≡ Menu

‘Nemesis’ and Orbital Change

The idea of ‘Nemesis,’ a hypothetical dark companion to the Sun, won’t quite go away, and it’s possible that the WISE mission may help us either identify such an object or else demonstrate that it’s not there. The idea is simple enough: Sol’s companion would perturb the Oort Cloud in its orbit, causing comets to enter the inner Solar System, thus increasing the likelihood of an impact with the Earth. Throw in an apparent periodicity in extinction events first described back in 1984 and you have an intriguing case.

But Adrian Melott (University of Kansas) and Richard Bambach (Smithsonian Institution) have reconsidered Nemesis in terms of extinction events in a new paper, one that looks at the timing of these incidents in light of the movements of Nemesis over time. They extend the original 26 million year extinction periodicity slightly, to 27 million years, and are careful to note that there is no consensus on the matter among paleontologists. But the real question they tackle is whether the apparent cycle really can be explained by the actions of a distant, massive object.

This is a useful contribution to the discussion particularly because the authors stretch the span over which periodic extinction events are studied to 500 million years. What they find is that the events seem to occur too regularly over that period to be tied to Nemesis. If that seems counterintuitive, realize that the effects of such an object depend upon the stability of its orbit. Two major causes of perturbation to that orbit have been considered, one the galactic tidal gravitational field, the other the effect of passing stars. Both carry a punch. From the paper:

Hut (1984) was specific that irregularity of the period of revolution of such an object over the past 250 My should be about 20% due to perturbation from the Galaxy tidal gravitational field and by passing stars, and sharp peaks should not be expected in Fourier analysis. Torbett & Smoluchowski (1984) reached the same conclusion, but with a somewhat larger estimate of the fluctuations from the Galactic tide alone, dependent on the inclination of the Nemesis orbit with respect to the Galactic disk. Hills (1984) estimated a period change of 4% per Nemesis orbital period from the effects of passing stars. Using a t1/2 amplitude scaling expected from a random walk, the orbital period should drift by 15 to 30% over the last 500 My. This change in the period will broaden or split any spectral peak in a time series frequency spectrum, so Nemesis as an extinction driver is inconsistent with a sharp peak.

So how sharp is the peak? Sharp enough for the authors to conclude that there is 99 percent confidence in rejecting the hypothesis that there is no association of mass extinctions with the 27 million year cycle. The periodicity is demonstrated for a much longer period of time, its timing revised to roughly 27 million years (over the previous 26 million) and the confidence level in the results has gone from 95% to 99%, surely a confirmation of a cycle, but what is causing the extinctions?

It’s not likely to be Nemesis, argue the researchers:

Fossil data, which motivated the idea of Nemesis, now militate against it and suggest another mechanism is needed to explain extinction periodicity. An attempt to associate the periodicity with passage through the Galactic mid-plane …has its own set of problems: 54 My is rather too short for most estimates of the period of the Sun normal to the plane, and our passage within the last My or so of the mid-plane… is inconsistent with the phase of the 27 My signal we have detected, with its recent maximum at 11 My ago.

This is a fascinating finding. On the one hand, we do see the expected timing for extinction events, but the very regularity of that timing argues against its being the result of Oort Cloud perturbations caused by a Nemesis-like object. Nemesis’ orbit couldn’t be that stable. We are left to ponder the cause of these events, now measured over a span of some 500 million years and found to meet the confidence levels of three different statistical tests.

The paper is Melott and Bambach, “Nemesis Reconsidered,” accepted by Monthly Notices of the Royal Astronomical Society and available as a preprint.


Comments on this entry are closed.

  • Will Baird July 14, 2010, 16:39

    A small note of caution.

    At least three of the ‘spot on’ mass extinctions that occurred on the 27 my basis have radically different causes (KT (bolide), TJ (vulcanism) and Ordovician (glaciation). It’s hard to see how all three could be caused by a regular X reason.

    There is also a VERY large chunk of time (approximately 360 mya to 260 mya) with virtually zilch mass extinctions. A mid Carboniferous extinction is disputed, to put it mildly.

    The strata sampling is really, really important. The more samples we have, the better the accuracy of what we are stating: there have been recent finds of Cambrian critters that were thought to have bought it in earlier mass extinctions that cropped up later (anomolorcarids in the Devonian and even what appears to be Ediacarans up through the Cambrian and, I think, Silurian).

