15

FLINGING COMETS FROM THE OORT CLOUD

You might have seen the coordinated dances performed by the Rockettes at New York City’s Radio City Music Hall or by ensembles on some old TV shows, in which a large number of beautifully dressed women synchronize graceful motions around a circle. Some of the formations involve spokes of dancers emanating from a common center, while others are composed of individuals forming rings—one inside the other. The performers seamlessly keep their circles intact, making it easy to forget how difficult the precise relationships among the individuals are to maintain. This applies especially to the outer members, who need to move more quickly and are also more distant from the inner region from which the coordination and directions emerge. You might occasionally see a dancer in the farthermost ring, who faces these greater challenges, mess up and get out of sync. But so long as she doesn’t fall, it’s not such a big deal. Though the error will take away from the performance’s beauty and perfection, which lies in the dancers’ synchronicity, nothing dramatic or disastrous ensues.

Icy bodies in the Oort cloud—tens of thousands of times farther away than the Earth is from the Sun—face challenges akin to dancers in the outer ring. Its members are so remote from the Sun’s gravitational pull that they are in a relatively precarious equilibrium. A sufficiently strong disturbance can cause a body, like the less precise outer dancer, to slowly move out of its expected position. If an Oort cloud object comes too close to the inner solar region, enough nudges—or one single more dramatic push—will send it out of its orbit altogether. When this happens, that body will deviate far more from its path than the errant dancer, running the risk of hurtling toward the inner Solar System and possibly even toward the Earth.

Near-Earth asteroids as well as some errant short-period comets can also be jostled by planets or other local objects so that they occasionally hit the Earth. But such impacts are almost certainly random. Mechanisms for triggering periodic disturbances have been proposed only for comets from the Oort cloud. The Oort cloud, the lone source of long-period comets entering the Solar System and also probably the source of most comets coming near the Sun, is also the only suggested source of regularly spaced comet strikes. The suggested periodicities in the extinction and crater records that we considered in the previous chapter have made for a good deal of interest in identifying what might trigger disturbances that could regularly send the Oort cloud’s icy bodies into the inner Solar System.

In this chapter, I’ll first briefly address the issue of whether comets or asteroids are more likely to have created big impacts. I’ll then review some of the original proposals for what might dislodge objects from the Oort cloud to create comets that can impact the Earth. Although these older ideas failed to account for the suggested regularity, they are nonetheless interesting in that they encouraged new ways of thinking about galaxy interactions. They also paved the way for our later, more promising proposal based on our newly suggested dark matter idea.

ASTEROIDS VERSUS COMETS

If the impactor responsible for Chicxulub was caused by an asteroid, dark matter had nothing to do with it. But if it were a comet that wreaked the devastation, an exotic dark matter trigger just might have been the culprit. In his book T. rex and the Crater of Doom, Walter Alvarez used “comet” as his default when discussing the impactor responsible for the K-Pg extinction, with the understanding that no one could definitively establish whether a comet or an asteroid had been responsible. Distinguishing the effects of comets and asteroids that have created craters—particularly those that fell to Earth millions of years ago—is difficult. If no one observed the trajectory, we usually have no way of knowing whether an object that hit was a comet or an asteroid. For the comet that destroyed the dinosaurs, the jury is still out.

We do know that comets and their fragments hit the Earth less often. Estimates of the relative frequency of comet impacts compared to those of asteroids range between 2 and 25 percent. This small rate corresponds to the low number of near-Earth comets. Of the more than 10,000 near-Earth objects that are now known, only about 100 of these are known to be comets, with the rest consisting of asteroids or smaller meteoroids.

