FT2008 - Alaska chapter 2

As I mentioned in my last post, the scientific program at FT2008 (The International Conference on Thermochronometry in Anchorage) was overall pretty impressive. I thought I'd highlight a few of the presentations that I found most interesting. As a side note, the extended abstracts for this meeting can be downloaded for free from the official meeting website here. The abstracts vary in length, but most are true extended abstracts with color figures. My discussion is by no means exhaustive, and who knows, I might augment it later. Here are some of my highlights, interspersed with random pictures of mine from the field trips. I am primarily sticking to methodological highlights today, I'll save the others for later posts.

• There were many discussions and presentations by one of the meeting sponsors, Autoscan. Autoscan is an Australian company that has been working to develop an automated fission-track counting system. I am still in the early stages of learning to count tracks, but I've observed the process and know most of the basics. Counting fission tracks is exactly what it sounds like, fission-tracks are etched in acid, and then using a microscope you count the number of tracks in your grain (gross oversimplification, I know, but to make a point). So counting tracks can be tedious, you need to count hundreds of them from dozens of grains to beef up your statistics. Anyways, as nice as it would be to have an automated counting system, the mechanics and potential complications of the process make me wary of trusting an algorithm. That being said, the Autoscan demonstrations are pretty convincing. You can download the demonstration and demo images from the Autoscan website here. Andy Gleadow gave the presentations on Autoscan and led the discussions. He went into detail about how the software deals with some of the more specific problems, comparing reflected and transmitted light images, evaluating overlapping tracks, distinguishing tracks from scratches and dust, etc. By the end I was sold. Again, I am not a certified fission track counter [yet], and therefore am undoubtably missing some important caveats, but Autoscan impresses me.
• Speaking of fission-tracks, there were another set of talks and posters by the group from Union College/SUNY Albany (John Garver and his student Matt Montario) about their recent work using a scanning electron microscope to date high track density zircon samples. The problem is this: Fission-track dating works because with time, tracks form in U-bearing minerals due to the spontaneous fission of ²³⁸U. Old and/or U-rich samples can accumulate so many tracks that they become impossible to count; they overlap and obscure each other too much. The Union/Albany group has developed techniques that allow them to count very high density samples. They do this by using a modified etchant (super secret recipe, well, until they get it published that is) and a scanning electron microscope. Typically, fission-tracks are etched with acid so they become large enough to see with an optical microscope. But, if you have a lot of tracks, this is a problem. So by using a less aggressive etchant, and more powerful microscope, they are able to effectively count samples that would otherwise be useless. I am assuming this will all be published soon, so I'll keep you updated.
• Barry Kohn presented some work he has been doing attempting to reduce single-grain apatite (U-Th)/He age spread in quickly cooled samples. Apatite (U-Th)/He thermochronology has been in widespread use for a little over a decade now, and as more and more data sets are collected, we are starting to identify and grapple with recurring problems. Perhaps the most significant issue are irreproducible single-grain ages. These are samples that appear well-suited for analysis, and have easily measurable quantities of U, Th, Sm, and He. Despite this, it is not uncommon for grains from the same hand sample to show significant scatter, well beyond what you'd expect from simple analytical uncertainty. There are many reasons why you'd actually expect significant single-grain scatter in slowly cooled samples. I won't go into it, but instead refer you to Fitzgerald et al., (2006) for a review. For quickly cooled samples, however, there shouldn't be as many complicating factors. Kohn presented results from his experiments where grains are abraded prior to analysis. Air-abrasion removes the outer rind of the crystals, leaving just a rounded core. Air-abrasion has the potential to deal with the "bad neighbor" problem in apatite (U-Th)/He thermochronology. "Bad neighbors" are U, Th, and/or Sm bearing phases that are close to or in contact with the apatite crystal in the rock. Because the He atoms move about 20 microns or so when they are expelled from their parent atom, He produced in neighboring phases can be implanted into the apatite. You end up with "parentless" He, which gives you artificially old ages. So, the idea is that if you abrade off the rind, you remove the region that could have had "parentless" He implanted into it. Kohn isn't trying to say that this is the only answer or that it always works, but in some of the samples he analyzed it certainly had the desired effect. Namely, abraded grains showed less scatter and were more consistent with fission-track ages and/or other constraints. Obviously still a lot of work to be done, but again, very intriguing.
  

