shells from living organisms they used an XRF, ran it up and down and they looked at iron and titanium and they see this waving pattern and what they found was that the titanium and iron increase when the ITCZ, in other words that's the band of storms across the equator that generates hurricanes when it moves south they end up with more rainfall in Venezuela and so there's more erosion of dirt and that dirt washes in and adds more titanium and iron into the system and that signature right there can tell them climate because if the ITCZ is north then there's less rainfall in the ocean the more rainfall is happening on land that's the relationship they see by the way, do you see this divot right here that is associated with the late Bronze Age collapse and that kind of training this would be the late Bronze Age hot period right there you also see the same thing happen in the little ice age where it drops and you can see the Holocene you can actually see the underdrives too the underdrives so basically the earth was recovering from the ice age and then it's like the earth well it was warming, it had a midlife crisis and it goes man, thank you so much for being on the ice age and it just freezes all of a sudden and you end up with this 1,000 year cold shark spat by the way, interestingly enough the underdrives 12,000 to 11,000 years ago and then it stabilizes right about 10,000 years ago but that's the underdrives and that shark drop of temperature basically went to a drought and the time you see this line drop that is a drought in South America because there was less rainfall happening so anyway, that's the system there and then we can see that this period, the Holocene this is the ultrothermal that we referred to when things were hot and dry oh this is cool, you guys fear of the ultrothermal in the Holocene, right? what was it like in the ultrothermal in California was it good or bad? bad why was it bad? all these water sources dried up you know why they dried up? because the storm system moved south so South America had a bounty but then all those storms weren't making it up to California and then it recovers down over here so here think about how humbling this is changes in the part per million of titanium in a bag in Venezuela tell us why the life was miserable in California for 4,000 years right? that's how powerful these elemental proxies are and if I'm studying, now that's cool for environmental reconstruction if I'm looking at ceramics the titanium in something that has a lot of calcium might be something relevant, right? so that's why these things are important think about how these plays form the more we can recognize when you look at the spectrum from this play from a Mycenaean you can look past the Mycenaeans is it Mycenaeans or Mycinians? Mycinians well, I mean in studying Mycinians you say Mycinians but in like the modern village today is Mycinians that's why I'm confused I talked to Greeks about it so Mycinians when you look at them when you look at their ceramic look past their culture and that will give you a clue as to what's going on but yeah, anyway there's your ultrothermal right there what year was this? I can send you the publication if you'd like I think this is like one of the most essential publications I would love to see this because people on the Great Basement Army are forever about how the ultrothermal affected these cultures this gives a pretty simple the thing is with ultrothermal is we don't see any ice sheets you only see it when you look at continental records like this by the way, just to get out of that like a soapbox here ice cores are what we use to define a little ice age, but this is from ice and grit terrestrial systems respond differently to these kinds of pressures that's what makes this type of work so important because here we see what's happening and what's affecting it and we can see the shift of that rain belt south and it's not hard to imagine why if the storm belt moves 500 miles to the south how that could lead to things being drier in North America because remember a lot of those rainstorms to get in North Atlantic Monsoon peel up the ITCZ this would have been the weakening of the Monsoon in the southwest so by the way ask me why is civilization it's sort of like you look at civilization in North America it kind of feels like it's 3,000 years behind everyone else like chocolate candy is a pretty good standing for the Neolithic why is it so late? I think because we basically lost 4,000 years it got too early, we couldn't do agriculture here and even if you wanted to, you couldn't do it it was too dry it's not until the mid-holocene and spits off enough storms to make rainfall predictable and out to do agriculture so this could be the constraint in the US southwest for why do we see such late development of crops if I had to guess I don't know for sure but I've often wondered that myself I think this is a pretty important paper I'm really looking for hog at all 2001 but I can send you a PDF if you want that's the one, hog at all titanium and iron? titanium and iron is what they looked at like a ratio? no, take it out there titanium was more instructive and the reason they could do it just normally is because it was just calcium carbonate the only need to normalize because the problem is they're surrounding the soil, there's so many different things all on top of each other the key with calcium carbonate like this is that the environment is so simple we don't have the ratio anymore we know exactly where the titanium comes from they have the benefit the ratios are how to work around problems basically so that's titanium here's another example of using titanium here's titanium in percent the same study here they can actually get a sense of aluminum variation as well that same so when we look at it even more closely we can see if that's the little icigs that they saw earlier and then this right here that's the late bronze age collapse as you can tell I have very strong opinions but you do see a significant climatic movement that's happening around that time alright so that's obviously the data from South America how does that relate specifically to the I'm assuming late bronze age collapse but from what I've seen the climate changes and you can't produce the same crops that have been shut down what's the connection between the ITCZ and what's happening in Greece I don't know for sure I know for sure that the late bronze age collapse is directly related to water temperature so you know how the Mediterranean is hot and humid in the summer summer that rains in the winter what what happens is in the summer the water warms so you have a very warm motion but the air is also warm does precipitation happen in warm air no and it's cold there in the winter the air chills so what happens that water falls out and importantly the water pools are slower which means for this short period of time the water is still warm but the air is cool so evaporation accelerates so it's like you take your coffee cup and put it in the freezer you can watch the evaporation increase so the evaporation increases and you have a localized rainfall that's the Mediterranean climate in a nutshell what happens if the water gets colder and we know the water gets colder at that time period I think the water gets colder and we have a lot of less hot airbursts coming up north along the periphery that's my idea but that's what if I had to ask you to mention basically I would say what's happening in Greece can be understood in global economies which would make sense to next step with this because in the winter you also get the Sahara rains yes exactly and it's that connection to the African climate that adds to it anyway so that's my idea you don't have to know how to publish me so next group of elements we're talking about redox elements I'm using Metallica, really heavy hard stuff this I think what I'm going to talk about here is what I think is absolutely essential to understanding the ceramic sourcing and archaeology if you're lucky enough to see this happen in a ceramic that is a sourceable ceramic what happens is in an ocean basin we start off with oxygenated water so there's lots of oxygen in the water fish are happy, everyone's happy and crucially crucially metals are soluble in oxygen so we'll live with them chromium all the elements we're going to talk about here are elements that float in water each of these reflect the oxygen that is present in the water when oxygen drops out these metals rain out so I'll end up with localized deposits of uranium, nickel, molybdenum, chromium vanadium I'll also see an increase in sulfur and that tells me this is anoxin it also leads to a very fine place so if I have an ocean basin well actually let me let me show you an example of this from the suit of reports core analysis so here's an example this is actual data from the core this is calcium carbonate and this is molybdenum and weight percent now what's happening what kind of rock would we call this if it's like 30% calcium limestone and limestone forms in the ocean so what's happening is I've got all this healthy oxygenated water there's all this pint and thriving and then the pint and dye and sink to the bottom of the ocean in the bottom of the ocean there's also oxygen so what ends up happening is other organisms can come and feast and let the trial matter the carbon leaves the system and the calcium shell is built up over eons and eons and that's what happens here and then oxygenated water I don't see metals like molybdenum at all then something happens all the calcium drops out of the system now plays are formed silicon, titanium, and iron and now molybdenum sky rockets up what's happening well if there's no oxygen the calcium carbonate falls down and nothing can survive to eat the carbon so the carbon stays in the environment the shells dissolve in the acidic waters so the calcium carbonate dissolves so calcium leaves the system and then clays are the only thing that can fill that void and so I end up with a fine-grained clay with lots of heavy metals and a lot of carbon so that an oil company comes along and they see look at all that molybdenum that means the carbon wasn't recycled to the ecosystem that is where I can get natural gas or oil or whatever carbon deposit and then the metals tell me if you see this in your ceramics that tells you you have an anoxic ocean basin that is defined in time and space just like a volcano so the specific ratios of those heavy metals can be used to source you don't always see them but if you do see them it's payday for ceramic sourcing because these are discrete so this is range of question do you normally then see high quality clays in close association with volcanic environments no no, it's two different processes two different processes this is an entirely sedimentological process we're talking about there's no igneous anything happening and if there is it's deep time it's not anything that's relevant to the ceramics we grab so we may think that the ceramics we grab are deep time but most of them are formed only in the past 500 or 1 billion years it's like the billion year stuff that you start to see that volcanic stuff play a bigger role so in any case and actually let's drill down a little bit I want to talk about exactly what it is we're seeing because this is it's very important I think for you to understand if you're going to do let's see if I can find the presentation I think this is it but I'm not, we'll see sure this is not one of my talks this is from someone else but let's take a look is this the same person that did that the one published yep is this the one? no he is a he's paying off the student loans right now anyway so let's stop here and talk a little bit about the neurology hold on to your head and your mind for a moment we'll go back to we'll move back to that idea of fresh I really need to focus on it so let me yeah here it is so here's what happens so I've got molybdenum and it's specifically molybdenum oxide and it's 2- now molybdenum oxide all that oxygen it's like having clouds attach to it it's floating it's and gliding in the ocean waters so our calcium molybdenum is falling along and then it falls into our oxygen base and there's no oxygen if there's no oxygen that water is hyperreactive so think about molybdenum think about uh molybdenum is a male baboon and these are the female baboons and the