How Do We Know That It’s a Rock from the Moon?

Many people have approached us over the years wanting to know if a rock that they possess is a Moon rock. The most common story we hear is that the rock was given to a relative in the 1970’s by an astronaut, a military person, or a NASA security guard. We have chemically tested several such rocks and none has been a Moon rock. Other people suspect that they have found a lunar meteorite. None of the many samples that we have been sent has been a lunar meteorite, except those from meteorite dealers, those persons who bought lunar meteorites from a dealer, or those from experienced meteorite prospectors who found them in the deserts of northern Africa or Oman.

Lunar meteorite QUE (Queen Alexandra Range Antarctica ) 94281 – An unattractive rock that could pass for a cinder or piece of slag. It weighed 23 grams, just less than an ounce. The cube is 1 cm on each side. Image credit: NASA photo S95-14590

No lunar meteorite has yet been found in North America, South America, or Europe. They undoubtedly exist, but the probability of finding a lunar meteorite in a temperate environment is incredibly low. Many experienced meteorite collectors have been looking and none have yet succeeded. Realistically, the probability that an amateur will find a lunar meteorite is so low that I cannot raise much enthusiasm to examine the thousands of rocks and photos that I have been asked to examine. If I wanted to find a lunar meteorite myself, I would not scour the Mojave Desert. I would look through rock collections at colleges and universities. It is not unreasonable that a lunar meteorite exists in an old drawer somewhere because a sharp-eyed geology student or professor found a funny-looking rock years ago in a place it did not belong. It would not surprise me to learn that some “expert” proclaimed that the rock was not a meteorite because it didn’t look like an ordinary chondrite, it didn’t attract a magnet, or it didn’t contain a high concentration of nickel. Both visually and compositionally, lunar meteorites “look” more like terrestrial (Earth) rocks than do “normal” meteorites (ordinary chondrites). It would be easy to overlook a lunar meteorite. A weathered lunar meteorite would look remarkably unremarkable.

Weathered pebbles of lunar meteorite Northwest Africa 11788, 1-cm cube on right. If I found a few of these in my driveway, I wouldn’t give them a second look. Photo credit: Rob Wesel

Here I discuss some aspects of lunar geology, mineralogy, and chemistry that guide us in our attempts to identify lunar material.

Lunar Mineralogy

Only four minerals – plagioclase feldspar, pyroxene, olivine, and ilmenite – account for 98-99% of the crystalline material of the lunar crust. (Material at the lunar surface contains a high proportion of non-crystalline material, but most of this material is glass that formed from melting of rocks containing the four major minerals.) The remaining 1-2% is largely potassium feldspar, oxide minerals such as chromite, pleonaste, and rutile, calcium phosphates, zircon, troilite, and iron metal. Many other minerals have been identified, but most are rare and occur only as very small grains interstitial to the four major minerals and cannot be seen with the naked eye.

Some of the most common minerals at the surface of the Earth are rare or have never been found in lunar samples. These include quartz, calcite, magnetite, hematite, micas, amphiboles, and most sulfide minerals. Many terrestrial minerals contain water as part of their crystal structure. Micas and amphiboles are common examples. Hydrous (water containing) minerals have not been found on the Moon. The simplicity of lunar mineralogy often makes it very easy for me to say with great confidence “This is not a moon rock.” A rock that contains quartz, calcite, or mica as a primary mineral is not from the Moon. Some lunar meteorites do, in fact, contain calcite. However, the calcite was formed on Earth from exposure of the meteorite to air and water after it landed. The calcite occurs as a secondary mineral, one that fills cracks and voids (see Dhofar 025). Secondary minerals are easy to recognize when the meteorite is studied with a microscope.

pyroxene – A group of magnesium-iron-calcium silicates, common on the Earth and Moon.

clinopyroxene – A form of pyroxene; typically contains some calcium; most common in mare basalts [Ca(Mg,Fe)Si2O6].

orthopyroxene – A form of pyroxene; contains little calcium; most common in highlands rocks [(Mg,Fe)SiO3].

olivine – A magnesium-iron(II) silicate; common on the Earth and Moon [(Mg,Fe)2SiO4].

ilmenite – An iron(II)-titanium oxide; more common in lunar basalts than in terrestrial basalts [FeTiO3].

feldspar – A group of alumino-silicate minerals; common in the crusts of the Earth and Moon.

plagioclase – A form of feldspar; a calcium-sodium alumino-silicate [(CaAl,NaSi)AlSi2O8].

anorthite – A mineral; the calcium-rich extreme of the plagioclase feldspar; the most common mineral of the lunar crust, but not so common on Earth.

anorthosite – A rock consisting mainly of anorthite.

