Sunday, February 6, 2011

Clearest evidence that the Earth is 6,000 years old: helium diffusion in zircons

By way of preface, I do not intend this post to be a thorough scientific discussion about helium diffusion in zircons. If you are already familiar with the topic, that’s great. But in case you don’t consider yourself “geology-savvy”, or simply have no interest in understanding the scientific argument put forth by Answers in Genesis (AiG) and the Radioisotopes and the Age of The Earth (RATE) team, I would encourage you especially to read on. In particular, I want you to consider whether AiG functions as a scientific research institution, given the manner in which they have promoted the results the study below as compelling evidence for a young Earth. My critique today is not so much about the rocks, the data, or even the interpretations of those data—it is about AiG’s failure to work according to the scientific method. By ignoring healthy criticism of their research and avoiding followup experiments, they have foregone a valuable opportunity to take part in the scientific community and instead reduced themselves to dogmatism. How? Let’s take a closer look.

What is the clearest evidence that the Earth is less than 10,000 years old?

If you perform a quick search for the answer online, whether on a standard search engine, young-Earth ministry pages, or even YouTube, you will no doubt come across a study entitled “Helium Diffusion Rates Support Accelerated Nuclear Decay”, which was updated and reposted by Answers in Genesis last week. The authors analyzed samples of zircon minerals from a well bore in New Mexico as part of an ongoing project (the RATE team) to test the validity of radiometric dating techniques. In essence, the study argues that although Uranium-Lead (U-Pb) dating of the zircons suggests that ~1.5 billion years worth of uranium decay has indeed taken place, additional evidence shows that helium (a product of radioactive decay of uranium) has only been diffusing out of the minerals for about 6,000 years (±2,000 years). To account for the wide discrepancy in age estimates, the authors propose that accelerated nuclear decay occurred early in Earth history—that is, about 6,000 years ago. This phenomenon would undermine the validity of radiometric dating techniques, since those techniques assume the rate of radioactive decay to have been constant over time. According to the authors, one must consider the possibility of a recent creation to make sense of these contradictory geologic “clocks”.

Details of this study have not only been published privately in the young-Earth community (in books, journals, and online responses to critics), but were presented in the form of a poster at an American Geophysical Union conference (Humphreys et al., 2003). Since that time, numerous critics have raised concerns about the methodology of Dr. Humphreys and the RATE team, even going so far as to accuse them of academic dishonesty (“fudging of data”). While members of the RATE team have not simply ignored their critics, the fact that AiG continues to publicize the study with very few changes from 8 years ago implies that the authors consider the counter-arguments sufficiently rebutted. On the one hand, this may invoke confidence in the results of their study: why fix something that isn’t broken? I will argue, however, that the response of the RATE team rather reveals the anti-scientific character of researchers at Answers in Genesis.

The reason the study remains popular and the reason I won’t detail the scientific problems with it are one and the same

Before you read any further, I would strongly encourage you to look at the recently published summary here. If you prefer a succinct graphical presentation, you can also view the AGU poster here (opens a PDF). I’m not asking you to understand all the science, or even read through the whole article. Just take a couple minutes to skim through the methodology and graphs, and return here when you finish.

Now, what was your immediate response? Were you impressed by the presentation of data, numerous equations for the diffusion model, photos of the zircons? To be fair, it appears at first glance that Dr. Humphreys and the RATE team have done their homework, and are capable of conducting geological research. When the first report was published some 10 years ago, I immediately thought: “Well, I don’t know how they did it, but it’s obvious these guys know what they are doing!” In the years to follow, AiG published responses by members of the RATE team to their critics, and made it seem that everyone was desperate to discredit the study, but to no avail. To be honest, I didn’t really understand the questions back then—or the answers for that matter—but as far as I could tell, Dr. Humphreys was prepared for the criticism.

So the more important question becomes, did you understand any details of the article/poster beyond their conclusion regarding the age of the Earth? Even if you can explain the use of helium concentrations in U-Pb-dated zircons, could you explain Dr. Humphrey’s method of predicting helium diffusivity given a 6,000-year timescale? If you agree with the RATE team’s interpretation of the data, can you explain the strong points in the study, and detail where the uncertainties lie and what assumptions were made? If you disagree with the RATE team’s interpretation of the data, can you suggest how to properly understand them, or cite the weak points in their study?

The reason I pose this seemingly rhetorical string of questions is simple. I am willing to bet that if you presented Dr. Humphrey’s research in detail to the geology department at any university, you would find a handful of geologists at most that would be able to follow the arguments without spending several hours reading up on the methodology. The modeling of helium diffusivity in zircon grains is a very complicated, math-intensive process that requires a particular expertise to apply properly. This method was developed only recently (hence the lack of data before 1999), and has been studied by a relatively small number of researchers over the past 10 years. The results can be extremely useful to geologists, and so hundreds of published studies have made use of helium data from zircons, but very few people are familiar with the laboratory procedure.