But larger impacts don’t necessarily arise solely from objects that are already nearby. Distant comets can sometimes escape their orbits to occasionally hit the Earth too. An intriguing study by the extraordinary astronomer Gene Shoemaker argued that although asteroids predominate in smaller impacts, comets might be more important for larger ones. Shoemaker plotted the number of impacts versus size, and found that the results seemed to depend on two separate populations. The smaller impacts all fell along a nice curve, but there were many more large impacts than this simple curve could accommodate. Knowing that asteroids produced the smaller impacts, Shoemaker hypothesized that a new source of impacts must have produced the larger ones—arguing that what he was witnessing was the sum of two separate curves representing two independent contributions. His guess was that the source of the larger impacts was comets.

Comets have the further feature that they carry a disproportionate amount of energy compared to asteroids, since they are generally moving at faster speeds—up to 70 or more km/sec as opposed to 10 to 30 km/sec for asteroids. Typically, a ballistic missile travels at less than 11 km/sec, an asteroid at about 20 km/sec, a short-period comet at more like 35 km/sec, and a long-period comet at 55 km/sec, though faster speeds occur too. (See Figure 32.) Kinetic energy grows not only with mass, but also with the square of the velocity. Comets’ greater speeds mean that even less frequent comet impacts, or ones from smaller objects, could in principle do greater damage than slower-moving asteroids.

[FIGURE 32] Average velocity of impacts on the Earth from asteroids, short-period comets, and long-period comets in kilometers/second. The curve also illustrates the expected relative fluxes of the three types of objects.

Shoemaker furthermore did chemical analyses that supported the comet proposal—though in fairness it should be noted that scientists doing such analyses have argued both ways. In favor of the competing asteroid hypothesis are isotope ratios and surviving meteorite fragments matching those of chondritic asteroids, which contain millimeter-size spherical pieces that were once molten droplets created in nebular storms 4.56 billion years ago, during the Solar System’s formation. But the evidence is not yet decisive. We don’t know the isotope ratios in comets, so we might find that they are similar too. Furthermore, more recent research argues for a lower iridium and osmium level than had formerly been believed, which would be more consistent with a comet interpretation.

In 1990, the astrophysicists Kevin Zahnle and David Grinspoon argued for a comet impact at Chicxulub using very different reasoning. They proposed that comet dust entered the Earth before and after the K-Pg extinction event in order to explain the amino acids found in the sediments surrounding the K-Pg layer. Since dust particles get suspended in the atmosphere and fall slowly and thus would reach the ground intact, the dust could in principle be the result of a comet that disintegrated over a long period of time—raining material down onto the planet.

One reason that comet strikes might happen more frequently than would otherwise be expected is that when Jupiter swings comets around, it sometimes breaks them up into fragments. If and when this happens, the likelihood of connecting with Earth increases since several of these fragments might then cross the Earth’s orbit. Some astronomers speculate that this occurred as recently as a few thousand years ago and cite excessive comet dust in the inner Solar System as evidence.

The relatively fresh Shoemaker-Levy comet hit on Jupiter was a spectacular illustration of the destruction such comet fragments can do. Carolyn Shoemaker first spotted the comet near Jupiter in 1993, and followed it along with her husband, Gene, and another colleague, David Levy. They noticed that the comet had an unusual appearance, appearing not as a single streak in the sky but as an arc punctuated by spherical bright spots. Soon afterward, through a more precise observation, the astronomers Jane Luu and David Jewitt were able to identify at least 17 separate pieces that spanned an arc resembling a string of pearls.

Astronomer Brian Marsden from the aptly named Central Bureau for Astronomical Telegrams deduced from its trajectory that its unusual structure was the result of a too-close flyby of Jupiter, whose gravity broke the comet into smaller fragments. He suggested the possibility of a future Jupiter close approach or even impact. Astronomers followed up and calculated that Jupiter’s gravity would indeed trap the pieces, which would return for a head-on collision between July 16 and July 22, 1994.