female baboons are happy with their molybdenum when they get to the analytic base there's many more attractive baboons in that water and so the oxygen that's organic rich so then the oxygen exits molybdenum we've got all these better ones so now the baboon has to settle with older less attractive baboon and that's the salt and that weighs him down and he sinks to the bottom of the ocean and now that molybdenum starts to accumulate as a heavy metal in those sediments because the oxygen leaves an anoxic system if any oxygen gets into that anoxic system it is leaving and going for something else as fast as it can so the molybdenum is stripped and the sulfur replaced it do you remember the periodic table? it's just down from oxygen so sulfur subs in for oxygen it sinks and precipitates out and that's what leads to my heavy metal deposit now what makes this important is we come over here to this talk these metals all do the same thing they precipitate out but they precipitate out oxygen is only what happens to the equation acidity other elements available all that constraints the redox reactions that occur so as a consequence no two anoxic basins look identical to each other they all look different so that is why it's important for clay sources because these basins produce clays that can be used in the ceramics and are used in ceramics you will see these sediments show up in the archaeological record you want to look for molybdenum and uranium those are your two most important elements molybdenum and uranium by the way let's see how good you guys know your tracer what filter would be the best for molybdenum and uranium very good, the green filter that's why you want to calibrate to the green filter this is why you don't want to take brookers calibration because brooker calibrates to the yellow filter for the tracer fiber when I re-randed the brookers we had green, yellow and these I've had all the way from you know 3 to you got it, 30 so molybdenum is the key molybdenum is the key see molybdenum, you're at the street I can't guarantee you'll never see it you do see it, you've got admittedly sourceable ceramics uranium is a good follow up nickel and copperonium are both very good too but they tend to precipitate out in the blue waters by the way, molybdenum what we're talking about here is how we can tell when photosynthesis is starting without any fossils because the problem I'm curious they don't form fossils very well at all we think they do sometimes, but we usually just confuse so how do we know when photosynthesis began well if I'm on early earth 3 billion years ago what is not in the water so molybdenum there's no oxygen in the water, very good if there's no oxygen in the water is there molybdenum in the water no, so I'll see molybdenum and the sediments on the ground molybdenum can't stay suspended in the water so if I look at the early history of earth so there's no oxygen there's itch no, it's just brackish water it's not really like gross swamp water so molybdenum can't be there it's just no oxygen so what happens is is perversely speaking, if I analyze a sediment from that early on in earth's history I actually see lots of molybdenum because it wasn't in the water if it's not in the water, it's still in the sand and the dirt, it's everywhere when you add water with oxygen the molybdenum floats up into it so in my sediment cork when I see molybdenum, that tells me there's no oxygen when molybdenum disappears that tells me there's oxygen first so in the earth if I have a formation let's say the entire earth was represented the floor here is the creation and the top is yesterday and I measure right about here and fill up the wall all of a sudden molybdenum shows up and what does that tell me? there's oxygen and what does the oxygen tell me? there's life, photosynthesis so anyway, that's the crux of it so like oxygen is spelled by planets it's simply cyanobacteria yeah, plants, we know them don't jump up until about 500 million years ago on the continent but photosynthesis, the process of taking the light, splitting it out carbon into the body, oxygen out that's 2 billion years ago 2.2 to be exact and we can date that because of the molybdenum that we see in the ocean in ocean stratigraphy so anyway, if you ever want to see if you think you have a moroc maybe 3 billion years old that's what you can tell that's a nice marker for how old the rock is but locally in small pockets you will get molybdenum showing up because so for example, by the ocean current and the ocean current changes direction that I might have a little pooled bag that does not have oxygen circulating in anymore and then it can turn brackish and when that happens it gets really nasty by the way, almost every time you read about a shipwreck I went to the Mediterranean, like the Carinia Wreck or the Ulugru Wreck those are because of this so I had this crazy trip where I went to Cyprus and to get to Cyprus I had to take a train to through the demilitarized zone through to the Carinia Shipwreck and when I looked at their coins on all of the bronze coins found in the wooden wreck they had uranium uranium was bonding with the copper in the coins, why? for every reason they knew they had a shipwreck in the anoxic environment the wood, the planes didn't decay and if you still look at pictures from those wrecks you almost always see the wood come up and then there's like a straight line that leaves them that's where the oxygen in the water is and so every shipwreck in the Mediterranean you find is because of these elements, that's what tells us that so what happens is is if I'm scuba diving in the ocean in any organic matter that's old like wood from a shipwreck the only reason I can see the wood is because the anoxia of the water prevented other organisms from consuming and recycling that carbon of the biosphere so yeah, absolutely if I were going to Mediterranean and if I had a probe that could tell me oxygen in the water, I could use that to predict shipwrecks because it doesn't mean that that's going to be where the shipwreck that's going to be where I can see the wood where it might still be there so you have shipwrecks everywhere, that's just the most visible exactly, by the way, can I tell you something hilarious the Institute of Nautical Archaeology who does a lot of the Mediterranean shipwreck work has two tracers, one in Kuala Station one in Bodur Turkey and one of the big problems they need a tracer for is to sort, if they find a serenity they want to know which shipwreck it comes from and I was like, you mean they're super stacked on each other and they're like, yeah they hit the same rock that happens all the time in the Mediterranean there's the same big shallow rock that they can't see and they scrape the hull against it and they drop on top of an existing shipwreck so in one case I had to sort bronze age copper from bronze from a Roman shipwreck and then brass from a medieval shipwreck all sandwiched on top of each other from the same rock but in this case the wood didn't preserve because it was ox of waters because the rock, the waves and all that agitated everything the really rare rats are when they fall on an anoxic zone but they've had coins in its rubies yeah, you can still see those but those coins would not be loaded with uranium what I measured it was super cool, it was a brass Hellenistic coin but it had like 0.25 white percent uranium I could have discovered this thing with a Geiger counter if I wanted to because the uranium in that water was so thin because it was precipitating out yeah so anyway, but that's why these elements can be important in ceramic sourcing because they create these hyperunique it's sort of like the first line from Lea Tolstoy's Anna Carina is all happy families are the same but each one's happy family isn't happy in its own way in the same sense, oxic marine deposits are all the same except for anoxic deposits which are all anoxic in their own way so the specific ratios of these can tell you about that by the way, oil companies use that to predict how much oil they'll give the ratios of these metals can help them predict how much carbon is present in those deposition sequences it's super cool, but it works out really well I mean, it's not a theoretical for them they make a lot of money they ride or die based on the validity of it like how long it was anoxic or something how anoxic the pH all that can tell them how much carbon can be sequestered in those deposits locally anyway, the chemistry is super cool and here's an example of how you can use it so here I do you notice what are they ratioing to? aluminum because it's not reactive so here, vanadium to aluminum, what do they see? boom, right there copper to aluminum nickel to aluminum to aluminum and that tells them the specific anoxic rates notice that the nickel and the molybdenum tell different stories what's the y-axis there? y-axis is depth, so this is a core so think about this as a set of a core the size of this wall and this is where we're going down here we can see these local zones of anoxia that means in the history of the ocean fish are happy, fish are happy zone of depth and then that zone of depth is going to have a lot of carbon so yeah, but the thing is if you have a ceramic source this clay might make for a better ceramic in this clay and you can even tell in the same pit where the clay is coming from possibly but yeah, that's the idea so that's why these elements are important they're very reactive, very responsive all the clays you have are coming from water environments and this is true, by the way salt water and fresh water both create these environments it's not in either war you can see these same processes happening in fresh water ecosystems so yeah, but anyway that's the key there alright, the roots these are your volcanic indicators it's one of your ash layers if you are trying to find an ash layer you can find with these so for example, let's say you're trying to find the Buzama eruption in a sediment sequence you can use these elements to find it and I know, because now I'm a grown-up scientist I actually am on a committee of a student and that student has found the Buzama Buzama eruption in sequences and that we were proving in a late bomb bill using the tracer using these elements in the ratios someone to college in Missoula had Buzama ash layer yeah, that was a heck of an eruption and it was 7700 cal years BP it's a great way yeah, I think so anyway, yeah, so that's the example you know about these ones here too these two are also indicative of volcanoes, they're not as useful for sourcing obsidian, which is why you don't use them but they can indicate volcanic activity so if you're trying to find ash and by the way, this can be important too do you know what they use for temper sometimes these remnants? volcanic ash same rules as obsidian you can source the ash one of my favorite examples of this this is from Bruce, so take it for what it's worth you know he saw the data but he saw it, he was with an archaeologist in Guatemala and they had ash ceramics from two villages and in one of those villages they saw ash and in the other ceramics, they didn't and so in that case they could identify the ash that was moving to this village being brought over not necessarily from the volcano itself but they discovered that this ash is indicative of this particular volcano so you can even source the components of the ceramic, if you get this right and by the way, the ash the what, this is key, this is important the ash will have more rubidium and more itrium so if you find ash and ceramic do not use the strontium do not use the zirconium use the rubidium, itrium, and niobium instead because these elements are not common in regular ceramics but they're very common they occur three to ten times more often in the volcanic material do not use strontium and zirconium no no, do not use strontium and zirconium if you're trying to do ash use rubidium, itrium, and niobium because there's lots of strontium and zirconium in regular ceramics that's all, it's not that if you have obsidian then strontium and zirconium are great but if I have a regular sediment maybe it's 50% ash 50% sediment the signature from the strontium and zirconium can be stronger from the sediment than it is in the ash but the ash will be stronger from the rubidium and itrium and niobium that is what we're right into that problem