plot of impact rate
Table of contents
When an extraterrestrial impact takes place on the Moon, some of the target rock vaporizes and recondenses into small droplets of glass called spherules. These land from a few meters to thousands of kilometers from the impact site. As a consequence, every gram of lunar soil contains spherules from a hundred or more impacts. These spherules are only 150 to 250 µm in diameter, but modern methods of 40Ar/39Ar dating can be used to determine their ages. For images and a better idea of what spherules are like, we have posted detailed information for two of our spherules, Lu09 and Lu25. A dramatic image of an Apollo-11 spherule, used in our LBL press release, is also posted. It is 250 microns = 1/4 mm in diameter, about the size of a grain of sand.
In 1993, I speculated that most of these spherules would come from different craters [1], a result that was later confirmed. Thus, even though we don't know which crater was the source of each spherule, the distribution of the ages of the spherules from a single lunar site reflects the age distribution of craters on the Moon.
We have measured the ages of 155 spherules from the Apollo-14 site [2], and from these determined the time variation of cratering on the Moon - and presumably also on the Earth. The results show a gradual decrease in the cratering rate from 3 Gyr until about 0.4 Gyr, when the rate increased by a factor of 3.7 ± 1.2, back to the level that it had been 3 billion years earlier. Statistical analysis indicates that approximately 94% of the spherules come from different craters [3], a result that is confirmed by chemical analysis [4]. Potential biases from lunar gardening and nearby craters have been analyzed, and appear to be negligible.
Approximately 1 gram samples of lunar soil were obtained from NASA, from each of three lunar sites: Apollo-11, 12, and 14. The spherule ages were measured at the Berkeley Geochronology Center using the 40Ar/39Ar isochron technique. Measurements of the spherule ages from Apollo-11 and 12 yielded inaccurate dates, due to low potassium content. Based on a chemical analysis of these spherules, we concluded that most of the spherules came from local impacts [4]. To improve the accuracy, we obtained lunar soil samples from the Apollo-14 site which was known to have high potassium content. This was indeed reflected in the spherules chemical composition, and led to improved age accuracy. The median age uncertainty for our 155 spherules had a one standard-deviation error of 0.15 Gyr.
The distribution of ages for these spherules is shown in the figure. The histogram shows the number of spherules with ages in each 0.4 Gyr bin. The number of spherules in each bin is printed above the histogram. At the bottom of the plot we have drawn 155 Gaussian curves, one for each spherule. Each Gaussian has unit area, but the width (RMS deviation) for each is set to the 1 standard deviation error in the age uncertainty for that spherule. The sum of the individual Gaussians is shown as the smooth curve. This curve represents the best estimate that we have, in a statistical sense, for the cratering rate. Note, however, that the sharp peaks are artifacts of several highly-accurate ages.
plot of impact rate
The gradual increase in impacts from 4.5 Gyr to 3 Gyr does not imply an increasing impact rate, but is instead a result of the age of the Apollo-14 site, estimated at 3.85 Gyr. The spherules older than this were probably first deposited at distant sites and then transported to the Apollo-14 site by subsequent impacts. The decrease from 3 Gyr to 0.4 Gyr probably reflects a decrease in the density of potential impactors (comets and asteroids) in the solar system. The comets and asteroids causing the impacts were slowly being eliminated by Jupiter, which deflected them into the sun or out to infinity. The observed decrease is consistent with previously reported estimates. [5]
The most surprising result is the increase seen in the most recent bin (the one on the right, with 26 events), representing the last 0.4 Gyr. We estimate that the rate increased (compared to the previous 1 Gyr) by a factor of 3.7 ± 1.2, to a level comparable to the highest it had been at in the previous 3 Gyr. (Note that prior to 3.5 Gyr, the rate is believed to be much higher, but that fact doesn't show in our plot because of the relatively young age of the site.) A similar increase of a factor of two had previously been suggested based on measurements on the Earth and crater counting on the moon. [6-9]
To test if the recent increase is due to an experimental bias, we analyzed several possible systematic errors. The increase can not be accounted for as an excess due to the nearby Cone crater, both because of the diversity of chemical composition, and the incompatibility of ages.[2] Of the 11 spherules compatible with the Cone crater age (25 Myr), 3 are black, 3 are yellow, 3 are orange, 1 is green, and 1 is white. A statistical analysis of the spherule ages [3] shows that we expect 94% of the spherules to be from different craters, i.e. a total of 9 spherules in the entire distribution come from the same craters as others. (For a pdf version of this reference, click here.) It is even less likely that all nine would be in the same bin. Lunar gardening could, in principle, distort the distribution. However the sample was collected from the ejecta blanket of Cone crater, and represents a well-mixed sample of lunar soil. Removal of all 11 recent spherules (0-25 Myr) would completely eliminate this systematic effect, if it were there. That is probably an overcompensation for the bias, and it leaves 15 spherules in the bin, still a significant increase.
