REVIEWSOF GEOPHYSICS,VOL. 25, NO. 2, PAGES277-284, MARCH1987 U.S. NATIONAL REPORT TO INTERNATIONAL UNION OF GEODESY AND GEOPHYSICS 1983-1986 The Moon GRAHAM RYDER Lunar and Planetary Institute, Houston, Texas INTRODUCTION on the origin of the Moon have been written [Wood, 1986; Boss, 1986a; Stevenson,1986]. S.R. Taylor [1982] wrote a book that summarizedthe Moon The collisional ejection hypothesisinvokes a catastrophic as we knew it at the start of the quadrennium.Since then lunar collison late in or after the formation of the Earth, ejection sciencehas continued (perhapsto the surpriseof many), and of material to produce a disk of material (not a fully-formed provided us with many new ideas and "facts" about the Moon Moon), and a coalescenceof this material to producethe Moon and its past and presentenvironment. Lunar scienceis not [e.g., Wood, 1986; Hartrnann, 1986]. There are severalvariants moribund. Some aspectshave receivedinsufficient attention of this basicmodel: for instance,the collisionneed not be singular despitethe potential for advances.For instance,there has been [e.g., Ringwood, 1986a,b]; and Boss[1986b] suggestedthat spin little experimentalphase petrology relevant to lunar rocks.Some up of the Earth by an oblique large-bodyimpact could cause fieldslack the possibilityof new data, e.g., geophysics(although massshedding of mantle material, contributingto the disk. The reinspectionof the data continues).But the lunar samplesprovide concept of collisional impact is not new, as it derives from a continuing resourcefrom which some of the Moon's secrets Hartmann and Davis [1975] and Cameron and Ward [1976]. can, with carefulwork, be revealed,and terrestrialspectroscopic What really happened was a conflation of developments:(1) observationsprovide new data. Most important,perhaps, is that a realizationthat the traditional hypothesesof capture,fission, the framework for interpreting data is continually being and coaccretioncould not explain the M oon'sfeatures or were improved, so that our level of sophisticationis continually in some way implausible;indeed, collisional ejection appears increased. Lessons from terrestrial and meteorite studies are used to be the only mechanismwhich is not ruled out by dynamical to elucidate our lunar data, and in return, the Moon remains constraints[Boss and Peale, 1986]; (2) developmentof a better a major proving ground for ideas about the evolved bodies understandingof the growth of planetsfrom planetesimalsand of the solar system.In this review, prominenceis given to two the dynamicsof the early solar system,which showsthat large major developments which have taken place in the last bodyimpacts would be bothpossible and expected [ Wetherill, quadrennium:interest in and a possibleconsensus on the origin 1986;Hartmann and Vail, 1986],rather than ad hoc aspreviously of the Moon; and the recognitionand study of meteoritesfrom supposed;(3) the stimulationof the conferencefocussed attention the Moon. Less spaceis devotedto more "traditional" topics, on the hypothesis[Hartmann et al., 1986]. While the hypothesis not becausethey are actually lessimportant to lunar science, is now popular, it has not been proven, and it is not clear but because the advances are less dramatic. that it satisfies,or can satisfy,all observationaldata. Important calculations,for instanceon emplacementof ejectainto Earth THE ORIGIN OF THE MOON orbit rather than its reaccretion or total loss, and on three- The Apollo mission samplesdemonstrated that the Moon dimensionalsimulations of impact, are in progress[review in is a differentiated,evolved body, and not the primitive, cold Stevenson,1986]. The improvementin computingfacilities over body whichUrey had supposed.No clearreading of lunar origin the last decade have been very important in making the came from early Apollo science,which did however provide calculationsnecessary to amplify and constrainthe hypothesis. clues and constraints by better characterizing the Moon, Wetherill [1986] followed the natural orbital and collisional including demonstratinga very hot early period, an age the evolution of 500 initial planetesimalsto planetary formation. same as the Earth, and oxygen isotopeslying on the Earth's Later accumulationin all simulationsis characterizedby giant massfractionation line but differingfrom mostmeteorites. There impacts (up to 3X Mars size) which are sufficientto explain followed a period during which little attentionwas paid to the the large angular momentum of the Earth-Moon system(which origin of the Moon. However, during the last quadrennium, is not explainedby coatcreation).Such impacts can alsoexplain the origin of the Moon became a very active field of study, the Earth's obliquity [Hartmann and Vail, 1986]. One problem and