where sourcing artifacts in a great basin that are highly eroded and pitted they have ash or sediment and coming into this pitted it was really, you know, it's like 30 ppm above exactly right exactly right so you can't like, you can't so it's soil in the pit it's not the pitted exactly yeah, sediment is coming in it's like little lowest deposits there you just need to try to not scan the pit the other 5 lines would be appropriate for that you can shrink it the other option is drop strontium and zirconium and use these guys because those guys would be much more common in the obsidian than whatever sediment came in you do not find more than 10 ppm in a regular sediment unless it's a rare earth mine you don't see it it's pretty predictable but we are looking at the distribution of obsidian you can see if you just ignore strontium you get a good cluster and as soon as you interview strontium you can predict how wet it is exactly where strontium is in there yep yep, that's it so it's a solved problem but anyway, those are your volcanic elements and this is an example of where they can do here's where you can find when you do elements just on the earlier ratios you'll see those volcanic sections just spiking it's just finding the right combination of ratios to figure it out the really good one we found or much more than my student found is you add a trium niobium and zirconium and then divide by ravide and that spikes out volcanic ash of X in a core for example but they can be very thin horizons you need to measure very very consistently let's come over here these are your clay elements these all correlate with each other if you have a clay there is a set ratio for each of these these will be important to teasing out different clays for you the coolphanoxia may or may not happen this will always the ratio of these elements should be unique to specific clays and I say it shouldn't be, it's possible for two different clays to have similar ratios for a couple of these elements but not for all of them so do you ratio all of these to aluminum? I think that's a good starting point but maybe the titanium potassium signature is more unique for this clay so you might want to play with all of these though I would absolutely start with aluminum since that's the most inert because the problem is so aluminum is inert titanium is inert silicon is inert is iron inert? can it be diagenetic? can it be pyrite? can it have other ways of coming in? this potassium inert not necessarily, more biotic activity can lead to a different form of potassium present potassium can also be in an oxide or in some of clay molecule too so I would still I would divide ratio of these into different combos but what you could do is build an index of all these ratios and combine that index as your clay signature and then compare it to others and what people like to do is try to do clay sourcing like they do obsidian sourcing that is look for rare elements of rubidium and transgenarconium I don't think that's the best way I think you want to use the light elements because all of these you can measure at 15 keb and use these ratios and use all of them together to form a fingerprint for clay and then compare that one across the board that I think is a better strategy to be reflected in ratio of these guys so an example of that here this is monsoon activity in India here where you can see for climatic reconstruction the presence of clays can suggest a change in formation environment so if I measure a geologic horizon the clay change can tell me something about the river systems but for your purpose you're going off of the fact that it's highly unlikely that the aluminum silicon fasting aluminum, iron titanium and all of these ratios are identical in the same group that's basically the idea I think if I'm not mistaken I have an example of this from an archeological context so let's come over here to documents University of Chicago diagnostic box there we go aluminum silicon these are for different artifacts right here, just one thing do you see that artifact there it has a different ratio it's slope remember that one of the slopes was the important thing so aluminum and silicon will always correlate each other but each clay will have a different ratio between the three so this guy breaks out consistently in three different measures do you always want to measure molten time for the ceramic by the way I can measure this guy once three to five times right, front, back try to get as many as possible you've got it you want to do as many as possible if it's a lead glaze you're not if it's a lead glaze you're kind of doomed because that lead is a substantive chemical change if it's picking up you can get just a clay on the side but that lead is going to throw you off all the time but I can start the lead I can take a spectrum that doesn't activate the lead I'm trying to get it right that way do you notice things moving like the gravity doing that I've never seen that have you maybe with metals metals you definitely see let me show you with metals actually just how much can happen so let's come over here when I was in Taiwan I was in Taiwan last month and I saw something that shocked me they had a Mycenae they had a sword in their collection and so I asked them they also had M6 which is like the tracer but the spot size is 25 microns and it can map and so I asked them to make a map of the entire sword BMP Rooker that's it so this is the copper green is sorry this doesn't have the labels copper is green tin is red isn't that funky so when I measure this of the tracer 35% tin when I measure this 12% tin so there's more tin on the blades than there is in the center and I see this consistently when someone brings one of these Mycenae swords up to me and asks me if it's real or fake I take two measurements if the tin varies it's real and by the way I don't know if this is a product of bad manufacturing that the tin is sinking because it's heavier or if it's on purpose if it's a functional reason for doing this I've been debating with this with metallurgists in Europe and America I cannot get a consensus on it at all I want to write a paper on the bronze age that's about this very because if I have 35% here and 10% here that's technically not a bronze anymore it's a bronze in the sense that it's carbonate tin but it's not metallurgically bronze so I want to write a paper that's called I can't believe it's not bronze and go over all these hundreds of issues in bronzes but I see this all the time where you have strong variation across the board but is it usually like this or is it very different I don't know yet we've only been able to map one at 25 microns this isn't hit up one right now we're excavating a cemetery we have just some high ends of solar so I have a research project for you this is not your specialty but it's some well-hanging fruit take the XRF next time you're in Greece and measure like this and see if you see that tin copper variation I guarantee you will so when you're doing this you're literally going millimeter by millimeter across the entire area this is 25 microns this is close to a million you don't have to do that I would measure it diagnostically here here flip it over to the sail but you can see here the variation is huge by the way it's not just tin either it's also that lead tracks the tin so I think what's happening is it's sinking so when they pour the mold in like the dirt or something when they pour the alloy in the lead drops straight to the bottom and the copper tends to float to the top it starts to segregate I'm also making this as a a formal thing because we see the same relationship there's still more tin and lead on the other side of this too the bronze age so anyway but I was talking to the metallurgist in the University of Vienna by the way I traveled we were talking about this and he said you were wrong these are 10% tin 90% copper they could not afford to make them this way and I'm like I agree tin is coming from a very long distance away so what I told him was here's the test I can't do it but you guys can do it if you have access to this so whenever he does it first I will go up there with you and we'll have a great publication measure it like this there's more tin in here and it's not 10% tin let's say they're not making their swords right and there's a lot more tin there that's present we should be able to measure the volume by putting it in sand or water, probably sand put it in sand measure the volume and measure the weight if it is truly a bronze we should be able to calculate 90% copper, 10% tin if we know the volume too but if there's more tin if there's a difference in its composition it'll weigh more and we can test it out that way does that make sense? if you can give me a plane ticket I will absolutely agree but no if we can do that test we could identify whether or not this is intentional or not I think it's intentional the tin is on the cutting edge so I think they're making different alloy properties in the edge in the center but I could be wrong if you're a polymobile and you're at hey, hold on this could be corrosion maybe the tin is corroding more readily than the copper and that's what you're seeing because you're only measuring the surface exactly which is why the volume measurement is important because if we measure its volume we can actually calculate if this is an anomaly or not very easily if this is represented with the inside of the blade too then this will weigh more XORF is only a part of it we have got to have the other variables too yeah so reshaping or something maybe a solid property of tin would be better suited so let me tell you this the reason I halide this so the guy who says corrosion could be it but I've measured bronze all over the world in multiple cultures the only time I see this the only time when I measure modern bronzes I don't see this so either this is a time reaction in which case tin is a proxy for age which is a publishable paper or they're using tin in a different way which is a publishable paper like it doesn't matter what the cause is if this is a case that older bronze objects tin corrode on the surface and the reason the late bronze age shows this is because it's older than the medieval stuff using this as a chronological proxy would be huge, right? it would mean there's a chemical reaction that is a reliable indicator of chronology and we can do it non-destructively that's extraordinary if however I'm right and it's compositionally different these guys are not using the bronze the way we understand it today and that's important to note too either way it's testable but we need a lot of source so if you have those sorts you also have other, like we have mirrors and other bronze objects that you could test against for example if they're doing it if the bronze mirror shows up with a more uniform pattern whereas this shows up with a different pattern see that? he started the thing and I thought this is testing framework with XRF that's very good, you need to get it I know that Kim and Jesse were working on it I know they submitted it they did, okay so all the bronze came from is that what Kim was saying? they're probably from Cyprus the tin is probably more I don't know I don't do metal so much I think the tin is mostly Anatolia the tin is Cornwall England or Afghanistan there's no local tin sources all there's tin the Hittites have tin they imported it too give me a moment here can we find it? I have a pop on this and I can't remember its name it's going to derive the nuts I can't remember where I can do this top I'm just going to go all the way this is the Mediterranean stuff that I was doing that earlier oh it's an agricultural conference that's right this is about agriculture here are tin sources major and minor that's true Afghanistan is a modern finding it postates this publication because after the United States invaded scientists got a chance to live in it you saw the news that they found all these economically valuable resources in the mountains of Afghanistan it's not reflected here but it is there Germany Spain England Brittany France by the way I mentioned there's a tonne in the ocean maybe that inspires some really interesting work maybe that's what the trope is maybe that's why what is it, Enelius Enelius yeah that's a healthy station that's why I'm right there it's almost like you can see the bank this is Malaysia that would