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

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

Saturday, July 28, 2012

RE: Comments on sediment subduction; case of the Rb-Sr decay system

Regarding the last post, Mike wrote:

"Thanks for posting this, but it's so technical that I doubt many people can follow it. Could you put in a little summary with a simple diagram or two? I think what it means is that the evidence shows that sediment from very old crust has been carried down the river, deposited on the ocean floor, subducted, and re-emerged in volcanism, and that all this is impossible in a short time frame. Is that the general idea??"

Thank your for the feedback. Please forgive me for the technical level. I found it difficult to condense the papers into a manageable narrative without oversimplifying. I hope to keep your attention, however, because these isotopic systems are discussed regularly and at length in YEC circles. In fact, they are even offered as evidence for a young Earth (!??!).

In any case, the short answer to your question is yes, you have understood correctly.

Isotopic ratios (especially of Sr, Pb, and Nd) in volcanic arcs show evidence of 'contamination' by oceanic sediments. But not just any sediments—as you keenly noted—very old sediments that eroded from very old continental crust. In the case of the Lesser Antilles, each element tells the same story: the Orinoco River basin had to be in place and depositing sediment into the western Atlantic before the ocean crust was subducted beneath the Caribbean tectonic plate. It is worth noting that this phenomenon appears not only in the Lesser Antilles but—well, every island arc around the world. Since YEC's and 'Flood geologists' would have us believe that plate tectonic movements happened very fast during/after the Flood (i.e. before these river basins formed), their model would not predict such trends in the isotopic data (in which you see perfect mixing lines between the upper mantle and modern oceanic sediments). Rather, their model is falsified by these data, even if we grant that the subduction of oceanic crust and the volcanic eruptions that produced the island chain could have occurred within ~5,000 years.

But we won't grant such a scenario, because it is physically absurd (not to mention, it misses the whole point of the flood narrative, both theologically and historically).

Hopefully this confirms that you followed the article better than you thought. However, I made a few quick figures to better explain the whole process, focusing on the Rb-Sr system. Let's begin with the first figure from the last post:

When the continental crust that now underlies much of South America formed, it incorporated a relatively high amount of Rubidium compared to what was originally in the mantleThe reason is that Rubidium does not 'fit well' into the solid structure of the mantle, and so it is excluded preferentially when part of the mantle melts. This process is well documented by experimental petrology. When you melt a rock like peridotite—a close match to the composition of the mantle—the bulky elements with large ionic radii (pretty much everything on the left side of the periodic table) are concentrated in the melt. That melt eventually produces continental crust. The result is that continental crust has, on average more radioactive Rubidium (87Rb) than the mantle. Over time, therefore, the crust will accumulate more radiogenic Strontium (87Sr) than the mantle. In this fashion, we can distinguish Strontium derived from old continental crust versus a young piece of basalt (the ratio of 87Sr/86Sr is much higher in the former; see Fig. 1).

To visualize the accumulation of radiogenic Strontium in various minerals, consider the following figure:

In this graph, there are three isotopes of which to keep track. The first is 86Sr, which is stable. It's concentration does not change over time, because it does not decay and is not produced by decay. The second/third are 87Rb and 87Sr, which are radioactive and radiogenic, respectively. Over time, 87Rb decays into 87Sr, so the abundance of the former drops while the latter grows.

Geochemists are more interested in ratios, however, and so I have plotted the ratio of 87Sr (radiogenic) to 86Sr (stable) in yellow. Notice that since there was relatively little 87Rb to begin (the blue is much smaller than the gray), the ratio of gray vs. green changes very little with time. From this graph, you can understand why the 87Sr/86Sr ratio of depleted mantle (which contains almost no Rubidium) has changed very little in the past 4.5 billion years (cf. Fig. 1).

For continental crust, which contains a relatively high concentration of Rubidium, the process is the same but the rates are different. In this case, we have more 87Rb to start—relative to 86Sr—so the ratio of 87Sr to 86Sr changes more rapidly with time. This graph explains the high slope of continental crust in Figure 1, as well as the high 87Sr/86Sr ratios found in very old granites (e.g. the Canadian Shield and, of course, the Orinoco watershed).

We might also point out that since the half-life of 87Rb is ~48 billion years, the most important element here is time. Without a lot of time, there is no known mechanism by which to form minerals with such drastically different 87Sr/86Sr ratios. Some YEC's will try to redirect your attention to things like 'accelerated nuclear decay', but the explanation is 1) ad hoc, made up only to rationalize why the data contradict their hypothesis; 2) physically absurd, since a 1-million-fold increase in decay rates would produce enough heat to melt the Earth; and 3) completely arbitrary and even contrary to orthodox notions of divine providence, in which God does not produce random miracles just to make it appear geochemically that mountain ranges and volcanic islands formed through a more elegant and ancient process.

Just to complete, the following (roughly drawn) figure illustrates how mixing appears between modern oceanic sediments and the upper mantle in the case of Strontium isotopes:

As I mentioned, the process is much like mixing yellow (mantle) and blue (sediment) paints, which produces some hue of green in between. The precise hue depends on the extent to which mixing occurs between the reservoirs. Nonetheless, data from multiple isotopic systems converge on one story—that of an incredibly slow process by which the mountains were 'transferred' from one place (the Orinoco watershed) to another (the Lesser Antilles island arc).

I hope this clears up any questions that arose from my previous post. Feel free to continue any discussion below or by e-mail.


  1. Thanks, that's a lot clearer to a non-geologist :-)

  2. It's also clearer to Mike's sister (me) who is not much of a scholar but wants, very much, to understand.