    Finally, a friend that is far smarter than I am pointed out that FFTs can and do give false positives. However, my stat isn’t strong enough to pick that fight.

  • Carl Keller July 14, 2010, 17:50

    Our nearest known star is something to nail down. It’s not certain that Proxima makes the remote component of a triple system. An interesting write-up is in Burnham’s Celestial Handbook, Vol. I:

    “There is some evidence that Proxima is in slow orbital revolution about Alpha Centauri, but the period must be extremely long, perhaps in the neighborhood of half a million years. The actual distance between Alpha and Proxima is approximately one trillion miles, or about 1/6 of a light year. This is an immense distance for any physical pair; it is nearly 300 times the greatest separation of the main pair A and B, and more than 400 times their mean separation. It is approximately 10,000 times the distance which separates the Earth and the Sun.”

    The real dynamics of such a system should eventually come to light. We have a close example of multiple-body interactions, whether it is a triple star system or not. Should there be a Centauri Oort cloud, there will be more for future generations to study in this regard.

  • Denver July 14, 2010, 20:13

    If there is a Centauri Oort cloud, it’s raining there right now.

    The Grand Coupure extinction may be associated with an apparent meteor swarm which occurred ~35My (Chesapeake Bay, Tom’s Canyon and Popigai). That, and the opening of the strait between Antarctica and South America. There is some disagreement.

  • philw1776 July 14, 2010, 20:18

    As Will Baird said, both the evidence of periodic extinctions themselves and the methods of statistical analysis used by the authors are in question. The MIT Technical Review comments section tears apart the flawed statistical analysis.


    Will Baird knows something about mass extinctions, et. al.


  • Zen Blade July 15, 2010, 17:07

    Interesting assertions… but I am not convinced by the data I saw.

    The 27 MYear interval actually looks quite weak to me. Yes, most (not all) of the largest events do appear on the “27 MY” track, but there are a lot of gaps… and the issue of “noise” and “error” on the Proportion of Genus Extinction axis is not known (to me)… If we assume 10% error, and undoubtedly some data points have higher error (20% or more perhaps), then a lot of their data is “in the weeds”. In other words, how do you classify something as a major extinction event.

    Starting from 500 MYA, there are no OBVIOUS outliers until ~450 mya extinction. The earlier events are so rapid that they are likely just by-products of rapid evolution, new species, or some other set of events that were common from ~540-480 mya. So, that time period should not be included.
    So, we have the ~450 event, and then there *might* be an event at ~375. However, that event could really just be a normal event for the 400-320 mya time period…
    So, then there is clearly a doublet (two peaks) at ~260 mya. That’s something distinctly different. So, that means we have a 450, 375, and 260.
    Because of the 260 doublet, the 250 peak is now meaningless. However, without the 260 doublet, the 250 peak would probably be worth mentioning.
    There is clearly an event at 200 mya. But between 200 and 100 mya, there are no events on par with the definitive (in my opinion) events I previously mentioned. Yes, there might be something at 180 and 150… but then we have to include a lot of peaks ignored earlier.

    But, then we get to ~65 mya, and there is another clear event. There is nothing else after that.

    So, 450, 375, 260, 200, and 65…

    Really, this data set looks like it can be mined… but I would not have subscribed to their interpretation of what equals a “mass extinction”. Granted, my research has nothing to do with any of these topics. However, I have generated data that looks “similar” to their second figure, and I would have produced a very different interpretation regarding what was or was not significant. However, depending on what experiments you have conducted and what you are looking for… this could be like comparing apples and hand grenades.

    However, assuming we are simply comparing apples+oranges, I think the first thing I would have looked at is nearby peaks and the relative heights both before and after (and not necessarily just within a couple of million of years, but perhaps 5-10 million years). I bring this point up because either the peaks between 400-350 mya are significant or they are not… However, the peaks between 250-200 mya appear (to me) to be just as significant (or not) as the 400-350mya peaks, yet the authors treat these two sets of peaks differently.

    The only way the set of peaks between 400-350 mya (as well as the peaks ~500 mya) can be significant is if you assume that the valleys (sometimes only a single point) are also significant and accurate. However, if this is true, then the peaks around 360 and 250 mya should also be marked as significant, which they are not.

    Anyways, I think that either they do a poor job explaining why their analysis is a proper one OR that their analysis is not a proper one… and really they are trying to fit their data to a model or hypothesis rather than trying to come up with a model that explains their data.