Sure enough, right on schedule, the first fragment dove into Jupiter’s atmosphere with a speed in excess of 60 km/sec. The region that was visibly affected was at least the size of the Earth. The atmosphere was illuminated by dust that preceded the actual fragments, which themselves created a brilliant flash. This generated effects similar to those surrounding Chicxulub—but this time the damage occurred on Jupiter. Since the fragments were less than 300 meters across and the initial comet that created the fragments was at most a few kilometers big, the energy released was far less than that from the object that created Chicxulub. Nonetheless, it was an impressive sight.

Impact craters on the Jovian moons indicate this was not the first time that such dramatic capture breakups and impacts have occurred in the region. And, if the periodic meteoroid idea turns out to be correct, it will be further evidence that comets have been important throughout the Solar System’s existence. The association of such astrophysical phenomena with planetary surfaces reminds us that even seemingly abstract theoretical research might ultimately help explain our own existence.

TRIGGERS

Although no one can be certain, I will assume for the rest of this book that comets from the Oort cloud are responsible for big impacts. It is the only known possibility for which we can potentially explain periodic hits. Although a perturbation of an icy body from the outer Solar System that sends it toward our planet’s path might sound like science fiction—not mistakenly so, since it often is—this sequence of events is also science.

Recall that the furthest reaches of the Solar System contain the Oort cloud—a hypothesized somewhat spherical collection of minor bodies that might extend beyond 50,000 times the Earth–Sun distance. The evidence for the existence of this huge source of comets—too far away to directly observe—is precisely the visible comets that have entered the inner Solar System.

In contrast to the situation for the dancers mentioned earlier, the pull of the Sun—and not the mutual interaction among Oort cloud objects—is responsible for keeping Oort cloud icy bodies in their orbits. But the Sun only weakly gravitationally binds the objects into the cloud, which is so enormously far away. Gravity’s strength decreases according to the inverse square of the distance, so its influence on an object tens of thousands of times further away is less than 100 million times less strong. The Sun’s pull on a comet in the Oort cloud is that much less strong than the Sun’s pull on the Earth. In such a loosely bound environment, even relatively small disturbances can potentially alter an Oort cloud object’s trajectory, ultimately kicking it out of its orbit—dispatching it out of the Solar System altogether or sending it hurtling down a path inward toward the Sun.

Though the astronomer Jan Oort later put the idea on firmer footing, the Estonian astrophysicist Ernst Julius Öpik proposed in 1932 that perturbations to comets at the outer edge of the Solar System (in what is now known as the Oort cloud and occasionally the Öpik-Oort cloud) sometimes send those icy bodies toward the inner region of the Solar System. Öpik had the whole story essentially correct—reasoning that some icy bodies would eventually become unstable and vulnerable to perturbations so that external influences would sometimes nudge them out of their orbits and into a path that was heading toward the Earth. He even suggested this could influence life here, though he didn’t necessarily envision global devastation of the sort accompanying the K-Pg extinction.

Öpik’s impressive work nonetheless left open the question of why the orbits became unstable or what the trigger was that precipitated their escape. These questions wouldn’t be addressed until many years later, when the Alvarez proposal (and the Cold War with its images of massive devastation) entered the public consciousness and resuscitated interest.

Examples of objects that astronomers have suggested as perturbations include nearby passing stars and giant molecular clouds—enormous concentrations of molecular gas with mass between 1,000 and 10,000,000 times that of the Sun. But although the former jostles the orbits and the latter has some effect, neither is the dominant mechanism for sending comets en route to the inner Solar System. A nudge’s influence depends on the magnitude and frequency with which it occurs as well as the density and mass of the icy bodies it acts on. Neither stars nor molecular clouds have both sufficient force and frequency to explain all the comets that we see.

In 1989, Julia Heisler and Scott Tremaine investigated a far more significant influence, which is the tidal force of the Milky Way. The Moon creates the familiar ocean tides through its distorting gravitational influence, causing the ocean to rise and fall by pulling differently on farther or closer regions of the Earth. In a similar manner, the galactic tide caused by the Milky Way bends the orbits of outer Solar System objects. The gravitational pull of the Milky Way acts differently on objects that are not in precisely the same location, deforming the otherwise spherical Oort cloud so that it is elongated toward the Sun and compressed along the other two directions.