be Vietnam Cambodia Myanmar actually that's Myanmar and then you're coming up into Bangladesh and then into southern China I don't think it was this stuff though that's too far so I think realistically speaking they do have a ship wreck they have bronze eggs so we do know these guys are contributing part of it but this gets to why it's so weird to see all that tin in the sword because you don't want to be throwing that stuff around it's coming away that long it's a long way to transport it anyway it's super interesting super interesting I got to analyze the tin ingots from the Ulu Brune wreck by the way when I was in Turkey it was super cool they were heavy as old get out do you think the tin from Ulu Brune is also coming from England? they don't know where it's coming from they have no idea but I can tell you how they can find out there's a trace of bismuth in that tin and we know that that tin is not recycled so wherever of these has about 100 ppm bismuth that's where you're going to find that tin but yeah anyway that's the best I got I haven't been in any of these mines I don't know what's in them but I do know that in Turkey all these guys had to transport a huge distance away and it almost certainly wasn't rumble either this is a mining deposit this is where you're getting that flux of tin which would make sense with the amber deposits that we have as well because typically you don't just have one resource come in unless it's Saudi Arabia there's something else that they're bringing along for the ride a little bit more money anyway yeah but that's the tin paradox so I have a question to get back to ceramics yes so you were saying it's good to analyze a ceramic multiple times as many times as possible yes for this for this reason because so practically speaking when you do that holding the instrument on the object and taking multiple scans is easy but then when you combine them and you create an average it depends so let's say you analyze them and you discover there's a residue on the inside that has phosphorous itself or that's localized let's say you measure the surface and you find out there's some kind of slip that was added you throw that out and then you use the insert so this is what I'm this is important to look at the spectrum before you process stuff so that you know how all those things are connected to each other and I'm working on a project in Israel where they did it beautifully they measured in the name they had the ID they had the sample a code for the position and then a measurement number sequentiality so then that way I could parse that information and throw in all the slips or you could measure the slips separately I could source the slips separate from the clay because if we know that aluminum is only coming from the top 20 micrometers right there it's where you have a surface treatment so you have a ceramic and a lighter ceramic and then that creates a coating of paint on the surface it's still a clay but it's a surface treatment like if you see those acrylmopots from New Mexico that are just beautiful those are almost always slipped that's a slip route of the clay exactly all that kind of pottery from America you have the other slips like leg lasers are silver based well leg lasers are I mean historically yes and then by the way the reason you want the measurements are for this ceramic here I can see that it's different so there are multiple points exactly so you look right here this guy is NNO260 I see that they're clustering together and that they're separate from this so that tells me this is from a different source I need multiple measurements because there's noise I need to see what the median plus and that's what it's about I recommend cloning a lot together not just the means because you kind of get a better sense of what the distribution is when they're all together like this but I can see that the majority of these come from the same place and others that one and then these two are obviously outliers that don't tell you much of anything if I'm just taking one measurement and I saw that and I attribute it to a different source because all the other measurements are up here so that blue dot is along with the other blue dots exactly you can see the axis right there and you can see there it's the slope not the actual value the slope of the ratio that's the most important which is why you want more than one measurement because that will tell you the slope a measurement of one does have a slope so that's the analysis you want to do for ceramic sourcing but for all of those look cute you want to discode yourself because then you're taking out the surface treatment if the clay that's important to you just put the clay in this project we're looking at well this project we're looking at from tablets so we knew it wasn't split but absolutely but you can also do the same thing where you come and take like four or five along the clay break and then like four or five in the slip itself you got it exactly right you got the picture so it's a lot of time for something like that don't use any filters or any of the calibration is first it's just good to just look at the spectrum and once you've done that you can apply the calibration you don't have to go back and shoot again with the calibration that's correct so let's say you measure all these and you measure them with the chroma and you're like oh shoot that's not how I calibrate it you just calibrate after the process the day if you want to but you can't post a filter it's like a photograph I can edit the photograph but I can't go back in time to change what the photo was exactly right yep you got it the ceramic sort of thing this is why I disagree with Robert's treatment on his criticism of ceramic because he's trying to do it like obsidian and I agree with him I wouldn't I don't think obsidian is the best framework this is very insightful I think you can still find clusters that are different do I think I can find a pit where this comes from that's different I don't know that I could do that but I can still tell that it's from somewhere different than any other pieces and if I have enough contextual information I can piece out what's brought in and what's not which is typically really what ceramic sourcing is about is figuring out local production versus imports so a paper that he's written the one that I just showed you earlier where he's critiquing the reliability of the broker mud rock standards versus the ones that he's created he's creating that those new standards to better more similarly mimic obsidian sourcing exactly and that's not the right framework I don't think that obsidian is strong in terms of coding and PPM is going to tell you anything about the sourcing of ceramic maybe in some extreme cases so for example the one time I would change that opinion is here no sorry that's not it oh darn it's not on here that's okay it's here do you see these blue ones break out and the green one that's susa the reason this is happening is because the strong team is 0.1 to 8% and the strong team is by the itria it's calibration so it's black firing calibration but that's still something that you can pick up with the mud rock yes the new mud rock the new mud rock would not have this problem the old mud rock has the problem but yeah in any case but obviously these I can source they're substantially different because it's strong healing blue and green are different or blue and green are different from that pink color so the blue and green are all susa because these are just the artifact numbers but I know from the data set this is all susa susa is like the southern iron second oldest city we don't tell them that they're very sensitive but yeah they heal them like no after I mean true true city not like got the right amount of agriculturalist like organized central redistributive economic church center I think oh what is the first first is in southern Iraq it's not her first third no it's not blue group or something like that Heruk Heruk yeah that's it we got that look it's Samaritan it's like old dirt yeah but anyway but yeah like I agree with them like here's the niobium I don't see much of anything happening with niobium at all right with the aluminum silicon which I would never use for obsidian because what's the point actually does break out groups but it doesn't break them up clusters it breaks them out in different slopes and that's what you want to use it's not as easy and we need to do a lot more legwork and this can be insightful anyway does that make sense there is here all right so continuing on if I have a couple really good buddies that would be iron and manganese the ratio of manganese to iron tells you how much oxygen is in the water manganese mn2 plus precipitates out of oxygenated water so it's the anti-molefin so the more manganese relative to iron or really the more manganese relative to titanium or aluminum the only way to do manganese iron is because they're next to each other and they fluoresce but that will tell you the oxygenation of the water so it's the opposite signal so if you don't have a read-up situation you can use this ratio to tell you the relative oxygen amount in the water and here's an example of this so here we have the manganese the manganese and iron ratio the manganese iron ratio pops up the more life is in the water column do you see that black dead zone that's when I have less manganese relative to iron so the manganese iron ratio is very much a good indicator for oxygen there are other things that can cause magnesium to get there I don't want to mislead you but if you have to bet make a bet manganese relative to iron is a very good prophecy for oxygen so now crap the air terrible versions and this would tell me pyrite and low oxygen conditions so iron and sulfur together so manganese iron is oxygen iron to sulfur iron and sulfur is low oxygen and that would be the formation of pyrite now if you may say I don't have fulls golden in my Xenia ceramic you would be wrong there might be little flecks of pyrite but nonetheless still in dig of the rapid ocean process would you get it from maybe even otter you would build flecks of pyrite so iron and sulfur that tells you right away those ceramics might be diagnosed with theoterep even if you can't see the flux by the way there's a source of ceramic that has a tone of anadian V it's an element of anadian I don't know where it is I don't know what's up but it's super unique on it but pyrite ceramics can sometimes show up as being highly different too pyrite phosphorus indicates nutrient availability so the more phosphorus you see the more that indicates bionic white at present that is just as true for ceramics as well sometimes that phosphorus is preserved in the heating side of all so with that play as present same deal in archeological sites obviously what do we use to identify human activations of phosphorus same rules phosphorus actually really indicates any life as present it's just humans concentrate in a place like deer don't and so they're a little bit unique spatially compared to other mammals on North America what are the critiques for using phosphorus in the Maya area for intensive activity areas as they haven't scanned all different activity areas to understand the variation in phosphorus that's fair, that's fair that's fair, so it's very fair if markets are into high phosphorus areas then living spaces shouldn't have high phosphorus because they're not as active as markets but the others are really dead in it that's fair, that's fair no, definitely worth looking at I've been very sold on the phosphorus thing from the data I've looked at in general but I couldn't get to the specifics of habitation versus marketer this activity with that activity that's one of the things that's hard to rate oh, very good good luck I hope that works out and I hope you get an XRF too to do that because the resolution of the XRF is just fantastic let me here, let me show you what I'm talking about it is very low but when you have specific activity it's high so for example this is from a room block from Turkey the east and the north and I can see highly localized phosphorus deposits and the phosphorus as you know, mineralogically is not common, it's very low that's why we have to use fertilizer to grow plants because it's so low but when you have human activity, cleaning and all that phosphorus tends to deposit like