Unfortunately, the technical aspect of the RATE team’s helium diffusion study works very well in their favor. If most professional geologists would have trouble following along, what will the average AiG reader take from the study besides unwarranted confidence in young-Earth creationism as a valid scientific endeavor? At the same time, the average critic of AiG is left scratching their head in search of a good reason to cast doubt on the results. Consequently, AiG can knock down the endless string of bad arguments to appear victorious before their audience.

Answers in Genesis has not managed to stump everyone, however, and several experts in the discipline have provided lengthy responses to their work. For those interested, Dr. Gary Loechelt published a semi-technical discussion of each model here, while a PDF of his full technical paper is available here. Therein, Dr. Loechelt’s offers constructive criticism based on his expertise as a researcher in material sciences. His analysis is eloquent and fair, focusing on the science behind Dr. Humphrey’s interpretation of the data rather than simply dismissing him on account of his association with AiG and the Institute for Creation Research (ICR). University of Kentucky geologist Dr. Kevin Henke published an even lengthier response here, offering a more in-depth analysis of the geological aspects of Dr. Humphreys’ original paper. Although some would find Dr. Henke’s tone too antagonistic for comfort, he raises valid objections that could and should be incorporated to Dr. Humphreys’ ongoing research.

Between all the links above, there is little scientifically that I could add to the discussion. Although I have spent many hours improving my understanding of noble-gas diffusion through mineral solids (hours well spent, to be sure!), I do not have the expertise of the aforementioned researchers in this particular discipline, and summarizing their arguments here would miss the aim of my post, which is to consider AiG’s application of the scientific method.

The goal of the scientific method is to communicate the results

When I teach introductory geology, I make a point to inform students that above all, geologists are writers. A majority of any scientist’s time, in fact, is devoted to the publication of research through three steps: 1) familiarizing oneself with the research of others in the field; 2) detailing the results and interpretations of one’s own research (writing and publishing articles), so that the experiment can be tested or improved upon by anyone in the field; 3) refining the results and interpretations of one’s research (authoring a new research proposal) by considering the criticism of others. Steps 1 and 3 overlap to form a never ending cycle for the research scientist, but also connects him/her to the scientific community. Without this constant interaction, a scientist’s work is (practically speaking) invalidated, either through silence or through dogmatism. In the former, the scientist simply never publishes his/her research, so the result is the same as if he/she had never run the experiment. In the latter case, the scientist rejects both criticism and contrary data from fellow researchers, and never improves his/her experiment. Consequently, his/her results and interpretations have no impact on the scientific community at large.

Now, which of these scenarios best describes the actions of the RATE team? I would argue that Dr. Humphreys and others were familiar with the scientific literature, and published their results with sufficient detail that the experiment could be repeated. In other words, they had no problem accomplishing steps 1 and 2 above. Furthermore, their research was capable of producing scientific hypotheses regarding helium concentrations in zircons elsewhere, given a measured diffusivity. Thus they were in a better position than most young-Earth creationists before them, having a large audience and a valid means to take part in active discussion within the scientific community. All they needed to do was: 1) collect more data and report whether their model was consistent and predictive of scientific phenomena; 2) refine the model according to critiques by other experts in the field to determine whether such critiques were valid. It has been more than 10 years since the original report was published; what did they decide to do?

Response of the RATE team to the scientific community

To his credit, Dr. Humphreys has devoted much time responding to some of his critics. Neither he nor any member of the RATE team have, however, produced new data, collected new samples, or modified the original model to account for its admitted shortcomings. Instead, he has challenged others to publish their critiques in peer-reviewed journals. Given that his own research can not be found in a peer-reviewed journal, this seems an inappropriate challenge (publishing in a journal costs money, and this money is not well spent critiquing an unpublished model through an article that adds nothing to the research community). Nonetheless, both Dr. Loechelt and Dr. Henke have offered models for interpreting the data consistent with the conventional geologic timeline and posted them for public access. The response is exceedingly generous on their part, given the number of hours required to draft a monograph that adds nothing to their salary or resumé.

The RATE team concluded in their helium diffusion study that the “uniformitarian model” was off by a factor of 100,000 in predicting the observed diffusivity values. When challenged that his model was over simplified, Dr. Humphreys’ alternative tactic has been to respond that the assumptions he used were “generous to uniformitarians” and that it doesn’t help their case. In doing so, he has not only stifled the scientific discussion by taking a dogmatic stance, he has abandoned the scientific method completely. Here are a few examples:

1) Estimated Q/Qo values: the rate team reported ratios of the modern helium concentration to the total helium expected from 1.5 billion years worth of decay (Q/Qo). To calculate this value, one must assume or estimate the initial concentration of uranium in the mineral, since the decay of uranium into lead produces the helium. How is this done? Values reported by the RATE team were estimated based on the lead (Pb) content (the radiogenic daughter product), which was assumed to be constant between mineral samples (though the original formulas were not reported). Not only is lead heterogeneous in igneous rocks and minerals, but it can also diffuse out of the rock over time. Thus Pb concentration is not a viable proxy for the original concentration of uranium, and Dr. Humphreys knows this. Modern methods use the ratio of thorium to uranium (Th/U) for a more accurate estimation. The Th/U ratio is easy to analyze (and inexpensive), but even after 10 years, it has not been implemented into their model. Why? Dr. Humphreys notes that adjusting the value would “reduce the percentage retentions by only a factor of two or so” (Humphreys, 2005), implying that it still wouldn’t fit the “uniformitarian model”. That may be true, but what does it do to his own predictions? The estimated age of the zircons is no longer a tidy 6,000 years—it becomes much closer to ~20,000 years. Apparently, this is not acceptable to the RATE team (6,000 years has a special ring to it in the young-Earth community), and so they stick with the simpler interpretation (but have yet to publish exactly how those values were calculated in the first place).