Over time, the gravitational force from the Milky Way will tweak the paths of minor bodies into very elongated, or eccentric, orbits. Once sufficiently eccentric, the perihelion—the distance of closest approach to the Sun—will be so small that objects can be more readily injected into the inner Solar System. The tidal force at this point can be sufficient to send icy bodies out of the Oort cloud to increase the comet flux in the interior. The result is a slow and steady flux of comets reaching the Earth.

To make matters more interesting, it turns out that the dominant mechanism for dislodging icy bodies to send comets into the inner Solar System is not solely dependent on the tides, but rather involves both stellar and tidal perturbations working in concert. Though stellar perturbations are not ultimately the ones that usually create the comet showers, since they occur over much larger time scales than the tidal influence, they are essential for prepping the Oort cloud to a point where a tidal interaction can be pivotal. It’s like a bicycling team in the Tour de France. The rest of the team helps the lead rider position himself so that he can make the final end run that earns him his yellow jersey. Because he crosses the finish line first, we generally only know the name of the winner—not the supporting domestiques. Even so, the other riders played an important role. Similarly, although the immediate trigger for dislodging the comets is the tidal force, the reason it can exert sufficient pull is that stellar perturbations have already jostled the orbits sufficiently that some are in precarious positions where only a relatively minor nudge will send the comet into the inner Solar System. Stellar encounters are essential, but the actual trigger for the comets—the one that gets the credit—is primarily the tidal force.

The distance at which the galactic tide dominates over the Sun’s gravitational inward pull is about 100,000–200,000 AU from the Sun. At the outer boundary of the Oort cloud, the Sun’s gravitational influence no longer suffices to maintain stable orbits. We have just seen how, farther in, tidal effects perturb borderline stable orbits, occasionally dislodging a minor Solar System body and sending it to the inner Solar System. Closer in still—in regions that have been accessible to observations—tidal effects pale in comparison to the Sun’s gravitational pull. So it is only in the Oort cloud that tidal effects can jostle weakly bound comets significantly. And in all likelihood these tidal effects are responsible for 90 percent of the comets that originate there.

So the Milky Way contains the means to disturb comets to send them into paths toward the inner Solar System through a gravitational influence that physicists and astronomers now understand. But this mechanism—although important and interesting—doesn’t suffice to account for all comet showers or periodicity in comet strikes. In the absence of additional complications, the tidal force I’ve described leads only to a slow but steady stream.

Astronomers trying to explain periodic enhancements therefore made further speculative attempts to explain why the triggers for these comets might not be completely random, but would instead occur at regular intervals in the range of tens of millions of years. I will say up front that the proposed explanations I’m about to present did not succeed. But understanding these suggestions and why they failed helps guide the search for alternatives. One of these suggestions was the precursor to the dark disk proposal that I’ll later describe.

THE NEMESIS PROPOSAL

The first—and the most colorful-sounding—suggestion for explaining periodic impacts was that the Sun had a companion star playfully termed Nemesis, and that Nemesis and the Sun orbited in a big binary system. Astronomers proposed a very elliptical orbit for this hypothetical companion to the Sun that would allow it to pass within about 30,000 AU of us every 26 million years. This 1984 proposal was an attempt to account for Raup and Sepkoski’s suggested extinction periodicity by Nemesis’s enhanced gravitational force on the Sun every 26 million years when it was closest. The suggestion was that at those times, Nemesis’ gravitational influence would dislodge from the Oort cloud minor Solar System bodies that could then bombard the Earth as comets.

The roughly 30-million-year period for the enhanced encounters (and hence a heightened rate of comet strikes) requires a very big system, with a semi-major axis (half the length of the ellipse) of order one or two light-years. One problem with this proposal is that stars or interstellar clouds would make such an enormous binary system unstable, destroying the regularity of the presumed encounters and causing the rate to vary over the last 250 million years. This variation hasn’t been seen.