hyper-locally, like in a little 30 square centimeter area as where you can see those pockets in this case we map with the tracer very slowly but that gives you an idea of how it works so what what's our scale there we did it every 40 centimeters maybe like pink to green this is met counts in our taps so this isn't quantitative so this would be the photon intensity so we went from zero counts to 900 counts per second it is, it's not nothing it's quite a bit we did the same thing for chlorine by the way to see where the salt was coming and here's a punctuated deposit and here's where it filtered out have I ever told you this story about chlorine in Chateauville so when I was in Chateauville they had this problem where salt was causing some erosion in the walls and like it was starting to degrade and crumble and all that so basically they wanted to use the tracer to figure out where it was coming from the thought process is that there was like some ground water with salt that was coming up causing the problem and I was with this brewing gel from Cardiff I can't remember her name and we were analyzing we went over it and we analyzed both the wall or the ground that has all the salt and what we found was that there was a silicon peak and a chlorine peak so we could see the salt just fine sodium would have been better but chlorine is still pretty indicative so I saw silicon and chlorine and it was a sediment just dirt right, clay and dirt so we measured that I can't see silicon I only see chlorine is that possible? no, it's an elaborate wall, of course we're siliconing it why would I only see chlorine and not silicon? what would possibly cause silicon not to show up in the spectrum if it was covered up, right? if something was on top of the silicon that told me the salt was in higher concentrations than the one brick wall it was layered on top than it was in the ground that's not consistent with salt water that's consistent with someone pouring salt water on top of it so that was like really confusing so we could rule out round water right there so then I said we need a control measure because if it's ground water it should be outside the site too so we walked outside because Ian Otter built these buildings around each of the sites we walked outside, we analyzed the ground no chlorine whatsoever again we measured every component that they used to manufacture the shelters and in the far right corner we found a vinyl window shade a PVC polyvinyl chloride what was happening is in the winter snow would build up outside this window, it would melt water would come through the polyvinyl chloride pull out chlorine anions and then wash it over the site and then that would react to sodium oxide that was naturally soil and then you have salt the art concert got a conservation student for this and she went off to tell Ian Otter and I went back to use the restroom and then I realized oh I forgot my filter box over at the building I got it over there, there was a construction crew already taking the building components down Ian Otter was like that I was wanting to do it I was shocked but that's what you see here this is where the chlorine was dripping in on the floor and you can see it wash over anyway, by the way don't use tarps if you can avoid archeology because if they have PVC and you get a couple winters you're actually dropping a ton of chlorine at the site and chemically altering it don't do that test it measure the tarp to see if it's PVC or not if it's not PVC, you're fine if it is PVC, don't use it I can't strongly recommend this enough it has long-term consequences we need to get it on something it's like an overnight thing you just use like tarp for shade when you put it this way it's going to, there's one winter that is going to pass don't use it because then the water can have to do it this actually upset me because I was at a shopping canyon they used PVC pipes to create a water system and so they're actually creating localized damage there with that system which is really frustrating but yeah, don't use PVC in archeological sense this is what is it is not enough can we take a break? yes, let's take a break it's already 2 o'clock, I'm sorry I warned you, I'll keep talking you weren't kidding, how do you eat do you eat? he lives on in Japan I do, I do, that's all I do he's got a two-year-animalizer he's got everything he needs two years ago there was a class in your computer yeah, well it's definitely not long yet to take a lunch break I don't want to turn it through we can just hang here until 3.30 and then exit out and call it a day and I actually wanted to there's a talk at 3 that I was going to video before okay, oh gotcha yeah, that's fine I'm almost to the end of this geochemistry thing anyway is it helpful at all? I don't know this is very good yeah, I've heard about this it's always the past an explanation of what's going on logistics, and they incorporate the logistics of doing this too that's why I was asking, okay if you're going to take 15 measurements what do you do with those 15 measurements after yeah, exactly I've seen people, when I was at University of Florida working in their curation there were people coming over and analyzing the ceramics from floor archeological sites and they had a tracer and they were scanning and I said, well what do you do after this this is before I started really learning about it like, oh we just average all of our signatures like, well if you don't ever scan the inside of the vessel at a cross-section you only have the exterior there's something that you're fundamentally missing from the variation and they just kind of shift their head and they're like, no we know what we're doing and I said, that's fair you're not very critique, but if I was going to create an average of anything I would want to analyze every side of it not just a body inside outside even if the vessel isn't split exactly right because the surface changes different than the middle changes in the firing process you got it so I recommend I like to measure like a cube six measurements all around as best as I can if I can't, I can't but yeah, that would be my recommendation like, forward like a cube treat the ceramic with a cube and measure all of it so that would be the best I think so those kind of things are arguably just logistically because I've read about ceramic sourcing but I've never read about anybody saying this is where we took them this is how long we scan them it just gives you the geocubistry and says, here's our clusters you're missing telling me how you actually pick it up and do something that's not included in the writing and also the inclusions are different you can't reject pretty big differences that's actually one of the reasons why XRF can be a negative that's why you weren't going to use EDS for smaller shards to measure that because then you can target one specific area we actually had a little workshop on the EDS component it was partial because it wasn't calibrated yet it couldn't really show us it has a fire wire plug so you could get us another XP last hop that's all of these machines but anyway, they were warning us that you can zoom in to whatever, 20,000 acts or something because it's an SCM but the EDS isn't that it can't go really that far it's pretty important but you actually it's much more pinpoint than XRF not as much as the optical or the SCM that makes sense yeah, that makes sense yeah, makes sense, makes sense forget about these so if I could ask them how are their workshops are they entertaining or are they fun I was like I was like moving their models in gotcha, so real quick we've got to close at 3 o'clock at 3.30 but you need to videotape something oh yeah, I have to take the channel yeah, so gotcha, that works so what we'll do is we'll aim to close around 3 and then, because I know you guys are hungry maybe do an early dinner if you're okay with that or like a mixed dinner lunch I thought that works for you that works nice yeah, no, good call I'm good, I'm good I'm good I usually don't eat lunch though but you guys saw I eat a phenomenal amount of dinner that's my meal of the day go ahead and grab some snacks sorry about this good book thing I didn't realize it was already 2 o'clock I'm an hour I've had some really embarrassing times talking and talking and people are just too shy anyway I thought it was going to be new yeah, that works it's very information dense this part and it's very advanced we weren't kidding when we said that that was a more advanced discussion I kind of feel like when the information density hits a certain point your brain just starts to slow down and if you want, I can create an edited version of this presentation and give it to you if that's helpful because I need to take out all the stuff that is relevant to an idea I don't know why they have this idea it's not like I guess it is in military base but anyway, it would be helpful I mean, I know we have video of it but no, how many to review and all that but yeah, the key takeaway Oak are getting at is that understanding the context of formation for the plague is the central understanding of the ability to source not that I disagree with speaking too much I really agree on the dynamics of things but I do think he's unduly critical of ceramic sourcing because using obsidian must stop the sourcing I think there's a lot you can do but I think the difference is you need to accept that it's not going to be sourcing to a specific composite so much as clearly being able to divide out and identify these areas and I think in your case since transport across an ocean it's a lot easier I think if you're looking at a cluster of sites outside of Chicago then it's a lot tougher but if you're dealing with two disparate cultures separated by a sea then it's an easier job though I also think there's a lot of other interesting questions that you can ask with the x-ray I think the I cannot believe it's not bronze aspect actually if I can give advice to someone who's going through grad school someone who's going through grad school and school was all I lived and breathed I was just fascinated with it I only did this dumb x-ray because a friend asked me to he really needed help and that ended up giving me a whole career you don't know what's going to be important and so don't box yourself in too much explore a lot of different things in grad school it is your time to explore as Lucas and Kathy and Nico can tell you once you're done through that process and you're focused on something like your kind of boxed in I'm doing this Monday, Tuesday and tomorrow I'm going to rehearse to do some photogrammetry on their Egyptology collection to create a 3D printed photo or collection of things that they can use in the exhibit it's coming out nice that's a great week to have all this happen well then tonight is the main night so I have to go and quick run you can actually have your rat so you can come and see pictures of a bronze so every year we do it's mostly age for the donors it's the main night we have a little reception beforehand so you can have some nice the main lines and then we have basically an hour-long presentation where my advisor walks through all the stuff we did in the last season and so she was on sabbatical last year so this is actually two years of exhibition which includes one big two one of four we have a lot of bronzes specifically we have Saigon it's just to the west of Numea it's just north of Marseille Marseille's had a long community project so Berkeley has owned the rights for Numea for a long time it's something like the 50s Steve Miller started there so we work there as well we haven't been digging there we live there too so she's basically been working on my CD for like over 15 years something like that and that's a ceramic production set which is the reason why I'm interested because the ceramics that are being produced there are actually from that specific workshop we're showing up in New Jersey which is interesting but then just in the last four or five years we started working at Agonia which is very close to Numea all of our materials from the museum in Numea we're kind of, but it's a bronze age chamber to the cemetery we've done we've got to, we've done four tunes now two of one has been almost completely looted one has been partially looted and the two of them have not been looted at all and those two we've got a lot of bronze materials so