RSS Feed
iTunes Link

Episode 112 - Is Mercury's Magnetic Field Decaying? Does that Prove Recent Creation?

Download the Episode

Recap: In 2008, and ever since, young-Earth creationists have claimed that MESSENGER spacecraft results from the planet Mercury indicate that Mercury's magnetic field has been decreasing, in-line with creationist Russell Humphreys' prediction back in 1984. Find out if the data support this, and if there are other explanations. Or, if we need to throw out an old Mercury for a recently created one.

Puzzler for Episode 112: There was no Puzzler in this episode.

Q&A: There was no Q&A in this episode.

Additional Materials:

Transcript

Claim: This claim is fairly recent, as far as young-Earth creationism claims go, originating on August 26, 2008, so far as I can tell, based on a prediction he published in 1984. It was put forth by young-Earth creationist Russell Humphreys, who’s brought us other YEC claims that I’ve addressed about Earth’s magnetic field in Episode 9, a changing speed of light in Episode 81, and claimed that spiral galaxies are one of the top ten proofs of creationism, that I dismantled in Episode 92. Humphreys wrote for Creation Ministries International that the MESSENGER spacecraft, when it flew by Mercury on January 14, 2008, measured a magnetic field that was 4% lower than what Mariner 10 measured in its flybys in 1974 and 1975. 4% in only 33 years. Something he predicted back in 1984 - that all planets’ magnetic fields are decreasing and have been decreasing since Creation, 6000 years ago, but that is almost impossible to explain in a secular framework.

Investigating the Claim

Okay, so that’s the claim: MESSENGER measured a magnetic field 4% smaller for Mercury than Mariner 10, fitting with a decaying field predicted by Humphreys in 1984. In fact, in his 1984 article, Humphreys wrote: “Mercury’s decay rate is so rapid that some future probe could detect it fairly soon. In 1990 the planet’s magnetic moment should be 1.8 percent smaller than its 1975 value.” In a February 5, 2008, article for CMI, he wrote, “At [this] rate, Mercury’s dipole magnetic moment would be 4.4 percent lower than it was in 1975,” when MESSENGER measures it this year.

He concluded his August 2008 article with this:

“But the first results seem clear enough for us to expect good agreement with the creationist model. None of the now-verified predictions of the model could work without the biblically-specified original created material of planets and the biblically-specified age of the solar system, 6,000 years. When NASA’s space program began many decades ago, nobody expected it to vindicate Scripture so strongly.”

A seemingly important validation, this claim was picked up by other YEC groups including the Institute for Creation Research in articles written by David Coppedge (see Episode 75 for more on him), Brian Thomas, and Jason Lisle. In fact, just a few years later, Brian Thomas wrote:

“The Science authors wrote that the field strength for Mercury is "~27% lower in magnitude than the centered-dipole estimate implied by the polar Mariner 10 flyby." This confirms that Mercury's magnetic field is rapidly diminishing, which in turn confirms that the field must only be thousands of years old—just as the creation model predicts.”

So, there are two parts to examining this, as with any claim: First, is it real? And second, if it’s real, what are all the possible explanations, and is the one chosen by the proponent of the idea the most likely?

As the Sound of Music song goes, let’s start at the very beginning - a very good place to start - which would be step 1: What does the data show, and does it show what he claims it shows?

First, Humpreys used a 1979 review by Ness et al. that summarized the magnetic field observations by Mariner 10. He said he used this because it had the smallest error bars, and that it gave a field strength of 360±22 nT*R3. Humphreys then used a 2008 paper in the journal Science - one of the two main science journals in the world - by Anderson and others. He said that paper gave a field strength of about 280±50 — 20% lower, not 4.4% lower.

Already, it looks like something fishy is going on.

How to Get 4.4% Decrease

How does he get a 4.4% decrease, when his own graph shows - forget units for a moment - it shows the field was 4.7 before and 3.8 now. That’s 20% difference, not 4.4% difference.

From what I can tell, it looks like he took the extreme low from his value of the Mariner 10 measurement — as in, the average minus the uncertainty — and the extreme high from the MESSENGER measurement — as in, the average plus the uncertainty. When I do that, I get anywhere from about 1-8% difference, depending on how exactly you get the numbers out of his error bars. He only gives the number he uses for Mariner 10, not for MESSENGER. Which is important.