But the real nail in the coffin for this idea is the much-improved infrared survey catalog of objects in the entire sky—which would now include Nemesis had it existed. Although measurements were inadequate in 1984 to decisively rule on the suggested object’s existence, observations have dramatically improved since that time. NASA’s Wide-Field Infrared Survey Explorer, which was launched in 2009 and collected relevant data until February 2011, should have seen this proposed red dwarf–type star had it existed—but it didn’t. Having not seen a proposed Jupiter-sized gas giant planet either, the infrared observations also ruled out another similar proposed explanation based on a hypothetical new planet, which the idea’s originator named Planet X.

PROPOSED TRIGGERS FROM GALACTIC MOTION

In light of these failed ideas, a few very different proposals based on the Solar System’s motion through the galaxy’s known components seemed like promising alternatives. These proposals didn’t introduce anything new and exotic, but instead suggested that existing density variations that the Solar System encounters when passing through the galactic spiral arms or in crossing the galactic plane could induce variations in the Oort cloud perturbation rate. These repeated passages through high-density regions could in principle account for periodic comet showers.

Recall that the Milky Way is a disk galaxy, meaning most of the stars and gas lie in a thin disk, about 130,000 light-years across but only roughly 2,000 light-years in thickness. The Sun is located at a distance of about 27,000 light-years from the galactic center, and happens at this moment to be close to the galactic midplane—less than 100 light-years away. It is also at the edge of a spiral arm.

The spiral arms of the Milky Way extend from the galactic center in the radial direction as they wind around. (See Figure 33.) These spiral arms contain more gas and dust than the regions in between, and consequently are areas where young stars are more likely to form. They are also the site of an enhanced concentration of giant molecular clouds—the enormous concentrations of molecular gas mentioned earlier. When the Sun crosses these denser regions, the molecular clouds exert a greater gravitational force that could in principle cause more extensive perturbations and thereby generate a periodic enhancement in impacts.

One potential problem with this proposal is that the spiral arms don’t exhibit perfect symmetry or even have a fixed rotation rate relative to the Sun. Therefore, the Sun probably doesn’t cross them at a precisely periodic rate. However, since the structure, kinematics, and evolution of the spiral arms are currently only poorly understood, any conclusion ruling out the spiral arm option on this basis alone might turn out to be premature. In any event, until periodicity is better established, this lack of perfect regularity in the predictions doesn’t necessarily rule out a match to the data, which might turn out to exhibit only approximate periodicity too.

[Figure 33] The spiral arms of the Milky Way with the location of the Sun (size not to scale) indicated.

However, two other factors make spiral arms a poor explanation for any observed enhancement in the impact rate. The first is that the average density of gas in the spiral arms is not elevated enough to account for periodic impact enhancements. If the density doesn’t change by enough, any enhancement during spiral arm crossings will be too small to register.

The further issue is that the Solar System doesn’t cross the galaxy’s spiral arms all that often. There are only four big spiral arms and maybe two smaller ones and a galactic year is pretty long, so there have been fewer than four crossings of the larger spiral arms in the last 250 million years. In fact, because the arms move in the same direction as the Solar System (though at different speeds), the crossings are probably 80–150 million years apart—far too rare to explain the record of extinctions or impact craters.

However, the failure of spiral arms to explain the period and periodic enhancements doesn’t rule out vertical variations in density as a potential impact trigger and this proposal might well prove to be the more promising suggestion. Superimposed on its circular motion, the Solar System oscillates in the vertical direction in which it covers a much smaller distance (compared to the 26,000 light-year radius where the Sun is located along the plane), as is illustrated in Figure 34. As it orbits the galaxy in a roughly circular orbit, taking about 240 million years to complete the circuit in what is known as a galactic year, the Sun also bobs slightly up and down. This much smaller oscillation amplitude in the vertical direction of the Sun depends on the matter distribution in the disk, but a reasonable estimate is roughly 200 light-years—though we are currently much closer to the midplane than the maximum height, perhaps 65 light-years away.