2) Geometry of the zircons and grain boundary conditions: if you take a look at Figure 7 in the most recent article, it reveals another assumption in the RATE team’s model. Zircon minerals were assumed to act like a 30-micron sphere during diffusion, rather than elongated prisms as seen in Figure 1. Scientific models commonly use mathematical simplifications, for obvious reasons, when the results are relatively unchanged. Dr. Loechelt points out in his technical paper, however, that the chosen radius is too large to accurately represent a 75-micron zircon. Thus the RATE team’s model significantly underestimated the amount of helium that should still be present. Furthermore, the model assumed a helium diffusivity for biotite equal to that of zircon. Though Dr. Humphreys knows and admits the assumption is not accurate, he avoids dealing with the correction because it is “generous to uniformitarians”. In other words, the assumption further underestimates the amount of helium that should be present so that the RATE model no longer fits. If you combine the effect of geometrical and boundary condition factors, the RATE team’s prediction is incorrect by more than an order of magnitude, and they could no longer report a 6,000 year age of the Earth.

3) Geothermal history: the rate of helium diffusion in minerals is temperature-dependent. In other words, higher temperatures would have cause more helium to escape in the past (hence the lower concentrations of helium in the samples at greater depths). Dr. Humphreys assumed a constant temperature to calculate the “uniformitarian model”, even though a thermal history of the region is available and reveals the samples were at far lower temperatures for a majority of the past 1.5 billion years. Dr. Humphreys originally misread the thermal history, and said his assumption was again “generous to uniformitarians”. He has been corrected, but still refuses to work the real thermal history into his model. Why? Even if the “uniformitarian model” still fails to accurately predict the data, according to his own calculations, it is bad (and dishonest) practice to discredit a model that has no basis in reality.

But here is the better question: what kind of thermal history should Dr. Humphreys use for his own model? Would it be reasonable to assume a constant temperature through time? As the RATE team is aware, radioactive decay produces heat. Key to their model is the notion that 1.5 billion years’ worth of radioactive decay occurred sometime in the past 6,000 years (during Creation Week or the Flood). Without invoking a change in physical laws, such accelerated decay would have caused the rocks to melt and there would be negligible helium in them. However Dr. Humphreys chooses to deal with the problem, he is forced to abandon the scientific method without begging the question, since the results will always be “proven” by his assumptions—namely, that the rocks are 6,000 years old no matter what the data and models say.

Lapsing into dogmatism by refusing future work

A number of other improvements could be made to the RATE team’s model, such as using a multi-domain diffusion model (actually, this is probably the most important one), but in more than 10 years, almost nothing has been refined. More time has been spent arguing that the model cannot accurately predict a “uniformitarian” history of Earth, despite the publication of Dr. Loechelt’s alternative model that does. I think that given the budget and claimed expertise of AiG’s research team, a much simpler solution can be found.

Why not repeat the experiment?

What prevents the RATE team from collecting and analyzing new samples, even in collaboration with their critics, to demonstrate that a young-Earth model can better predict the data? The problem is neither money, time, nor incompetence—what is it? Answers in Genesis and the Institute for Creation Research have a golden opportunity to take active part in the scientific community, but have made every effort to avoid proper contact. Perhaps I am wrong, but I suspect they already know the results will not be in their favor. All one needs to do is perform a literature search for “helium diffusion in zircon” or “uranium thorium helium zircon” in scientific journals. Since the RATE team began their helium project, hundreds of articles have used helium concentrations and measured diffusivity in zircon to date thermal events in geological strata (buried layers of rocks). How many of these studies do you expect to be consistent with the RATE team’s estimate of 6,000 years?

Exactly none.

And so AiG will instead devote their resources to dogmatically defending a failed model in hope that their readers will not understand the science behind it. I imagine they will continue to publicize this study until another rigged experiment can be found. Until then, we can only scratch our heads and hope that people will pay closer attention to the most basic aspect of science: the scientific method.


References Cited

Humphreys, D.R., Austin, S.A., Baumgardner, J.R., Snelling, A.A., 2003, Precambrian Zircons Yield a Helium Diffusion Age of 6,000 Years: American Geophysical Union Fall Conference, v. 84, Abstract V32C-1047. (http://www.icr.org/pdf/research/AGUHeliumPoster_Humphreys.pdf)

Humphreys, D.R., 2005, Helium Evidence For A Young World Remains Crystal-Clear: http://www.trueorigin.org/helium01.asp

2 comments: