Exploring the wonders of geology in response to young-Earth claims...

Never been here? Please read my guidelines and background posts before proceeding!

Saturday, February 12, 2011

Direct radiometric dating of dinosaur bones using the U-Pb method, and ICR's attempt to avoid the issue

Well, they've finally done it. In a Geology article published last month, Fassett et al. (2011) described their application of the Uranium-Lead (U-Pb) radiometric dating method directly to fossilized dinosaur bone. In addition to applying a known dating method to a completely new type of sample, the authors' specific goal was to demonstrate from independent evidence that dinosaur bones found in an early Paleocene sandstone were not inherited from older strata. Why is this important? For two reasons. Until now, only circumstantial evidence has been available that any (non-avian) dinosaur taxa survived past the Cretaceous-Paleocene boundary (65.5 million years ago), but the 64.8 (±0.9) Ma age obtained from their 'bone in question' provided a powerful argument that the creature lived after the global extinction event (albeit less than 1 million years after and in a single locality). The second reason is that the geologists were able to date directly something other than an igneous rock, and surely this has never been done before...right?

If you learned about radiometric dating from a textbook in school (like me), then you were likely taught geochronology as it was ~60 years ago, and in a simplified format: igneous rock cools to form crystals, single parent decays into single daughter element, crystal acts as a closed system, initial daughter concentration is known, decay rate is known and constant, etc. There is nothing wrong with this picture, and these principles are important in many radiometric dating methods employed today. But for anyone interested, geochronology has come a long way since its original applications. Scientists have since been able to place direct radiometric dates on lake sediments, calcite cements, black shale, speleothems (i.e. cave rocks), fossilized teeth, detrital minerals in ancient sediments, coral reefs, and even ground ice! These specimens range in age from Proterozoic (1.8 billion-year-old black shale) to present day (modern lakes, caves, and permafrost), and are commonly dated through multiple independent methods.

On the other hand, it would be unfair to give anyone the impression that U-Pb dating of a dinosaur bone is no big news. When dealing with non-igneous samples, many of the basic assumptions (mentioned above) go out the window and so the method becomes very complicated. For example, complete fossilization of the dinosaur bone may have taken several thousand years (or more) to complete, so the "age" of the bone is not the same across the whole sample. The apparent age depends on current uranium and lead concentrations, the former of which was incorporated into the bone as the original bone material was replaced by silica. Over time, both parent and daughter can be lost through fluids in a porous sandstone, or uranium can be added through further permineralization. In other words, it must be assumed that your 'crystal' acts as an open system and that many of the dates can or will be wrong. Normally, U-Pb data are plotted on a concordia diagram to quantify the loss of uranium or lead, but since dinosaur bones do not act (geochemically) like a magma chamber and the fluid chemistry is variable over time, there is no easy way to detect such disturbances since the time of deposition. Instead, the authors plotted U-Pb ages vs. the ratio of radioactive uranium to stable lead (238U/204Pb) to look for trends that would indicate uranium enrichment or lead loss (both of which can make the sample appear younger). In the first sample, they found good evidence that uranium enrichment had occurred (near 20 Ma) and that various diagenetic processes produced a scatter in U-Pb ages throughout the bone, while the second sample was less disturbed and yielded more consistent U-Pb ages. Available radiometric and geomagnetic dates for the rocks containing the fossils agreed very well with the obtained U-Pb age of each bone sample once diagenetic effects were accounted for (73.04 Ma vs. 73.6 Ma for Sample 1; 65.1–64.0 Ma vs. 64.8 Ma for Sample 2). In short, the authors' results are very encouraging for both geochronologists and paleontologists, and provide good evidence that the dinosaur bones were 73.6 and 64.8 million years old, respectively.

Of course, not everyone welcomed the study with enthusiasm, as elucidated by Institute for Creation Research (ICR) science writer Brian Thomas's review of the study. Therein, he claimed that the dino-bone ages were "not only the result of a broken radioisotope system," but were "contrived to agree with previously assigned dates for the samples." He advised the authors to search for remains of collagen, which could then be dated by the radiocarbon method, to prove that the samples were multimillions of years old. However, he does not expect the advice to be heeded, since it may rather provide evidence that the bones "are not millions of years old" and might "call the whole evolutionary picture into question, and...offer evidence for a recent creation."

Perhaps that's a valid question. Why not "take the scientific high road to legitimacy", as he puts it?