The Data

It’s important because the Anderson et al. paper gives a range, based on MESSENGER’s flyby of Mercury. They give a range of 230 to 290 nT*R3. Why Humphreys takes the average of that to be about 280, I don’t know. It should be, from a naive approach, 260±30. That would mean there’s a 28% decrease in the magnetic field, not 20%, and not 4%. That doesn’t fit the model he wants you to believe from his 1984 paper.

So, what about that Mariner 10 number? He used a source from 1979 that he said had the smallest error bars. The problem with this is that people have continued to study the Mariner 10 data to this day and built different models for the magnetic field based on the 3 flybys of Mariner 10. Mariner 10 measured a magnetic field anywhere from 159 to 401 nT. The Ness et al. review that Humphreys used gave a value at the high end of 360.

Measuring Magnetic Fields and Magnetic Field Shapes

One might think that this is a simple measurement to make. Why does it matter when the resource was written so long as it was after the measurements, and why on Earth would we ever have new models that could possibly somehow change what those measurements mean?

The answer comes from the difficulty in measuring magnetic fields, especially ones that are not spherically symmetric — they’re not the same no matter how you rotate or twist them, like a spherical one would be.

In Mercury’s case - and really, all the planets’ cases - the magnetic field is not spherical. So now, imagine that you have a balloon. Inside the balloon is some sort of material that is denser at the center and gets less and less dense as you approach the edge of the rubber. As soon as you cross over that rubber boundary to the outside, whatever’s in the balloon is no longer there - you can’t measure it.

Now, imagine that you have a probe that can measure the density of that material, and you’re going to fly it through the balloon. You fly through one random point on the outside of the balloon, and exit at another random point. You do this a total of three times.

So, you have three tracks, three traces through the material in the balloon, and density measurements all along those traces. Can you build up a model that can perfectly describe the density of the material, as a function of distance from the center of the balloon, if you took three random paths that didn’t go near the center?

You may be able to get a reasonable estimate. It may be pretty accurate.

But now, let’s repeat the experiment. Let’s put the balloon in a wave pool, such that as time passes, the balloon itself - that rubber boundary - is going to bend and flex as the waves go by because the density changes with the wave pool will interact with the boundary of the balloon. Correspondingly, as the boundary of the balloon bends and flexes, the density of the material inside will also change. And, let’s add two MORE balloons inside the main one, each with their own density profiles.

Now, send your probe through, three times, and take measurements. If I’ve lost some of you because this is hard to visualize, I think that my point, anyway, is made clear enough: This is a hard problem. Mariner 10 made three passes by Mercury, and we measured the magnetic field those three times. And, we tried to fit it to various models of the shape of the magnetic field. As computers became more powerful and as different techniques arose for solving this problem of figuring out the profile of a magnetic field and as we developed different models for the shape of magnetic fields around planets, then our interpretation of those data, of what they indicated for the strength of the magnetic field, changed.

By Russell Humphreys relying on a 1979 summary just because it had small error bars, he doomed his model straight away. Besides needing to fake it anyway by taking the lowest range of the highest estimates possible to get a decrease of 4.4%.

In general, the BEST-case scenario for the creationist is that the Mariner 10 data is just not good enough to truly say whether the magnetic field has decreased since Mariner passed by, since the estimates range from about 135 to 350 nT*R3.

And, this also ignores the later work, Anderson et al. (2011), which gives us a final number of about 195±10 nT*R3. If Humphreys wants to use his original 360 number, that would be a 45% decrease in the last 35 years. Larger than he predicted by a factor of 10, and clearly indicating that SOMETHING is wrong with what he’s doing, rather than claiming a modest 4.4% decrease.

Busted

But, most scientists say it hasn’t changed at all, and here we get beyond the creationist Cherry Picking - where Humphreys picked the data he wanted because it showed what he wanted - to another creationist stalwart: Quote-mining!

Let’s see … from the Anderson et al. (2008) paper that Humphreys used, the authors state in their next-to-last and last sentences of their short paper: “We find no evidence for a change in the planetary dipole since 1974 and also find that the planetary field is predominantly and possibly entirely dipolar. Although there are significant uncertainties associated with these results, they are consistent with the presence of a stagnant outermost core.”