This oscillatory vertical motion of the Solar System could potentially account for variations in tidal effects over time and thereby explain any periodic effects on the appropriate time scale. Because the concentration of stars and gas changes as the Solar System moves in and out of the somewhat denser region of the galactic midplane, the Solar System encounters different environments as it oscillates across it. If the density were to increase dramatically as the Solar System crossed the plane, the perturbations would as well, and consequently the rate of comets hitting the Earth could be enhanced at those times. Since galactic tides are the dominant perturber of the Oort cloud, density variations in the vertical direction within the galactic plane could potentially have sufficiently strong influence. The New York University professors Michael Rampino and Bruce Haggerty, who made this suggestion, gave it a colorful name too—the Shiva hypothesis—after the Hindu god of destruction and renewal.

Two features of the matter distribution in the galaxy are required for this scenario to match observations. First, the midplane density has to provide a gravitational potential that accounts for the correct oscillation period in the vertical direction. This condition is independent of any precise perturbation mechanism. If the Solar System doesn’t cross the midplane at the correct rate, any enhancements at those times won’t match the data.

The second feature is the one necessary to achieve the change in rate that could account for periodic comet showers—namely a sufficiently marked variation in density that would lead to a time-dependent influence on the Oort cloud as it passes through the galactic plane. These two features are relevant to any proposal of density enhancement at the galactic midplane. They rule out the proposals discussed here, and—as I will later explain—account for why the disk of dark matter, denser and thinner than the usual ordinary matter disk, might be a suitable alternative.

However, in 1984, Rampino and Stothers, relying on a more standard Milky Way composition, tried to explain the requisite density variations with giant molecular clouds—which are densest near the galactic midplane. Their reasoning was similar to that used in the spiral arm crossing suggestion—the matter concentration increases when the Solar System goes through the clouds. This proposal was quashed the following year when astronomers showed that the cloud layer is too big—it extends almost as far as the Sun’s vertical oscillation amplitude, so the variation along the Sun’s trajectory would be too small to register. Without additional matter, the encounters with molecular clouds are in any case too infrequent to account for an approximately 30 million year periodicity.

An alternative possibility was explored by Julia Heisler and Scott Tremaine—this time working together with the astrophysicist Charles Alcock. Having established the significance of the tidal influence from the Milky Way, they pointed out that although this effect alone predicted a fairly uniform comet rate, a kick from a nearby star did have the potential to create a comet shower. The question then becomes how frequently such encounters might occur and with what impact. How much of a variation in the rate of comets hitting the Earth should we expect?

The team estimated the expected rate by asking how often a star of a solar mass (the minimum mass needed to have the necessary impact when moving about 40 km/sec) comes within about 25,000 AU of an Oort cloud object (the minimum distance required to perturb it, since it is comparable to the distance of the Oort cloud from the Sun). It turned out one such encounter is expected every 70 million years. That isn’t often enough to account for the suggested periodicity, but it could in principle account for a few such events within the last 250 million years.

Heisler and collaborators subsequently did a more extensive numerical simulation to make better predictions—taking into account the extra push the tidal force could provide. They found that stars had to be a bit closer than the Sun than they had previously thought. The real rate of predicted showers is therefore even smaller—more like once every 100 million or even 150 million years—far too infrequent to account for any periodicity that might have been observed. Subsequent, more detailed numerical analysis found the role of stellar encounters for triggering impacts was greater than they had found, but it was still insufficient to explain the data.

The conclusion of all this research is that without any new ingredients, the Solar System’s gravitational potential doesn’t change dramatically enough over a short enough time frame to yield an observable difference in meteoroid strikes in which the rate at regular intervals would have demonstrated an observable spike that dominated over the background rate. Although the Solar System crosses the galactic midplane on a periodic basis, comet showering due to a conventional matter distribution is not particularly elevated at those times.