Despite Mr. Thomas's strong words, I think there is good reason that his critique will fall on deaf ears. First, his description of the method practically mirrors the wording of a Science Daily news report found here. Coupled with a handful of inaccurate statements about the technique—"an igneous rock...unlike sedimentary rock...has been considered 'datable' because the relevant isotopes are locked up in its tiny crystals", "radiodating", "it was taken on faith that the data indicating a younger age do not represent the real age"—, Mr. Thomas's report indicates that he is not very familiar with the Uranium-Lead system (beyond the basics). Second, he claims that samples were hand picked to match up with the "known" date of 64 million years, obviously missing the fact that two samples were dated: the first was bone from a well-dated horizon (73.04±0.25 Ma) and worked as a control sample; the second was a bone from a horizon with a range of ages (65.1–64.0 Ma, based on magnetostratigraphy), and was suspected to be older than the sedimentary rocks containing it (i.e. an old dinosaur bone washed into new sediments). While the first sample yielded a scatter in U-Pb ages (many of which were younger than expected), ages obtained for the second sample were all statistically within the age of the bed, and no "hand picking" was necessary. Third, he defers the discussion to a review article of the RATE team, which apparently shows that "evidence has mounted that radioisotopes underwent a period of radical acceleration of decay in the past." As discussed last week, this supposed evidence is not only highly circumstantial, but surrounded by poor and dishonest scientific practice. Fourth, he challenges the authors to search for collagen remains in the bone, apparently overlooking the fact that each bone sample was already mapped by a Scanning Electron Microprobe (SEM). In other words, if the bones contained any carbon-rich (organic) components, the authors would have reported it, and their results would be in Nature or Science, not Geology.

Most importantly, Mr. Thomas does not offer an alternative explanation for the geochronological data, but only claims that the system is broken. One must ask, however, why the ages agree at all if the system is simply broken? The sandstone containing the first bone sample was dated by a sanidine mineral from a nearby volcanic ash, using the 39Ar/40Ar method, while the range of expected ages for the second sample is the result of dating the ocean floor through K-Ar and other methods. In a young-Earth model, how are these processes even remotely linked? If the multimillion year ages are the result of accelerated nuclear decay in the past (an absurd scenario for multiple reasons), why would they even be remotely concordant across sample types and methods, let alone within one standard deviation?

I would suggest that Mr. Thomas simply does not have an explanation, and therefore must hope that his readers will not pick up on his false accusations about "hand picking" data. For in the world of young-Earth creationism, evidence of the Earth's antiquity simply cannot be accepted. It truly is that simple.

References Cited:

Fassett, J.E., Heaman, L.M., Simonetti, A., 2011, Direct U-Pb dating of Cretaceous and Paleocene dinosaur bones, San Juan Basin, New Mexico: Geology, v. 39, p. 159-162.


  1. There are limitations being revealed regarding your comments abour basic assumptions one of which the assumption of immense ages has actually not been vaguely close to being convincing, sorta proven let alone observed. simply because it was just too many years ago and we just simply were not even there! Don't forget the basic fundamental rule to this whole scientific method requrired to be valid in the first place is that whatever your testing has to be OBSERVED and repeatedly at that now today. in other words...what could of happened possibly in our imagination millions of years ago is just not convincing enough to give it the amount of credibililty your demand. An alternative argument is not necessary considering your limiting starting assumptions. A few credible starting assumptions to prevent questionable dating should include:
    1)a closed system (no chance of interference is unrealistic.)
    2)don't assume nothing has ever changed in the environment using uniformity when your talking millions of years! unrelaistic, you were just not there to know!
    3)don't assume your gathered samples are close to be the same age, most likely their not for obvious reasons i hope i dont have to spell out.

    1. Science does not require direct observation. An individual cannot precisely measure a temperature, we rely on an instrument to do that. And in order to make sure that the instrument is accurate, we check it against other different types of instruments, and against known points (freezing and melting of distilled water, for instance).

      When measuring ages of things very long ago, we rely on other tools. Radiometric dating. Magentometric dating, stratiagraphy, dendrochronology (for more recent periods than this, admittedly). And we calibrate against known events. Dendrochronology supports radiocarbon supports ocean sediment supports ice core supports other radiometric methods. To throw out one, even at the long end of the scale is to ignore them all, even those that can be checked against events in recorded history.

  2. Peter, thanks for your comments.

    I'm not sure I understand what you mean by "assumption of immense ages has not actually been vaguely close to being convincing, sorta proven let alone observed." Whether or not we can verify (or scientifically corroborate) events from history (either *in our own lives* or beyond) is a matter of philosophical discussion. We must first reject solipsism, affirm the reliability of senses, and assume the uniformity of natural law through time. At this point, the evidence for deep time is overwhelming. The only valid objection—though not from a Christian perspective, I believe—is that God/gods created the universe to simply appear older than it actually is. Such an objection undermines the validity of all scientific inquiry, however, and so the irony is that what formulates the objection in the first place (scientific evidence of deep time) no longer holds any meaning.

    With regard to your 'starting assumptions':

    1) Nobody in this study simply assumed a closed system in reporting a final date for each fossil. The assumption is used to calculate a model age, but understood to be invalid in many cases.

    2) The authors used various methods to falsify this supposed assumption. They did not assume that "nothing has ever changed in the environment" but rather demonstrated the contrary. This fact is inconsequential to the final dates reported, however, so your argument does not constitute a meaningful objection.

    3) Again, I'm not sure what you mean. Who assumed the samples were "close to be the same age"? What does this have to do with the study? The authors made a hypothesis: "if the highest U-Pb model ages are concordant with Ar/Ar ages from the bounding volcanic strata and show no geochemical evidence for U-enrichment or significant Pb, *then* we have corroborated the use of U-Pb dating as a valid technique in dinosaur bones." P, therefore, Q.