Or, there’s an abstract from this year’s Lunar and Planetary Science Conference, by Philpott et al., who concluded: “Our results, together with error estimates on the best-fit dipole moment, suggest that any variation in Mercury’s axial dipole [term] … over the past four decades [is] smaller than ~10 nT.” That would actually match Humphreys’ claimed value, but it would mean that the term would have to have been only 205 back in 1975, not the 360 he claimed.

As another example of quote-mining, there’s that quote from Brian Thomas over at the ICR:

“The Science authors wrote that the field strength for Mercury is "~27% lower in magnitude than the centered-dipole estimate implied by the polar Mariner 10 flyby." This confirms that Mercury's magnetic field is rapidly diminishing, which in turn confirms that the field must only be thousands of years old—just as the creation model predicts.”

If you’re paying close attention, that 27% is more than 4.4%, so again, inquiring minds might notice something is amiss. In this case, the entire sentence is: “The best estimate for g10 is taken to be –195±10 nT (1-SD uncertainty), ~27% lower in magnitude than the centered-dipole estimate implied by the polar Mariner 10 flyby.” There’s that little term in there that might clue some of you in that Mr. Thomas’ interpretation is wrong: “the best estimate for g10

Let’s go back to the balloon example. You have your balloon in the wave pool with two other balloons inside of it, all with varying densities. The big balloon represents Mercury’s extended magnetosphere, while the little ones inside of it, which are tied together in the middle, represent Mercury’s dipole — think of it as a little bar magnet inside of the planet. You take your three passes through the balloon, and you try to fit the most basic of models of the field shape. It works, somewhat.

The problem is that you can’t get any more complicated than that. You just don’t have enough data. Now, you have MESSENGER in orbit, and instead of three random passes through the balloon, it’s actually in orbit. It goes around and around and around in slightly different places each time, as waves go by, and gets all this data built up on the strength and direction as a function of where it is.

With all these data, you can then start to see the field more accurately, and the simple model from before may start to break down. With all these extra data, you can see that the dipole is not perfect, that it’s not spherical, it has a more complicated shape. That’s what the “g10” term refers to — the basic shape. You can make it more complicated by adding a “g20” term, which is a quadrupole term for those who are interested.

When you do this, the strength of that original term is going to be less, such that you can add strength to the other term. I tried really hard to come up with an analogy for this, and the best one I could think of is a slinky. Let’s say you have a slinky where one end is hooked up to a wall, and you wave the other end so there is a single up and a single down motion. For the math folks listening, that's a simple sine wave with the frequency being the length of the slinky. That’s like a simple dipole, a simple g10 term.

Now, while you’re still moving your arm up and down to get that single up and single down, wiggle it once more for every up and every down. So instead of slowly moving your hand up and down, as you go up, give it an extra wiggle on the way up and an extra wiggle on the way down.

The slinky is going to still have that main up and down pattern, but on top of that is going to be an extra up and extra down because of those wiggles. They’re going to be smaller than the main one - they’ll be added in - but they’ll be there.

But, let’s say that you can only measure that main up and down, that g10 term. Because of the extra wiggles, you’ll measure g10 to be bigger than it really is. Like what Mariner 10 did.

But, if you have enough resolution and can make out that extra wiggle, then you have the g10 term that is lower, back to the original level, and a non-zero g20 term to account for the extra wiggle on top.

Hopefully that sorta made sense for most of you; as I said, that was the best analogy I could come up with. But getting back to Mercury’s magnetic field and this quote-mine, that’s what happened. Mariner 10’s data was only good enough to really estimate the g10 term with all the complications of other, higher-order terms, and MESSENGER allowed for the pulling out of those other terms which brings down the strength of the g10 term to only account for the dipole, while g20 accounts for a quadrupole moment, and other terms account for other moments.

So, while it may seem damning that the paper said g10 was 27% lower, that’s just because of the simplifications from coarser data and models.

Wrap-Up

In wrap-up, hopefully this has been follow-able by most of you who are listening. If you’re left scratching your heads, here’s the up-shot:

Humphreys’ and others’ claims that Mercury’s magnetic field has declined in the last 35 years, therefore it was created 6000 years ago, is likely wrong. To make their claim, they do what we see all the time from creationists: Cherry picking data and manipulating it to give them the result they want, and quote-mining from scientists to make it seem like they are saying something that they are not actually saying. When in fact, the scientists are typically saying the exact opposite of what the creationists claim.

Provide Your Comments: