NASA is wrong - Greenland crater pair probably ARE related
Topography under the Greenland ice sheet: from JA MacGregor et al., Geophys. Res. Lett. (2019)
A couple of weeks ago, a second crater-like structure was found under the Greenland ice sheet. Although not confirmed by taking samples direct from the crater surface (which is very difficult), MacGregor et al. make it very clear they think the structure on the right of the above image is very likely a crater. It bears all the hallmarks of a crater, with a raised rim, central uplift etc.
Only a few months earlier, the Hiawatha (top left in the image above) crater was confirmed - the first crater ever discovered under an ice sheet. Most importantly, the Hiawatha crater is thought to be very young, geologically speaking. It might even be young enough to have been, at least partly, responsible for the Younger Dryas impact. This is the impact very likely recorded at Gobekli Tepe, thought to be responsible for changes in human cultures on both sides of the Atlantic, the extinction of many species of large animal and the Younger Dryas mini ice age, circa 12,900 BC.
In their paper MacGregor et al. conclude that despite being very close - only 183 kms apart, these craters are unlikely to be related. They are very likely wrong. This matters because if they are unrelated, they could have been made by any random pair of asteroids falling millions of years apart. However, if they resulted from the same impact event, perhaps only seconds apart, it implies the impactor was already split into at least two pieces before hitting Earth. This is unusual for asteroids, as they mostly orbit alone, and asteroid pairs of roughly equal size are very rare. However, comets are known to split into many pieces as they orbit within the inner solar system. So, if these impact craters are related, it very likely means the impactor was a comet and not an asteroid.
In my book 'Prehistory Decoded' I dedicate a chapter to discussing the differences and ramifications of comet versus asteroid impacts. Briefly, if this is a comet impact then it supports Clube and Napier's theory of Coherent Catastrophism, which is what all the fuss is about. For years, NASA has ignored their work, preferring to focus instead on the threat from near-Earth asteroids. They have spent a lot of tax-payers' money finding and cataloguing near-Earth asteroids. But they have completely ignored the threat of comet impacts, which forms the focus of Clube and Napier's theory. According to their theory, the 'impact' of comet impacts is much greater than the threat of asteroid impacts over the last few million years. Basically, they argue that NASA has been 'barking up the wrong tree'. My book describes some of the evidence that shows their theory is almost certainly correct.
And yet, their work continues to be ignored by NASA. Why? Probably because it's embarrassing. Some people at NASA will have a lot of explaining to do if Clube and Napier are right (as they almost certainly are). But also, I suspect they want to avoid alarming the general public that there is a greater cosmic threat than asteroids that they can do very little about. It's simply not possible with current technology to either detect all the hazardous comets beyond Jupiter (that might eventually enter the inner solar system and threaten Earth), or to deflect their debris when they fragment. Since it is a threat we can do almost nothing about, I suspect NASA reasons it's better to try and refute this possibility - to avoid alarm.
Anyway, that's the backstory. So why do I think MacGregor et al. have got their sums wrong? This is a tricky discussion, so please bear with me. It appears to me that the authors of this paper were split into two camps; the geologists and the astronomers. I suspect the astronomers were keen to avoid any conclusion which favoured a twin impact, at all costs, since this favoured a comet impact and Clube and Napier's theory. The geologists, on the other hand, didn't have this bias.
The authors present two inconsistent statistical methods for estimating the probability these craters are unrelated. They also provide other arguments, based on the geology and astronomy, as to why they might, or might not, be related. However, ultimately, they conclude that the craters are probably unrelated, and it is their conclusion I disagree with.
Their first statistical method estimates the probability that successive unrelated impacts of the required size could occur within 183 kms of each other. They find this is around 0.02%, a tiny probability. This is close to the correct answer, and their conclusions should have been based on this result.
However, it seems the astronomers developed another statistical method designed to counter this conclusion. Their second method, their so-called 'birthday problem' method, is an estimate of the probability that over the last 650 million years a pair of unrelated asteroids of the required size could fall within 183 kms of each other on stable land. Using a Monte Carlo method, they arrive at an estimate that only 1.4 pairs of craters should be this close on stable land - land stable enough for craters to have survived for 650 million years without being eroded away. In other words, only 1 or 2 pairs of craters are expected to be observed this close together.
But there are several problems with this second calculation. First, we already know of 2 pairs of unrelated craters this close together on land. So, finding a third pair is unlikely. Not impossible, of course, just unlikely.
Even worse, their calculations allow the crater pairs to be of any age, i.e. each crater of the pair can have any age up to 650 million years (so their ages are allowed to differ by 650 Myr). The problem is, in their paper the geologists also state that it is very unlikely that either of these Greenland craters is older than 2.6 million years. So the astronomers' calculation is inconsistent with the geologists' deductions. The reason each crater is thought, almost certainly, to be less than 2.6 million years old is because over this time the Greenland ice sheet has waxed and waned many times, and all this harsh glacial action would have eroded any craters older than that. This is the upper age limit quoted for the 1st crater found, which the geologists state should also apply to the 2nd. But the astronomers argue that this is just an assumption. In essence, the astronomers are arguing that Greenland is sufficiently stable to preserve craters up to 650 million years old, whereas the geologists (who presumably know better) argue that it can only be expected to preserve craters up to 2.6 millions old.
Now, if we accept the geologists' view and limit both crater ages to be less than 2.6 million years AS WELL AS closer than 183 kms on land, the probability they are unrelated falls to 13 x 0.3 x (2.6/650) x (2.6/650) = 0.006%. This means they are extremely likely to be related. And yet the authors conclude they are unlikely to be related. The authors also present other arguments for and against a twin impact. However, none of these other issues are quantifiable. So, they cannot be used to alter this statistical result.
Essentially, it seems the authors could not agree on how stable Greenland is for preserving impact structures. If the geologists are right, as they surely are on this subject, it follows the craters are extremely likely to be related. If, however, the astronomers are right then the craters are still likely to be related, but with much less certainty. Nevertheless, they conclude the opposite.
In a final twist, the geologists present a conference paper at the 50th Lunar and Planetary Science Conference 2019 in Texas, where they reaffirm their view that the 1st crater (and thereby, presumably the 2nd) is less than 2.5 Myr old (see https://cosmictusk.com/)
(Ps: the astronomers didn't need to use a complicated Monte Carlo method to make their statistical estimate - in this particular case a simple analytical method would have worked just as well).
Comments
Post a Comment