So those are some of the presentations I have thought about the most since I got back from Alaska. I'll have more highlights in later posts. I'll also have more pictures, including a special Alaskan wildlife post, and a brief discussion of our stop at the Wasilla town hall. Yes, we stopped in Wasilla. But before I leave, here is a picture I took of the Exit Glacier, I tried to get the glacial striae in the foreground with the big looming wall o' ice in the background. Unfortunately I couldn't Photoshop out the guard rope and warning sign.

Fitzgerald, P. G., S. L. Baldwin, L. E. Webb, and P. B. O'Sullivan (2006), Interpretation of (U-Th)/He single grain ages from slowly cooled crustal terranes: A case study from the Transantarctic Mountains of southern Victoria Land, Chemical Geology, 225, 91-120.

Zotero and Citation Managing

Like many paper-writing academic types, I typically employ a haphazard and ill-planned technique for reference-handling when writing papers. I have EndNote and BibTeX, and I have used them both at times, but I really have never been entirely satisfied. I have liked EndNote at times, but the truth is the version I have is old and clunky, and I am not all that thrilled about plunging another $100 some-odd dollars into an updated version. BibTeX was great while I was using LaTeX, but most earth science journals do not accept LaTeX file submissions, and none of my collaborators are ever comfortable with LaTeX, which means I have to constantly import and export from Microsoft Word to LaTeX and back. A pain. Plus, BibTeX is great for most things, but putting out a .rtf formatted bibliography is not one of them. So to summarize, I want something cheap (read: free) that works well with Microsoft Word (the standard for my collaborators).

Enter Loose Baggy Monster, my better half. One of her favorite things in the world is playing with new word processing related software. Seriously, I'm not joking. She is especially excited about reference management utilities and other things that let her release her humanities dorkosity.

She pointed me toward Zotero, a free reference and bibliography management utility that works through FireFox, and integrates beautifully with Word 2004. I've now become a convert, and decided it is time to spread the word. Now, there are many features in Zotero I don't yet use or understand, and I am sure in the future I'll do a better job showing off it's power, but here are some screen shots describing why I like Zotero. Oh, and if you want, click on the pictures to see higher resolution versions.

First off, Zotero makes importing references into your database very easy. I do all of my reference searching on web-based applications, such as my current GeoRef interface (shown below).

You will notice that in the browser address bar, there is a little orange folder. That means that this page has enbedded bibliographic information that Zotero can read, and import directly into your database. This is a GeoRef search for the term "jimboite."


If you click the orange folder, a window appears with all of references that appear on the page. You select which ones you want in your database, and once you click OK that are imported. You can see that there is another window on the bottom right showing the references that are currently being imported.You will also notice that on the bottom of my browser window, there is the little "Zotero" icon. If you click this, you are taken to your personal database. You can see that one of the articles I had searched for and selected in my "jimboite" search is now in my database. I should mention here that this database is held on your hard drive, so even though it runs through a web browser, you don't have to be online to access your downloaded or otherwise entered database. Nor must you be on the interwebs to use the "cite while you write" functionality in Word 2004.


So now lets go to Word and start writing that landmark paper. You can see that because I installed the free Zotero plugin for Word, I have a little Zotero menu bar up on the top of the screen shot. So I've written my first stunning introductory sentence, and now want to insert a reference. I click one of the menu bar buttons.


And this window comes up. I can select what citation style I want to use. Truth is there aren't as of yet that many earth science citation styles available, but someone did make the AGU general style, which I tend to like. On the Zotero web site you can download and install the AGU style (it does not come standard), and there is also a style-maker that I have yet to figure out.


After I click AGU, I am then shown a window where I select the reference that I want to include. You can chose one or multiple, either way it is pretty simple.
Click "OK" and voila! the paranthetical citation is entered into the word file.

So now I'll add some more text and reference, and then click the menu bar button that says "Zotero Insert Bibliography" and Ka-Bam! Instant formatted bibliography.


Now GeoRef isn't the only web search engine that lets you download references straight away. Most of the individual publisher or journal web pages do as well. There are plenty of other features I like, for example you can take snapshots of web pages if you wish to record and later reference those; you can link references in your database directly to the .pdf's on your hard drive, and so on.

There are some things I want Zotero to improve on. First I'd like more citation styles. Zotero also does not have any decent way to sort out duplicate references, which is a headache. But, Zotero is one of those programs it seems is improving quickly. And hell, it's free!

So as of now I am a convert.

Eocene Farallon Lithospheric Chalupa

Also see my second post relating to the Farallon Plate here.

This past Friday I was talking with some of the people in my research group and we ended up discussing our favorite published figures. This is easy for me, hands down, it is the Eocene Farallon Lithospheric Chalupa of Humphreys (1995). I'll explain why I love this figure in a minute, but first...

I don't think I'd be out on a limb to say that when many geologists first look at a paper, they begin by studying the figures. Many branches of geology are very visual, the ultimate goal often being complete 4 dimensional reconstructions of complicated events. As such, most earth scientists spend a great deal of time learning how to make good figures. For many geologists, the steep part of this learning curve is field camp, when they first learn how to make geologic maps.

I think if you want to consider the information density of a figure, nothing can beat a geologic map. I am continually amazed at the amount of information that trained geologists can gather from geologic maps. The amount of information packed into a geologic map of course depends on the quality and scale of the map itself, but really good maps can detail billions of years of faulting, folding, erosion, magmatism, deposition, and metamorphism.

And one more "quick" aside. I think that my general reverence for geologic maps has played into my slight annoyance with Edward Tufte. This is probably entirely stupid and absurd on my part; I apologize in advance and fully accept criticism, but I am continually amazed that Tufte doesn't spend more time singing the praises of geologic maps. I had a chance to attend a public lecture by Tufte a few years ago. In general, I am a big fan of many of Tufte's ideas. I think that many people, geologists included, need to think seriously about how they visually represent their information. I think that geologists already have to spend a lot of time doing this, and so some of the examples seem kind of basic. I think this molehil of an annoyance of mine became a small end moraine of annoyance when I saw Tufte speak. First the background, this was a public lecture attended by about 600 people. Although Tufte is not categorically against things like PowerPoint or slides, he is much more in favor of archival quality paper. Unfortunately, these are expensive, and impractical for 600 people. So, the first 200 or so people who got into this lecture had beautiful large images to look at during the talk, images on such incredible paper that they are undoubtedly as bright and colorful today as they were 2 years ago. The rest of us got to crowd a dozen or so people around a single piece of paper, trying to follow along with his message. I would have loved a crappy PowerPoint slide complete with kitten background and lightning sound effects. I guess I am saying that I think it is important to strike a balance between quality and accessibility.

Long story short, I get a little bit stuck-up when it comes to geologists and visual representations of data. That is why I've decided to start blogging about some of my favorite figures.

So to start we have the Eocene Farallon Lithospheric Chalupa, published in 1995 in a paper by Eugene Humphreys (Geology, v. 23, n. 11, pp. 987-990). What this figure shows is his preferred model for the removal of the Eocene Farallon slab. The idea is that during the Laramide orogeny, the Farallon slab was subducting at a very shallow angle, effectively transferring plate boundary forces into the interior of the continent during the Laramide orogeny (making some of the Rocky Mountain structures). Beginning in the Eocene this slab was somehow removed, allowing hot asthenosphere to rise up and introduce a great deal of heat and basaltic magma into the western North American lithosphere. This led to the "ingimbrite flare-up," a period of intense volcanic activity in western North America, from Washington in the north down into northern Mexico. Volcanic activity did not begin everywhere at the same time, in fact it seems to have started at the northern and southern margins of the province in the Eocene. The volcanic front moved south from Washington, and north from Mexico, eventually meeting around Las Vegas in the Miocene (Vegas baby!) Many geologists interpret this to be the result of the gradual peeling away of the Farallon slab from both the north and south at roughly the same time.

So, here is the figure, the Eocene (~35 Ma) Farallon Slab!



So, with that background, what Humphreys attempts to explain in this paper is that pattern of magmatism, and how it could be related to slab removal. My favorite figure is his preferred model for Farallon slab removal. Why do I love this figure? Well.

1. It make sense. It takes a while to wrap your head around, but it explains the removal of the slab, and the associated patterns of magmatism.
2. The perspective. The view is looking up from the mantle somewhere underneath eastern Canada. The view is looking to the west, with the subducting Farallon slab coming right at you. I am used to map view and cross-sectional views, but I love the moxy to use this perspective. I wish google earth had this view.
   
3. The approximation. I like the fact that this is simplified and hand made. I don't think Illustrator has a feature that makes buckled subducting slabs. In my geologic education I have definitely felt pressure to make diagrams crisp and Illustrator-ific. I can imagine this figure originated on a sketchpad or (if I were to write this as a legend) cocktail napkin, and didn't get caught up in the refinement process.
4. It reminds me of a chalupa. Not that I consider Taco Bell excellent cuisine, but the buckled Farallon slab is a dead ringer for an enormous chalupa. And, it should be noted, that downing chalupas can also lead to flare-up events. I love mexican food, and Taco Bell is a distant cousin of mexican food, so that makes me smile.
   



5. I think my favorite thing though, is the fact that I didn't buy this figure at first. I first saw it in a seminar on Cordilleran geology when I was in graduate school. I remember spending a great deal of time trying to figure out what it was representing, and why this was the preferred model. At first I thought it was absurd and overly complicated. I am not an expert on slab removal or the ingimbrite flare-up, but I am now pro-Farallon Chalupa. Every time this figure comes up people who've never seen it are always unsure. Geology can be complicated. Geologic maps are complicated if you aren't used to looking at them. So I like that this image makes brains work. I don't think anyone immediately accepts this diagram, which is good. It is immediately interactive.

So what are your favorite figures? Any nominations? Feel free to nominate yourself if you have a figure you are particularly proud of.

Wunderground and free public data

The following post may seem like I am meandering but there is a point, I promise, stay with me.

We in the American northeast have been enjoying a pretty late winter this year, meaning, it's a hell of a lot colder than April is supposed to be. The place I used to live was, in general, much warmer, and I got to thinking of the temperature difference between where I used to live. This got me hitting Wunderground, a weather site I started checking thanks to my Mom (an amateur weather guru who often knows more about my weather from 3000 miles away than I do). One of the cool things about Wunderground is that they make it easy to download historical weather data. For example, after about 10 minutes of clicking and Excelling I made this
Which shows the temperature difference between where I lived in grad school and where I live now. The y-axis is the difference in average daily temperature (°F) between old home and new home; values above 0 mean it is warmer at BWRU, and values below 0 mean it is warmer here at ESRU. The line is a five day running average. Why 5 days? I have no idea, but I'd appreciate a good reason if anyone can think of one. There are a couple things I noticed after putting this together. First, we had a really warm winter here at ESRU, until well into January it really wasn't much warmer back at BWRU. Second, there is an odd apparent periodicity in the running average, with a fortnightly recurrence. You can sum up my statistical analysis of this feature with the words "Jack Squat," but it is still potentially interesting. The maximum difference was just a month ago, March 6th when it was 59° colder here than there.