  • James M. Essig July 15, 2010, 18:28

    Hi Folks;

    There are several potential candidates for Nemesis including all of the following: 1) A cooled white dwarf or a black dwarf; 2) A non-magnetically active neutron star or a neutron star whose rF beam does not weep by Earth; 3)A stellar mass range black hole; 4) A large cold brown dwarf; 5) and a Cold Dark Matter body.

    In addition, perhaps the cause is a non-electrically conducting and oscillating cosmic string that is several to many light years long and which has a vibrational amplitude velocity swing relative to Earth on the rough order of that of the random or semirandom stellar velocities of our local stellar neighboors with respect to Earth.

    One thing that seems extremely likely is that their is infact a causative mechanism for the mass extinctions that operates in a cyclical manner with a period of about 28 Myr.

  • Zen Blade July 15, 2010, 20:48

    Perhaps I should also point out that random events will always lead to the formation of “patterns”. That is, patterns that we deduce that others do not. I don’t think people should take the occurrence of random events to be “a causative mechanism for the mass extinctions that operates in a cyclical manner with a period…” or any sort.

  • Michael Simmons July 15, 2010, 22:46

    It would be that the cyclic event is a minor effect which makes the truly random events much worst. I don’t think anyone is saying all extinction events have a common cause.

    What about something orbiting the super massive black hole at the centre of the galaxy? If we are half way though the cycle it would be on the far side of the galaxy from us and may not be detectable separately from the centre.

  • Tim Whitworth July 16, 2010, 7:26

    How about if instead of one single body, nemsis turns out to be a widely spread swarm of bodies, perhaps gravitationally bound?

  • Zen Blade July 16, 2010, 9:58

    Actually Michael, there clearly were people (and probably still are people) who believe that there is a common cause… hence “Nemesis”.

  • Bynaus July 23, 2010, 8:09

    I wonder if the orbital period of a putative Nemesis could be decoupled from this 27 Myrs extinction cycle (if it exists at all, that is). Yes, the original idea was that Nemesis crosses the Oort cloud every 27 Myrs, resulting in a rain of comets. But maybe this scenario was to simple. Maybe there are two “Nemesis”-type brown dwarfs / giant planets instead, on much closer (and therefore much more stable) orbits, going through a resonance cycle that lasts 27 Myrs. At the peak of the cycle, the orbit of one of the two brown darfs / giant planets gets so excentric it crosses a denser part of the Oort cloud, causing the comet shower…

  • Mike Lorrey July 23, 2010, 18:04

    The orbital stability problem has already been treated by Muller, who found Nemesis’ orbit altered significantly at 400Mya triggering the late heavy bombardment. Muller shows that Nemesis’ current orbit would remain stable for 1 billion years from that point, so it should remain in its current orbit for another 600 million years.

    As for the claims of greater variation in orbital period via random walk, one has to also note that the sheer distance from the Oort Cloud to Earth is so great, and the possible orbital paths so numerous, that the true variation in the periodicity should be due to the variance in the amount of time passing between a Nemesis perigee and a disturbed comet striking Earth.

    There simply isn’t such an exact periodicity of exactly 27 million years, there is a deviation of a few million years, which can easily be explained by both Nemesis variability and the random chance of Earth intercepting an orbit-crossing perturbed comet.

  • Will Baird July 28, 2010, 22:35


    Except for the fact that at least two of the top six mass extinctions had zilch to do with extraterrestrial causation: CAMP and the Siberian Traps are not caused by ‘roids. The 6th mass extinction is not caused by a comet either.

    Only the KT has been finger printed for sure as such. The Devonian is up in the air (there’s multiple pulses, but there are nicely timed impacts for at least one).

    The Ordovician plausibly could be caused by an extended global cooling brought on by bollide strikes, but…there’s not supporting evidence. More likely, it was caused by the weathering of the Appalachians.

    A single cause – which periodicity would require – doesn’t fit the signatures of the extinctions either. Different things kill differently. Volcanoes kill one way. Bollides another. Looking at the patterns of extinction can tell us a lot. In this case, depsite the valiant (?) efforts of a few, there’s not a single cause for the mass dyings.

  • ljk February 15, 2011, 13:08

    So Nemesis is now called Tyche – will that make a difference in its reality?