The increase in cratering was roughly coincident with the Cambrian explosion of complex life on Earth. As Darwin originally conceived evolution, survival of the fittest meant superiority in conflict with other species. But an increased crater rate also increases the stress on species to be able to survive catastrophe. It is conceivable that this increased stress helped drive evolution towards species with greater adaptability and flexibility, away from instinct and towards greater intelligence.
Implications for the Solar System:
The lunar spherule project was originally conceived [1] as a method to search for periodicities in comet showers, a prediction of the Nemesis theory [10]. This theory postulates that there is a companion star to the Sun, orbiting at a distance of about 3 light years, with a period of 26 million years. There has been a great deal of speculation that the orbit is insufficiently stable for the theory, but a detailed analysis by Hut has shown that the stability is adequate [11]. The strongest argument against the Nemesis idea is the fact that it predicts that most of the mass extinctions of Raup and Sepkoski [12,13] should have been caused by impacts, and little evidence has been adduced since 1984 in favor of this conclusion. For more on the Nemesis theory and its current status, click here.
Unfortunately, the age accuracy achieved thus far in the lunar spherule project is insufficient to see a 26 Myr cycle. The median age uncertainty is 150 million years; a factor of 10 improvement is necessary to see the predicted cyclicity. Nevertheless, the Nemesis hypothesis provides a ready explanation for the 0.4 Gyr increase in cratering rate. If the orbit of Nemesis were nearly circular prior to this time (eccentricity < 0.5), then its orbit would not bring it sufficiently close to the inner Oort comet cloud to trigger periodic comet showers. But in its large orbit, perturbations from passing stars are frequent. We speculate that Nemesis was perturbed into a more eccentric orbit (eccentricity > 0.7) at 0.4 Gyr. This would be adequate to cause it to trigger comet showers at every subsequent perihelion, and that could account for the recent increase in impact rate.
Ultimately, of course, the existence of Nemesis must be confirmed by its direct observation. In the original theory, we assumed that Nemesis is a red dwarf star, and should be readily visible from the Earth. The Hipparcos satellite, unfortunately, surveyed only about 1/4 of the known candidates. Future parallax surveys, if they reach stars as dim as 10th magnitude, should find the star if it is there, or prove (by lack of discovery) that it is not there.
Implications for future measurements:
The spherule method has proven extremely successful. The increase in the last 0.4 Gyr has important implications, so it is important that it be confirmed at a different lunar site. This requires either another high-potassium site, or the use of larger spherules. Use of such spherules may also allow the detection of comet showers and a search for periodicity in impacts. The absence of such periodicity would disprove the Nemesis hypothesis. Measurements in the lunar highlands could potentially determine older cratering rates, and answer lingering questions about the existence and intensity of the late heavy bombardment.
[1] Richard A. Muller (1993), Cratering rates from lunar spherules, Lawrence Berkeley Laboratory Report LBL-34168. For a pdf version, click here.
[2] Timothy S. Culler, Timothy A. Becker, Richard A. Muller, and Paul R. Renne, (2000), Lunar impact history from 40Ar-39Ar dating of glass spherules, Science, issue of March 10.
[3] Richard A. Muller, Timothy A. Becker., Timothy S. Culler, Paul R. Renne (2000), Solar System impact rates measured from lunar spherule ages, in Accretion of Extraterrestrial Matter Throughout Earth's History, edited by B. Peucker-Ehrenbrink and B. Schmitz, Kluwer Publishers, in press. For a pdf copy of this paper, click here.
[4] Timothy S. Culler and Richard A. Muller (1999), Use of surface features and chemistry to determine the origin of fourteen Apollo 11 glass spherules, Meteoritics and Planetary Science (submitted).
[5] BVSP (1981), Basaltic Volcanism on the Terrestrial Planets (Pergamon, New York).
[6] McEwen, A.S. et al. (1997) JGR 102, 9231-9241.
[7] Grieve, R., and E. M. Shoemaker, (1994), in Hazards due to Comets and Asteroids, T. Gehrels ed., Univ Arizona Press, pp 417-462.
[8] Shoemaker, E. M., et al. (1990), in Spec. Pap. Geol. Soc. Am., L. Sharpton and P. Ward eds, vol 247, pp 155-170.
[9] Shoemaker, E. M. et al. (1997), in Hazards due to Comets and Asteroids, T. Gehrels ed., Univ Arizona Press, pp. 313-336.
[10] Davis, M., P. Hut, and R. A. Muller (1984), Extinction of Species by Periodic Comet Showers, Nature 308, pp 715-717.
[11] Hut, P. (1984), Nature, vol 311, pp. 636-640
[12] Raup, D., J. Sepkoski (1984) Proc. Nat. Acad. Sci. USA, vol 81, pp 801-805.
[13] Sepkoski, J., J. (1989) Geol. Soc. London, vol 146, pp 7-19. For some of his figures, click here.