we've got a lot of gold and other stuff like that so if you are free you want to come tonight and I'm flight leave at seven so I'm guessing it's not going to help but I wouldn't know what to talk about that's alright but anyway before we're talking about the my brain is sort of sorry Mike bronze it to me the key point is do the bronze thing too take it and own it I will tell you right now if you build the equipment you own it and you build this because the one that you showed is that the only sword that we've done that on only mine because it was a human museum so it was part of someone's art collection so they're negative that very well can be looted it's a billionaire who donated to an art museum I knew what it was and I thought we've got this there's like ten of these that exist on earth we have the interest let's do it the rest of them but I'll tell you right now if you send those to Taiwan they would be interesting because the gal who did this is interested in that archaeology that's published on it she's a scientist in Taiwan but her specialization is our conservation she's just got an academic interest of us so if there is an ability to do this analysis you do have a partner I can hook you up with to do this fine level work we just they can't leave the country they're able to leave the museum that's fair is it possible in a circus it's very possible but if it's affecting my measurement there it's affecting my measurement the entire time I've ever looked at any of them on stage and I just have a hard time believing but that's what's happening the only one because of the equipment it's the only one I've looked at with a quarter million dollar microXRF I've looked at probably 50 of these things with a handheld remember that sparta site in the news Lake Bronze Age yeah so I was invited by the Greek Ministry of Culture to look at that sort of cluster it's like the size of half this table and it's just like 30 of the things all scuffed together and I just sequentially boom boom boom boom boom but I showed them that 20 30% tin that was on those guys too just on the edges I can only mention the blade on that one because most of them have blades and stuff and they went on top I was able to show the difference where the middle had lower tin now again maybe this is an 8th dependent corrosion thing in which case maybe there's a way to date these things based on the corrosion of tin I don't know but I suspect my instinct is is that this is related to corrosion because I don't see the same thing in Asia when I look at Asian bronzes but it is or is not related to corrosion I don't think it's related to corrosion I hope it is I hope that criticism is right I don't think it's corrosion but again that's why we need to test them the thing is because we can only measure the surface we need to be able to test the weight if it really is 30 to 10% tin that means its average is around 20% which means it should be measurably heavier for its volume more dense than it would be if it was just copper so we need to use Archimedes on the assist to figure it out because again the extra Archimedes surface method or we drill for the center which I do not want to do especially not to the extent that it's just heterogeneity so anyway that'd be fun if for whatever reason if you can work it out if you can get me a plane ticket I can bring this and figure that out and then I'd be just toxic to someone or we'll figure it out but that'd be fun is that a way to rent these? not really I guess I can rent this one but I wouldn't feel right doing it but anyway typically speaking no one rents them because the risk is too high like $50,000 imagine I was going to say rent a car for people without time that's because there's a huge market for it like if something goes wrong is that your fault or is the person who rents it to you that's really the reason why you don't see too many rentals so yeah anyway I was just going to say the instrument I was using for the election was his 1878 just bring that up again that was my European instrument you can hire a power western to analyze promises oh yeah that would be a market for us there you go all together that might be a series I'll need your knowledge of sample what were you going to ask Kathy? so you have these profiles for different kinds of surrounding how discrete are these geologically and how homogenous do you have the basin you've got your clay so in three dimensions if it's a basin it's going to look like a cereal bowl that's maybe it could be the size of two footballs it could be the size of a city it could scale up to incredible dimensions it's totally random but it's typically a soup bowl shaped thing where you've got a center and then it kind of wings out the edges so technically though how does this group have these profiles these elements of profiles depends on where you are and how much oxygen is in this system the ocean is not binary it's not like there's a cube in the ocean so it's going to be a gradient that's true that's true for almost all formations though no formation is pure there's always changes and by the way the other thing that will mess people up is you can have stratigraphic horizons and volcanic eruptions that are in the interviews hold show up in them too for example I can show this data let me see I can find it down so desktop, projects echo it upon back analysis so this so this here let me bring up photos that so here is the Dakota hog down you can drive up to the Rocky Mountains this is the entire age of the dinosaurs we're getting to end in the one hillside a rocky mountain uplift just tilted all this way so when you find a dinosaur in Colorado it's usually along here but it stretches from New Mexico to Wyoming so it lends the dinosaurs straight line anyway I thought you would be fun let's take a tracer and let's measure before you send it across the whole darn thing so that way I'm doing a whole formation analysis by hand so I can see how different is this one how different is this one you're doing coordinates on the side of the hill exactly right and here is my result for entry this is a ppm by the way these are volcanic eruptions and this that ash layer 300 ppm hit tree have you ever seen that this is a unique volcanic system that existed somewhere under it I had no idea why I'm not the first to see this but you can use these plays to make ceremonies what happens if you grab here versus here you have a totally different elemental signature for the same geologic formation so yes the geologic formation this is another reason why clay sourcing is so hard I don't think you can ever truly get to the pit that this surrounding comes from but can you tell differences to answer archaeological questions absolutely no question but that would be the scenario but that's an example of how much variation you can have in a single formation it's huge it's absolutely huge by the way there are other ratios we can do here that can do climate do I have I hope I have it here by the way this is a zirconium to rubidium ratio so I can see sediment size there so you can do rubidium instead of zirconium and rubidium is the supreme rubidium that's the sediment size higher zrrv means higher part of the size let me see if I can find it here the really exciting is nb to rubidium in what geology is that what that's in the marine layers this is marine this is marine that's terrestrial but the answer to your question is in both in both marine and terrestrial zirconium rubidium should be part of the size basically the bigger the sediment the more likely the zircon crystals is what's going on the smaller it is the less likely because zirconium crystals have a certain size so if you shrink down below there's just not enough space to hold them anymore and so that's why the part of the size that that works do you think it's possible that there's a relationship like as if an entire new gene could be it very well could be I don't know for sure whether or not that would be the case my understanding was zirconium crystals but I could be wrong is it really harder to measure if it needs small particles but the x-ray the particle size shouldn't matter it shouldn't be narrow gaps or anything like that I think it is just it is really honestly as embarrassingly simple as you see more zirconium by the way here's an example for this so we've got two formations this is the cretaceous this gray stuff is the duress here if I use zirconium to rubidium and potassium to titanium I can almost perfectly pull the two out from each other so this could be an example of formation clustering rather than within the formation cluster and here that volcanic ash you can see the volcanic ash spike there but it's not as big of a deal so you can absolutely pick out different signatures where is that? in coda? coda is the name of the geologic formation and then mortison is the other formation mortison colorado mortison colorado is a town that's red rose mortison is also the name of the formation where you find all the dinosaurs that you see in museums like allosaurus diplodosis, stegosaurus they all come from not just that formation but from the town of the two but here's another one that works particularly well so you can source formations this way too but there's a difference between formations and within a formation and you'll need to do both for ceramics anyway that's just proof of concept I just had a lot of fun doing this it was a good couple of days but it was a ton of work because it took us two whole days just slowly walking half a meter and you can get a sense of how big an area it is from the currents it was a lot of work but it was cool we need to publish it at some point because I think there's some cool information about the fluid dynamics how long was your assay time? we did 30 seconds and then we did we had two tracers one person did light one person did trace so we had the guy doing trace with a 5i and I was coming back with a 3sd with a vacuum and measuring right behind it so yeah it was good times I also got I think the best single action shot of any human using the tracer because you're right on that formation boundary right student luckily went to college we were worried it was from a real poor family couldn't afford it but we thought it was cool let's go back over here here it is on phosphorus and I interviewed the nutrients and you can see when phosphorus peaks up where you've got to biotically have to zone same is true for your ceramics for whatever reason you've got a super productive ocean you're more likely to find phosphorus in that but be careful that could also be organic residues from u-square too so that's why you want to measure multiple parts of its u-square it won't be on the outside but it will be on the inside so you'll localize the pulse and as a bonus you can use that send it out for analysis and figure out what proteins are there and you can actually maybe identify what you're eating alright next up bromine so bromine likely indicates biological activity as well but you will see it on the trace scale it's not the only way you see bromine but you can't see bromine as a trace it's a little awkward to use but yeah and here's an example where we've got little organic content and then bromine and you can see the alignment right there so bromine can indicate organic content as well I don't know why this one's truly mysterious to me but I know when I've worked in settings where we know a lake in an archeological context has basically gone dead and brackish the bromine drops so bromine is definitely tracing biotic activity but I don't know the direct connection between why where is the TOC organic content TOC inorganic would be TIC you got it here's a separate paper bromine this is just an extra account it's a pretty sharp relationship so bromine can be a source or two and what's nice about it is phosphorus can be recycled back into the ecosystem that's the problem if you use this by most indicator bromine is not really useful but it is so bromine can be a nice passive indicator biotic activity so if you have a Mycenaean ceramic what a bromine Mycenaean and then you've got an egyptian ceramic with less and you see bromine is trending that tells you Mycenaean play source had more biologic activity and now you can focus in on that and the reason I bring this up is it's not a dome correlation you know how to tell the difference chlorine chlorine is primarily a problem that usually indicates a problem with data collection in course so for example what happens is if I take a core from an ocean and I set it on the ground what happens to the water inside that core it evaporates and as it evaporates it carries chlorine to the top and deposits it so for people trying to do environmental reconstruction chlorine is a massive problem it will be for you too each where site leaves that can cause chlorine to you so I don't recommend making any arguments on chlorine it's almost certainly not going to be from the clay it's going to be something else that's added in after the fact or an auditory preservation so do not use chlorine do not make insects from it all that jazz I avoid it as much as we can it doesn't mean chlorine in principle can't be used scientifically or in environmental reconstruction it's more of a distraction which is why I use the E as A I can tell you it's hard to contaminate your site oh yeah that's a negative one so let's go over this again earlier we'll find that discussion but the same principles are true for your ceramics just in a different context calcium is typically relative because there's always calcium but it's relative changes to something else can tell you something about the environment to say clay to do clay to clay fractions we want to use correlations between aluminum, silicon, potassium, titanium, iron and A if you can see it that's also typically a clay and that will tell you the clay fraction so we use that in pourers because if I use the correlation between silicon and potassium for example if I get an R squared of 70 that can sometimes tell me about 70% is clay it sounds weird but that correlation actually is a big proxy for the total content of plays because that tells me 30% of the iron is not bound to clay potassium thus there's a non-clay component for organics you want to check for chromium, nickel, aluminum, and uranium also maroon can indicate activity as well typically speaking these guys, the redox elements the nadium, chromium, nickel, and aluminum, and uranium they tell you about the dead stuff baromi tells you about the living stuff if you see baromi everything was living there if you see chromium, nickel, and aluminum they died there and the bodies stayed there that's basically the difference and that of course that way you can both tell where the clay comes from in a more local way sense and then finally nutrients which is, this is stuff you probably want to avoid in ceramics since it could be a use phosphorus, sulfur, and potassium those typically can be different but that said with one big exception I need to add here when I was doing ceramics, working with someone to do ceramic sourcing in Okinawa in Japan, there were two islands they were trying to figure out which ceramic came from which we found that sulfur was half a weight percent of one island sediment so in this case sulfur was the source for that so I wouldn't rule out sulfur entirely but just be cautious you see it and that's the extent of my geochemistry on all that I hope it's helpful, but that gives you a rough idea of the elements you can use in general I would put, create three bins the clay bin and that's aluminum, potassium, titanium iron, and silicon I would put the redox bin and that's molybdenum, vanadium, chromium, uranium and nickel and then I would have the biotic bin and that's phosphorus, sulfur, bromine and those three give you clusters of elements to infer from so any questions about all of that environmental reconstruction style what did you keep back to your slide that had calcium yep this one? oh ok basically calcium there's a lot of stuff there's a lot of different contexts but it's use relative is great it's use in absolute terms in terms of weight percent actually doesn't tell you much at all it's you almost always a reference point for some things by the way out of curiosity have you guys seen those red vents like if you drive in the west and you see these dark red soils what are you talking about like cliff sides that are bright red do you know what those are from? they're related to calcium let me show you an example of ceramic by the way this will teach you about what's in ceramics too in a useful way workshop 6 reports tomois ceramics so this is the set of ceramics they're blue clays and there are red clays so they're blue clays and red clays what would you guess makes a red clay what element oxidizes red? iron so most people say red clay is iron just like those red vents have iron if you've been to chocolate candy that orange soil has iron but here's the catch the katabua ceramics actually don't have any changes in iron it's calcium more calcium means less red less calcium means more red I can have iron 4% in two ceramics if iron has if there's 4% calcium 4% iron it's going to look like a brown one if there's 12% calcium 4% iron it's going to look blue if I have zero calcium and 4% iron it's going to be bright red calcium is white it dilutes it scatters the light actually so when you see something and you see it looks more red that doesn't tell you about its iron content it only tells you about the lack of calcium important point and that's a relative it's a calcium relative to the iron that actually predicts the color of the properties so this katabua ceramic they were asking what clay sources were they're using the answer was the same one the iron everything was the same it was only calcium that was changing and I can see that here blue clays have lots of calcium right clays have less elements like potassium didn't change at all were they adding the calcium as a temper could be we couldn't see the modular structure we didn't know if the calcium was geologic or if they were crushing up seashells we don't know what was it different across the types of pots it was just the color the color of the clay corresponded to the calcium that's basic so it wasn't something like all of the blue clays are used for cooking calcium, tempura and nothing like that it was stylistic as far as we can tell it was purely stylistic anyway the reason I bring this up is those bright red vents that you see those red cliffs that you sometimes see in the west those come from the permean triassic extinction that's the one where 96% of life died what we know happened then is that the ocean temperatures reached a boiling point by the way so it was super hot hurricanes benefit from warm water like the warm of the water moving the hurricane how big do the hurricanes get to be when you've got boiling water the answer is it's called a hypercane and it's about the size of the continent and their maximum wind speed on the interior of the cone is about 600 miles per hour and those hurricanes were wiping over North America regularly and we know that because we can find the ocean salt as far inland as Wyoming so it was a crazy time crazy crazy crazy time those red vents are that way because that high acidity extreme environment meant calcium dissolved almost immediately so calcium didn't get deposited which left that stark red interestingly enough those red vent formations tend to follow mass extinctions there's a red vent layer that follows the KT extinction of the dinosaurs too in the top in some areas so no one knows fully what the connection is but whenever you see a red cliff side that tells you there's a mass extinction happening either just before or during and we might know exactly the relationship but it means no calcium in the continent so yeah anyway fun times anyway but that's one thing we noticed as well anyway so I've gone through hopefully that all helps the clay we've got an hour before we've got a close up shot what questions do you have that you'd like us to cover as a group that would be helpful for your x-ray work could be anything not just missile ceramics in city and chemo stratigraphy, calibrations what's the latest on magnetite in color in the city I was learning that you couldn't really differentiate just because it's red and it looks kind of the same but it should be something measurable it's the oxygen isn't it it means it's oxidized iron not reduced right yeah because it's still iron so if you look here by the way look close can you see the black line on that guy that banding, my understanding is that those are local lines of deposits that are produced iron on this guy but let's test it right so the magnetite refers to the specific oxygenation is it right is my understanding correct that's how I always interpret it anyway some other let's see I would recommend four year transform well no I wouldn't so funny story I tried to use lids on obsidian because I thought well let's do it it'll be fun I couldn't get a spectrum of obsidian it just goes through you don't create a plasma ball you only create a plasma ball with just reflections by the way that's another problem I see the reason it's got a problem to use a laser the color of the sample changes what dissolves in the laser so you get color changes which is a huge problem in that field that's one of the reasons why they call it some quantitative the color of the sample the color of the sample changes what elements fluoresce black absorbs more of the laser so things don't fluoresce as well white overheats it and you get a ton of feedback so you get different results with the lasers yeah it's been a huge problem I got to meet a guy he worked with someone on the Manhattan project and he's like this famous laser scientist and he described to me he had this fun project this is kind of an aside where the Russian government asked him do we make laser cutting ice breakers the idea is I make a ship that shoots a laser from the bow and then it just tunnels through the ice and then the stern can just crunch through and separate the ice breakers because Russia's got tons of ice block this is important for them so he created a laser and he fired it to a block of ice and anyway so he's firing into this ice no laser comes out and so he opens it up and is looking and you just see this high powered laser and it kills to the first sunlight it's shooting through this ice and then the ice is just absorbing it all and it's not melting and so he's like what's going on so he bends down to look closer and then the laser breaks up the side of the ice the crystalline structure had caused it to arc and turn and his hand was there laser blast to himself do you know what happens when a laser hits you so if you see a Star Wars that a laser blast or hits him, physically what happens when a laser hits you well it only does one thing, it evaporates water so when the laser hit his skin, all the water immediately evaporated from his thumb and the skin bound to the bone like a monkey and so it left the permanent scar like the genus from a monkey that's what a laser does to you as it goes, the Russian government decided not to precipitate it for them yeah right they'll have plenty of time now right let's see if I can connect yeah, alright, awesome, we're on so you need to measure it method, we're gonna do this with the yellow filter because we're interested in seeing the higher what are the settings that you're doing I'll pull you through all these let's use the obsidian as an example in order to alright, so let me take you through a tour of all these settings we're gonna do 40 10 and no filter this is our test? this is our test, yeah is there a new version? there is no connection to the tracer 5.5 I can send you a drop obviously or give you a USB, I can throw it on and it installs a lot easier so I recommend using it and it's perfectly compatible with old data so you can use it now if you want to you don't need a new tracer to enjoy the benefits alright, so measurement method, we're doing 40 KB, 10 microamps, no filter turn it away from me easy cal doesn't quite work you don't want to use it that's when cal process is better I think anyway it's not, is it still working on that? no, it's um that's the one that's not in Excel yeah exactly, and I can send you a copy of yourself when you tend to need it but I recommend using it I recommend CloudCal what are we even talking about use that for your old tracers it can read in PD's heads now anyway so you're good to go anyway, so here it is no filter data, what do you notice about this? a lot of backscatter here, right? remember the iceberg analogy from yesterday most of these peaks are under here so we're gonna come over here I see what you mean, I missed that the the backscatter is obscuring yeah, exactly the filter just takes the backscatter it doesn't do anything else, anything with the other peaks exactly right so we're gonna use 25 microamps titanium here that's the yellow filter filters are the same, but you know what, the names, the color names don't work as well, it's inside the instrument they help me they're not really helpful for the newer user no, they are, I will never not think of it as the yellow or green filter but yeah, so here in red is no filter, in green is with the yellow filter, what happened? you're gonna work a lot of bad noise what happened to my light audience? they came down they're gone I destroyed them so, the benefits but over here I don't know if this is noise or not here, I can see that right there that zinc that right there, that's gallium I can see them clearly here whereas here, there's too much going on so this is why this is why you use a filter I get benefits here but I lose my elements over here because I'm not sending in anything to accept them now this is the green filter this is the yellow filter, 25 microns titanium, 300 microns aluminum what does that mean? it's a thickness so it's aluminum metal or titanium metal for analysis of this is what we typically use from metals, by the way this is the filter you'd want to use so yeah so this, along here and what's happening here an element like here, I see so clearly because this excites it but if I'm studying these elements over here I'm losing a lot of signature so anyway so that's the problem there if I come over here to measurement method we're gonna up this to 40 and we're gonna use 100 microns copper we'll do this for a 30 second this is the red one? this is the green it can't aid the obsidian filter it was developed by Robert Spiekman when he was in the Smithsonian Bruce blew it out to him this is where Bruce knew what obsidian was why obsidian was important and Spiekman sat down with him and showed him how to do it this was developed for obsidian it's been used for a lot of other things but that's main uses of obsidian alright so this is with the obsidian filter so you didn't produce those at Murrow or was it here in the Smithsonian? I think so when was it at Murrow? in the 2000s and then in the Smithsonian my recollection is the budgets were splashed as well as this travel budget and everything like that things I've been up and trying to finish this project I see so what's different about the pink? it's nice and flat right here that's why you use that for your obsidian that's why it's better than the yellow by the way you remember how yesterday you were telling me that maybe using higher is better what do you think? it's an identical match unless you're sampling small source samples yeah so what I was telling you yesterday is normalized there right next to it and then I think you'll get a better ratio but you have to be able to export the counts after normalizing and you need to use something other than software to do that that spreadsheet is not going to cut it so I recommend Cloud Cloud personally I'm biased but anyway that's the deal and you can see here the peaks are of a similar size all we've done is we've changed that background together how do you what's your strategy for normalizing? I usually aim for rhodium but it changes the results right so you notice how the green filter looks really fat back here compared to these guys that's actually an illusion if I go over here and normalize back here you can see what's going on this is unaffected by the filter and then you can see the filter that's the slope right there that's the slope right there and the mist doesn't have a slope without the filter so each filter creates a ski slope and what you're aiming for is you're aiming for the elements at the bottom of the ski slope so here the ski slope leads to these guys so the way that can you unnormalize for a minute's leak? yeah we're going to do options per second so this is unnormalized so what Bruce was explaining to me was you basically have three different areas you've got this area this area and then this area yeah I would agree with that I would qualify a little more yeah so for the green filter they stop and start in particular elements where you overlap so the green filter optimizes between about 13 and 16 KED and that's where rubidium, strontium, uranium all those guys call rents that are great for a city the yellow filter or the 25 micron titanium filter it's real it's both uranium and flores is next to rubidium and strontium so yeah it's both the traditional yellow filter now 25 micron titanium 200 micron alumina that is for between 10 and 6 KDB that's the twin filter the red filter or the 75 micron copper fills the gap between the two it does 10 to 13 so it depends on what element you're looking at which one you want to use what do you usually use red for? usually it's used for toxic things usually used in safety contexts so if someone is trying to see if the lab isn't drinking water they'll use the red filter that's why it's called the red yeah it has it the joke I used to make is gold also shows up really good with the red filter so my go-to is you're either dead or you're rich that's your gold and then for pottery so for pottery, here's the problem what do you use to source pottery? you use a whole bunch of different things that's why you calibrated for all of it so that way you don't have to choose so you can choose on the spot what's the element for this project and it's already optimized, it's already calibrated the problem you have is that for pottery it's just $2,000 for calibration and each of those faces is for calibration so if I made I would have 4 different calibrations which would be $8,000 so that's why it's better to do it yourself mud rock mud rock is still you still have to choose how you take the data from mud rock though do I do mud rock with this filter or this filter same object, better than different results now you're looking at $4,000 for two calibrations exactly and then you have to scan you have to use four different calibrations scanning them six or eight times on each object so how does that work if you're flying to do the mud stone mud rock how do you actually do that with this device what does that mean that you're doing the calibration I measure my standards with the same parameters I measure all those standards and then I apply it to the menu samples but you have to think about it how much time do you have in the field when you go up there we're usually there for three months so you've got the time to sit in the lab and crank through hundreds of ceramics but it's time that's going to be your enemy so for something like that if you wanted to build your own calibrations based on nice and easy ceramics you go through a crack on a pot where you have a huge data set and then you use that to calibrate those calibrations I know you know how much calcium iron could cause you can't use those Ted Pena has been working with Mer with Mer so those are presumably no you see here in the department I would ask him do you have those ceramic shards that have no value and then you can calibrate those to this and you can build a roman calibration it might be covered but you need to do the thing to remember by the way with calibrations let me hop over to cloud caliper this bit so cloud calipers so if I go over here to spectrum so this is the this is the obsidian calibration this is all the spectrum together the red represents the minimum and the maximum and the dotted line represents the mean if you want to see all the spectrum together I just take out variants all the spectrum my goal with the calibration is to get from the minimum and the maximum for every element that's the only objective for a calibration if I have the minimum and the maximum I mean the same material so if those ceramics cover the minimum and the maximum that you're going to see they're great if they don't then mud rock is bad the reason mud rock is good is because we got every single geologic formation we've got about the biggest range that you would expect to ever see in ceramics almost every archaeological ceramic should fit in the minimum and maximum of this that's the idea how many samples are part of the mud rock calipers? mud rock one is 26 mud rock two is 42 I want to add two together because then you've got it you're not over laughing you're totally suffering so the reason they had to do it is they ran out of mud rock one a lot of the geologic formations I think they couldn't drill again so whatever exists is all that exists that's it end of story that's why they didn't use mud rock two brookers still had mud rock one so brookers could add them together and make a complete sweep but it'd be hyper convenient I think that would be good to do they're not as interested in the science of it they're much more interested in making money so then if I have my calibration I choose the elements that I want to calibrate for and then I add my concentration so my weight percents here so that's what I mean if you know what's in the samples then you can build a calibration curve if you don't you can't and then the goal is to make the calibration curves now this is my calibration curve for sodium and it's not bad right by the way one thing you'll notice when you see how the regression line falls off are these dotted points closer to the regression or closer to the true value the dotted line yeah so this calibration is actually great even though the regression slope is driven by this guy these guys are still doing great so that's the trick with the machine learning stuff is you want to look at where it falls rather than the dotted line because the mechanics get a lot more different than Lucas too but anyway the point here is if I have a value between 5.5 and 2.5 can I use this calibration to quantify it yes if it's less than 2.5 and more than 5.5 can I use it you can but the error is increasing and you see that here the error on this extreme end it's getting really hot it did if I take this guy out that should shrink out it's exactly right so wouldn't you take that one? well in a Lucas 2.5 way watch what happens in machine learning I just clicked it plug Cal you can just click on the graph to take out points that's all guess what happened these guys split out so taking that out actually hurt the calibration didn't help them having the outlayer in made it better are these guys farther far away from the dotted line now if I click this this is very exciting that's what's up what's happening is outliers are not as bad in machine learning I still have a problem with that point but it helps me understand these guys better it's sort of like knowing David Bowie exists he's an outlier but I learned something about him that different so that's the way to think machine learning the rules are a lot different what you want is actually to have points here fill it out and constrain it is that more samples so actually the best thing you could do is you've got these roman ceramics measure the roman ceramics the same way there's nothing more the better the more data we have the better we get that's exactly right and I think the same is true for XORF do you know why we limited to the sample sizes because Excel started to get funny if we had to sit more than 60 that's it that's it it was Excel it wasn't the math the more samples is still better it was the limitations of Excel that constrained those calibrations we don't operate under those constraints anymore so yeah the more the merrier and you're welcome to use my cloud cloud software you're welcome to use the s1 cal process spreadsheet you can try with easy cal but easy cal you'll find if you have a trace of 5i it's worth it because that's what it is here and then I'll hold it but if I can't put it on here then you have to go on the computer anyway cloud cal is actually easier to use so using cal 1 is on minisee? no the file is produced from easy cal operator minisee so when I make a calibration with this guy right here if I look at that unit and I could spring up windows whereas cloud cal doesn't produce those files no it does not yeah I cannot put cloud cal files on the tracer is it a proprietary format? not really it's just the line definitions everything is too different maybe if I spent 6 months and mimicked it I could force it to happen yeah no worries that's right are you going to be back me go