So in broad overview, the situation turns out to be reminiscent of that with the spiral arms. The period predicted is too small and the change in density not sufficiently pronounced to give rise to a measurable periodic cratering of the sort the proposers had hoped to explain. Initial density measurements had suggested otherwise, but later calculations taking into account more recent data about the galaxy showed that suggestions that predated our work didn’t generate the right frequency or correct episodic enhancement to match the crater record. The too-long prediction for the period rules out all the galactic plane proposals unless some new and as-yet-undetected component of matter is present in the disk.

Putting together the best available measurements—which, like the evidence for periodicity itself, changed a lot over time—Matt Reece and I ultimately concluded that without a component of so-far undetected matter in the disk, the up-and-down oscillation period was too long to account for the data suggesting periodicity. Not only was the distribution too smooth to generate an impulsive change in the cratering rate, but the familiar Milky Way disk, if composed only of normal matter, is too diffuse to give rise to the correct periodicity.

Though not in themselves sufficient to explain any potential periodicity, the above proposals nonetheless taught me and Matt the fundamentals we needed in order to proceed. We learned that tidal effects create sufficient perturbations to drive comets into the inner Solar System during and near disk crossing. But we also learned that known astrophysical sources do not create the hoped-for periodic effects. None generate a sufficiently abrupt tidal influence to account for an enhancement in comets reaching the Earth.

This left us with two possibilities. Perhaps the more likely one is that the observed periodicity isn’t a real effect. The evidence is not all that strong in the first place and many accidents can conspire to give the appearance of a periodic effect. The second, more speculative, but far more interesting, option is that the structure of the galaxy is different from what is commonly assumed, in which case the tidal effect could be bigger and more dramatically varying than anticipated. This was the path we decided to explore. And it paid off.

As the next part of the book explains, when Matt Reece and I accounted for what is known about the density of ordinary matter in the Milky Way plane and the measured position and velocity of the Sun, we found that agreement with the crater record fared better when confronted with our proposed dark matter model. A dark matter disk in the usual Milky Way plane with the appropriate density and thickness could adjust the predicted magnitude and time dependence of the galactic plane’s tidal force so that both the impact period and the trigger pulse match the data reasonably well.

As a nice bonus, in this way of thinking the look-elsewhere effect of the previous chapter is less compromising than previously thought. We no longer have to think about all possible periods—but only those that take account of the measured ordinary matter density in the galaxy. Armed with the admittedly imprecise measurements of the Solar System and a suitable model of the dark matter disk, we can restrict the range of possible oscillation periods to be only those predictions consistent with existing Milky Way disk density measurements. Matt and I found that with existing data taken into account, the periodic assumption had roughly three times the likelihood of random hits. Though not strong enough statistical evidence to establish the existence of the dark matter disk that we had proposed, the result was promising enough to merit further study.

The best part of this approach is that our knowledge of the galaxy’s gravitational potential will continue to improve. Our method, which takes account of all the available information about the galaxy, will become increasingly reliable as more accurate data about the galaxy and the Sun’s motion is collected. Scientists today are measuring the distribution of matter in the galaxy. Current satellite observations are recording the positions and velocities of stars, helping us to infer the gravitational potential they experience—namely the potential that binds them to the Milky Way. This in turn will tell us more about the structure of the galactic plane.

In what promises to be some truly exciting results, theory and measurements will tie together Solar System motion with data here on Earth. More data in the future will lead to more reliable predictions, making for yet more trustworthy results.

The next part of the book returns to dark matter models, and closes with the particular model that might explain periodicity in the crater record. The study of periodicity and the Earth’s history is an excellent justification to explore—both our immediate visible vicinity and the more ethereal world of dark matter—allowing us to consider the remarkable possibilities for what might be out there invisibly populating our Universe.