OK, great, it is colder in the northeast than a western school at a lower latitude, big shocker. Doing this though reminded me of an exercise I came up with when teaching an earth hazards course in 2004, again using Wunderground. In February of 2004 there was flash flooding in the southern part of San Francisco. Pretty exciting, there was water rushing down stairwells and elevator shafts at San Francisco State University. At this time I was teaching this earth hazards course at an area school, so I used Wunderground to come up with this homework. I had the students get the rainfall data for two days from the same station, the flash flood day (2/25/2004) and another day, a few months earlier that had some serious rainfall but resulted in no flash flooding (12/29/2003). This is the graph they were supposed to come up with

The y-axis is rainfall rate (in inches per hour), and the x-axis is time. The total rainfall for both days was actually remarkably similar, actually 12/29 had more cumulative rainfall, but the flash flood day had all that rain packed into less than an hour, instead of spread out over the day. Basically I was driving home the point that flash floods, unlike regional floods, come from large amounts of rain in small amounts of time.
What both of these exercises got me thinking about is free public data sets. Wunderground does not provide the kind of really rigorous data I'd use if I was working on an actual publishable study, but as a teaching tool or means of illustrating basic earth science concepts, it is fantastic. So Wunderground is great for relatively recent weather data. What about other public earth science related data repositories?

Can't we all just propagate our errors?

This weekend I continued to work on a lab software upgrade for one of the labs I work in. This is all part of an effort on my part to streamline sample analysis, automate as much as possible, and make the data reduction smoother and more versatile. For those of you who care, I am doing everything with LabView, my programming platform of choice. My first step in the process is to go through the existing sets of code, figure out what does what, and decide if I want to copy the process or change things around.

This got me thinking, again, about error propagation. Most of the significant errors in the existing code are dealt with appropriately, but there are some very easy ones that are just ignored. Truth is, most calculations are fairly easy to propagate errors through, and I am a firm believer that uncertainty must be dealt with honestly, even if in the end it is insignificant. The most popular geochronologic data reduction tool (I am guessing) is Ken Ludwig's Isoplot. This does an excellent job with uncertainties, assuming of course the users report them all. But, Isoplot doesn't interface with the machines, and it can only work with the data it is given.

For this reason I think that the book An Introduction to Error Analysis by John R. Taylor (a physicist from the University of Colorado) should be required for anyone who works in a lab or with the data from a lab. All of the basic methods to propagate uncertainties are covered and explained very well. This doesn't discuss some of the more complicated methods common in geochronology (e.g. the MSWD), but for a starter and reference text it is well worth it. It also isn't directed towards the earth sciences, but again, the basics are the basics. For example, to calculate the size and correct for the depletion of a spike or standard shot of gas, you need to know the reservoir tank and pipette volumes, and the pressure of the gas in the tank (partial pressure if your gas isn't pure). Associating uncertainties with each of these measurements is fairly straightforward, so any number that comes out of these values, say size of spike, or a correction for ionization efficiency, should include those propagated uncertainties. Numbers without uncertainties really don't exist in geology, but they show up in data reduction all the time.

Dealing with errors in a realistic and representative way is the first step. Getting people to report them, even when they are crappy, is the second.

How Old is this Rock?

I had originally meant to take on the Just Science week Challenge, where I'd write a new entry every day about science and science only, no Bears, no music of the week, no asinine and unproductive bill brought up by some pandering politician. But, I flaked. Brian of ..Or Something fame kicked this challenge's ass. But, I decided that instead of daily posts, why not one ginormous one, guaranteed not to be read by even my most dedicated friend? So here you go, the science of Geochronology. I intend to come back to different aspect of this topic in the future, just as I have discussed it in the recent past.