the hypothesisthat the Moon was producedfollowing the with the collisionalejection hypothesis is showinghow material impact of a large (Mars-sized?)body into the Earth emerged can be emplacedinto orbit (rather than reaccretedor lost). One as a front-runner. This interest stemsfrom work inspired by way might be for the material to be vapor, not solid [Cameron, the Conferenceon the Origin of the Moon held in Kona, Hawaii, 1985, 1986; Thompsonand Stevenson,1983; Stevenson,1986], in October, 1984, itself inspired by the Lunar and Planetary so that gas pressureeffects become important. Viscousstresses Sample Team, and cosponsoredby the Lunar and Planetary or gravitational torques can redistribute some angular Institute, NASA, and the Division of Planetary Sciencesof the momentum [Stevenson, 1986] to help achieve orbit. Such an American AstronomicalSociety. The conferenceproduced an expanding vapor model can place enough material into abstractvolume [Hartmann et al., 1984] and a book of pertinent geocentric orbit to make the Moon [Cameron, 1985, 1986; papers[Hartmann et al., 1986]. Three important review papers Stevenson,1986]. Thompsonand Stevenson[1983] arguedthat the disk would remain hot and largely vaporizedthrough most Copyright 1987 by the •merican Geophysœcal Union. of its existencebut nonethelessexpand outsidethe Roche limit and cool within 100 years, and coalesce.The short time-scale Paper number 7R0059. 8755-1209/87/007R-0059515. O0 of accretionand the relativelyhigh temperatureof the solids 277 RYDER:THE MOON would producea partly or wholly molten Moon. Boss[1986b] new information about the characterand evolutionof the Moon, found that although dynamic fission was unlikely, spin up and prove that rocks can be ejectedfrom planetarybodies up produced by tangential impact could have contributed to at least lunar size without melting or even appreciable substantialmatter to a prelunardisk. Durisenand Scott [ 1984] shocking.Much of the work on ALHA 81005 was reported had removedan objectionto fission,but Bossand Mizuni [ 1985] at the 14th Lunar and Planetary ScienceConference [Abstracts found that planetaryviscosities, even of molten bodies,are too for Special Sessionon Meteorites from Earth's Moon, Lunar high for natural fissionto occur. and PlanetaryScience XIV] and in a specialissue of Geophysical The postulatedimpact is beyond any human experience,and ResearchLetters [Vol. 10, 9, 1983]; much of the work on the attempts to "scale" known impacts are probably doomed other meteoriteswas presented in two SpecialSessions on Lunar [Stevenson,1986]. Numerical three-dimensional simulations are Meteorites at the Tenth and Eleventh Symposiaon Antarctic in their infancy [Cameron, 1985; Benz et al., 1986a,b; Kipp Meteoritesin Japan (Special Session:Lunar Meteorites, 1985, and Melosh, 1986],and it would be prematureto usethe results 1986). as firm constraints. ALHA 81005 was the last of 373 samplescollected in the The Moon's compositionis at least roughly similar to that 1981- 1982season, in January, 1982.It wasrecognized as unique of the Earth'smantle [review by Drake, 1986a]and this evidence by its collector,John Schutt[Marvin, 1983].Its unusualnature wasa powerfulmotivator of the fissionhypothesis. Some authors wasdescribed in initial processing[Score, 1982], and its similarity have suggestedthat the compositionsare very similar indeed, to somelunar rockswas observed in thin section[Mason, 1982]. e.g., Ringwood [1986a,b] believesthat the siderophilepatterns Y-791197 was collected in November, 1979, but was not in the Earth are controlled by complex processesrelated to recognizedas lunar until much later [ Yanaiand Kojima, 1984]. core formation and uniqueto the Earth; he alsofinds the lunar Y-82192 and Y-82193, collectedclose together, were recognized siderophile abundance patterns to be similar to the Earth's as probablelunar samplesduring their preliminaryexamination mantle. He thus postulatesa number of large impactsinto the and are probablyfragments of the samefall [ Yanaiet al., 1986]. Earth producinga ring of mantle-derivedplanetesimals. Wiinke The evidence that these meteorites are lunar is manifold. The and Dreibus [1986] find the chemical similarities to be samples are all anorthositic regolith breccias, with the compelling.Others stronglydisagree with supposedsimilarities petrographic and chemical characteristicsof lunar rocks, [e.g., Drake, 1983, 1986a,b; Taylor, 1986; Kreutzbergeret al., including Fe/Mn about 70, from the highlands[e.g., Warren 1986; and othersin Hartmann et al., 1986],finding differences et al., 1983c; Warrenand Kallemeyn, 1986; Ostertaget al., 1985; in siderophiles,refractory elements, and Mg/Fe, among others. Lindstrom et al., 1985]. Perhapsmost decisiveare the oxygen Warren [1986] however