Geochronology is an umbrella term for all of the techniques we use to determine the absolute ages of rocks and/or individual minerals. All modern geochronology is based on measuring the stable daughter atoms produced during the radioactive decay of naturally occurring elements, most notably uranium (U), thorium (Th), potassium (K), rubidium (Rb), and lutecium (Lu). It is now common and fairly simple to measure the absolute abundances of these elements in rocks, or in many cases individual crystals, with a high degree of precision. As a geochronologist, the real work comes with figuring out what those ages mean.

The title of this post is a question I've been asked many times, and in truth is impossible to answer. Rocks have very complex lives and change continually on a variety of timescales. Each radioactive parent and stable daughter nuclide has specific characteristics that determine their inclusion and retention in geologic materials. When you date a rock you are really dating some event or process that the rock experienced. Some systems can be interpreted to record the age of formation of a rock. Others record the time when a rock cooled to a certain temperature. I've seen geochronologic data misused by lots of different kind of people. There is of course the young-earth crowd, but I really don't want to waste time on them (in this post). What is more disconcerning from my standpoint is how often I have seen data misused by trained geologists. In the past year I've seen talks where sanidine K-Ar ages from the 1970's were compared with U-Pb ages from the 1950's; I could tell that the speakers grasp on the philosophy and method of geochronology was simply nonexistent. I am more lenient when non-geologists misunderstand radiometric dating techniques, but when the data is something that you depend on, well....What it really comes down to is:

1. What are the chemical and physical properties of the parent and daughter nuclides?
2. When did the mineral that contains the parent and daughter nuclides form? Is it igneous, metamorphic, has it been deformed or recrystallized?

Calculating an age is pretty straightforward, all you need to know if

1. the number of parent atoms
2. the number of daughter atoms
3. the decay rate of the radioactive parent

When I've taught radioactive decay to classes in the past, I've found that one of the biggest mental stumbling blocks that students have to overcome is how a process which is completely random on the scale of an individual atom (whether or not it will disintegrate) can, on a larger scale, be completely predictable. It is of course all a probability thing, each radioactive atom has a specific probability that it will decay (the decay constant). Below is a graph of observed decay rates from a "radioactive parent nuclide," compared to a expected line.

Actually, these particles aren't radioactive, they are chocolate. These are the results of the great M&M experiment from a class I taught in 2004. M&M's, it turns out, are fantastic for illustrating this principle (pennies would work just as well, but for some reason the classes are never as enthusiastic about those.) M&M's have an "M" printed on only one side of the candy. So, you put 100 of these in a paper cup, dump them out onto your desk, and count the number of pieces that landed with the "M" side up. These are considered to have decayed, so you remove them, place the undecayed candies back in the cup, and repeat until all of the M&M's are gone. I even had them calculate an M&M "half life", the result of which you can see in the corner of the graph (± 1 sigma). I'd have groups of students do this with 100 pieces 5 times each. Then we'd compile the data as a class. Most people are comfortable with the idea that you have a 50/50 chance of an M&M landing heads up, but if a particular piece decays is random. This is directly analogous to the decay constant for a radioactive element. Each atom has a specific chance of decaying. The only real difference is that decay constants are in terms of some time value (per year, second, etc..), while the M&M decay constant is in terms of "rounds." These results are always impressive to me, that we can get such good results from ~4000 "atoms," especially when the average zircon used for U-Pb geochronology contains hundreds of trillions of U atoms, boo-yah, counting statistics.

U-Pb geochronology is unique among geochronometers in that it is actually two separate decay systems in one. There are two relatively abundant naturally occurring isotopes of uranium, ²³⁵U (which decays to ²⁰⁷Pb) and ²³⁸U (which decays to ²⁰⁶Pb). ²³⁸U is by far the most common (now), but every crystal or rock or anything that has ²³⁸U also has ²³⁵U. The ²³⁸U/²³⁵U ratio is actually exceedingly well known; it probably ranks up there among the best known numbers in physical science (thanks to the nuclear weapons industry). Why is this number so constant? Well, because different isotopes of U have the exact same chemical behavior, so they are never fractionated. But, fractionation is a topic for another post. What is important though, from a geochronological standpoint, is that every time you measure a U-Pb age, you are actually making two completely independent and separate geochronologic age measurements, one for the ²³⁸U-²⁰⁶Pb system, and another for ²⁰⁷U-²³⁵Pb. ²³⁸U has a longer half-life, and therefore decays more slowly.



This figure shows the evolution of different U and Pb isotopes in a closed system over time. Basically, start out with 1000 atoms of ²³⁸U and ²³⁵U, and let them evolve for 4.5 Ga. In truth, of course, the materials we use have trillions of times more U, and ²³⁸U and ²³⁵U are never equally abundant, so I've normalized the graph. But, you see the point. The solid lines are U, with red being ²³⁵U and blue ²³⁸U. The dashed lines are Pb, with red ²⁰⁷Pb and blue ²⁰⁶Pb. Either one of these can be used to calculate an age. But, the real power of the technique comes when we use both. Because the decay schemes are independant, the ratios evolve separately. But, if ur measurements of the decay rates are correct (which they are), then the ages from the two systems should agree. The collection of agreeable isotopic ratios is referred to as concordia.
The red dots represent specific time values, and the blue line is called concordia. U-Pb age determinations which fall on this line are called "concordant." There are a variety of processes which can cause ages to fall off of concordia (Pb-loss, multiple crystallization events, etc..), which actually lends more power to the technique. So, every time someone calculates U-Pb ages, they are actually also double-checking the decay constants of the isotopes. If these changed over time, as a few non-scientists claim, then concordia could never work. We actually use the above diagram, called a Wetherill plot, and a similar plot


called a Tera-Wasserburg diagram, to evaluate U-Pb data. These two methods have their different strengths, which I won't go into here (right now at least.)

U-Pb geochronology has a lot of other advantages. The primary crystal we analyze, zircon, does not like to incorporate Pb into it's crystal structure, which means that almost all of the Pb we measure is the product of radioactive decay (U likes to substitute for the Zr in the zircon structure.) Zircon is also great, because it is very common in rocks, and, as people like the famous ..Or Something wrote about, is tough enough to survive significant erosion and transportation and therefore pops up in sedimentary rocks all the time, giving a great deal of insight into sources and transport of sediment.

It also turns out that Pb does not diffuse out of zircon very quickly. In many cases, even if the system has experience really high temepratures, high enough to melt many common rock forming minerals (quartz, felspar..), the zircon will survive and retain it's U and Pb. If you have carefully evaluated the textural relationship of the zircon to the rock, you can often say that the zircon U-Pb age marks the time of crystallization of that crystal, and perhaps that rock.

K-Ar dating, or the more powerful Ar/Ar variant, works essentially the same way. Minerals like to take up potassium, and one isotope of potassium is radioactive. Minerals don't generally take up any argon, at least none in the crystal structure. So we measure the amount of K and the amount of Ar and we can calculate an age. The reason Ar/Ar is different is that we can use a proxy for measuring the K content, this makes the measurement process and the information we can recover from a crystal much more powerful. I will post about Ar/Ar geochronology later, I don't want to go into the details. I think one of the most amazing and interesting proofs of the K-Ar system was published in a 1997 paper by Paul Renne et al.,

P. R. Renne, W. D. Sharp, A. L. Deino, G. Orsi, L. Civetta (1997) ⁴⁰Ar/³⁹Ar Dating into the Historical Realm: Calibration Against Pliny the Younger, Science, v. 277, p. 1279-1280.

In this study they dated ash from the 79 A.D. eruption of Mt. Vesuvius. This is one of the hardest kinds of analysis to do, on a geologic time scale barely any time has passed for Ar to accumulate. But, as shown below, they actually nail the date reported by Pliny the Younger (a recording witness) well within error.

Pliny the Younger would have predicted 1918 years, they got 1925 ± 94.

The chemical and physical behavior of Ar makes the K-Ar system very powerful. Unlike Pb, Ar is a light gas, and when a rock is hot, Ar can diffuse out of the system almost instantaneously. We know a lot of Ar exits rocks into the atmosphere, about 1% of what we breathe is argon. Anyways, K-Ar, or Ar/Ar dates don't really tell you when a rock or crystal formed, but more when it cooled to a certain temperature. This is analogous to the (U-Th)/He system I blogged about here. Different minerals retain Ar differently. So, in the same rock, Ar/Ar ages from 2 or 3 different minerals can help chart the cooling history of the sample. Radiometric techniques that are primarily dependent on temperature are called thermochronometers. The elite earth scientists who use thermochronometers are thermochronologists. Think of thermochronologists as the navy seal or olympic athlete of the geology world.

So how old is this rock? Well, it depends. Do you want to know when that crystal formed? When the rock cooled near the surface? When the crystals that make up that sandstone formed? When the sandstone was deposited? All ages are not the same, but interesting things can happen when you hit the geochronic.

More Urlacher Information

I realized that it might help my readers if I provided some useful values in the new format, the Ur.

Gas Constant = R = 8.314 X 10^-3 Ur mol^-1 deg^-1

Activation Energies commonly used by Noble Gas Thermochronologists
He in apatite = 138.1 Ur/mol
He in zircon = 169.0 Ur/mol
He in sphene = 186.6 Ur/mol
He in titanite = 186.6 Ur/mol (har har)
Ar in kspar = 183.3 Ur/mol
Ar in biotite = 197.1 Ur/mol
Ar in muscovite = 179.9 Ur/mol
Ar in hornblende = 276.1 Ur/mol

To help move us all into the new system, it is good to remember that 1 Ur = 1 kJ

Fitting line through Data #2

So here is the published version of the graph I posted previously, this time you can see the y-axis value and the line drawn through the data points. As I mentioned in my fist post on this topic, my biggest beef is that the uncertainties associated with the salinity measurements, or even the time measurements, are not included in the graph or mentioned in the article or figure caption. One can imagine that if the uncertainties associated with these measurements are large, maybe 1 or 2 ppt, then drawing any kind of curve, let alone one with so many precise dips and peaks, is very sketchy.

This will hopefully be a regular feature of this blog. As I mentioned, I left graduate school with a very heightened awareness of error propagation and a very critical eye for how data are presented. Just because something is published does not mean it is flawless.

It is entirely possible that the shape of this curve is based on other factors, perhaps other data sets, or more likely some known and predictable relationship between time and salinity that these data happen to agree with. If that is so, it was not mentioned in the text of the article.

Show your uncertainties!

Drawing lines through data, predetermined patterns, and good imaginations

One of the topics my PhD advisor was the most concerned about was user bias in presenting data. I spent a lot of time learning how to propagate errors through calculations, and how to decide what kind of curves are appropriate to fit to various 2D data sets. I came across an article in a recent publication (pub and author name withheld for now) that got me thinking. Above I have posted a modified figure from this publication, the x-axis is time, the y-axis is a value measured in sediment cores. Take a few minutes and decide how you would draw a line through this data if you needed to, I will come back with their version in a few days. I have erased everything but the data from this figure.

Correlating data with lines or curves can of course be informed by expected patterns, where one value has been shown to vary regularly with the second. I am not an expert in the branch of earth science discussed in this article, so perhaps I am missing something. That being said, enjoy.

One more note, my first beef is the lack of uncertainties associated with the data points. They should either be on the figure or there should have been a disclaimer in the figure caption mentioning that the uncertainties are smaller than the size of the symbol. You will have to take my word that it doesn't.