Lunar Rocks – Breccias

Pieces of Apollo 16 anorthosite sample 60025. This particular sample is nearly pure anorthite; there are no dark iron-bearing minerals. Photo credit: Randy Korotev

Most of the lunar crust, that part called the Feldspathic Highlands Terrane or simply the feldspathic highlands, consists of rocks that are rich in a particular variety of plagioclase feldspar known as anorthite. As a consequence, rocks of the lunar crust are said to be anorthositic because they are plagioclase-rich rocks with names like anorthosite, noritic anorthosite, or anorthositic troctolite (see table below). The ratio of iron-bearing minerals to plagioclase probably increases with depth in the feldspathic highlands at most places. For example, rocks exposed in the giant South Pole – Aitken impact basin on the far side are richer in pyroxene than typical feldspathic highlands.

rock name mineralogy
anorthosite >90% plagioclase
noritic anorthosite and anorthositic norite 60-90% plagioclase, the rest mostly orthopyroxene
gabbroic anorthosite and anorthositic gabbro 60-90% plagioclase, the rest mostly clinopyroxene
troctolitic anorthosite and anorthositic troctolite 60-90% plagioclase, the rest mostly olivine
norite 10-60% plagioclase, the rest mostly orthopyroxene
gabbro 10-60% plagioclase, the rest mostly clinopyroxene
troctolite 10-60% plagioclase, the rest mostly olivine

In much of the northwest quadrant of the nearside of the Moon, in the region known as the Procellarum KREEP Terrane, the crust contains less plagioclase and more pyroxene. The original rocks of this anomalous crust were probably mostly norites and gabbros. The feldspathic crust of the Moon began to form about 4.5 billion years ago. While it was forming and for some time afterwards, it experienced intense bombardment from meteoroids and asteroids. The rocks of the lunar crust have been repeatedly broken apart by some impacts and glued back together by other impacts. As a consequence, most rocks from the lunar highlands are breccias (brech’-chee-uz), a word for a rock composed of fragments of older rocks. Breccias occur on Earth, but they are much less common than on the Moon. Also, most terrestrial breccias were not formed by meteoroid impacts but by faulting. Lunar breccias are subdivided into a variety of categories such as impact-melt, granulitic, glassy, fragmental, and regolith breccias. In impact-melt and glassy breccias, rock fragments called clasts are suspended in a solidified (crystalline or glassy) melt matrix formed by meteorite impact.

More information on lunar breccias and terrestrial look-alike rocks.

In fragmental and regolith breccias, there is little or no molten portion, just fragmental debris that was lithified (formed into a rock) by the shock pressure of an impact. Because breccia refers to texture and anorthositic or feldspathic refers to mineralogy, rocks from the lunar highlands are variously called anorthositic breccias, feldspathic breccias, or highlands breccias. Because the lunar crust has been battered so intensely, there were very few hand-sized rocks collected on the Apollo missions that are unbrecciated remnants of the early igneous crust of the Moon. Thus it is no surprise that all of the lunar meteorites from the Feldspathic Highlands Terrane and the Procellarum KREEP Terrane are breccias.

Lunar Rocks – Mare Basalts

Left: Mount Erebus in Antarctica, the Earth’s most southern volcano. Image credit: Randy Korotev. Right: Basalt-filled impact basins on the Moon. Image credit: NASA/GSFC/Arizona State University.

On Earth, volcanoes are often cone-shaped mountains because they are a pile of ash and lava ejected from a vent. The lavas are viscous and solidify before they flow very far. Because of their iron-rich composition and lack of water, lunar lavas were much less viscous, more like motor oil. When lunar lavas erupted onto the surface they did not form volcanoes, they simply flowed and filled low spots. As a result, lunar lava deposits are flat, thin, and cover wide areas. Also, because the Moon has no atmosphere and little gravity, ejected ash dissipated widely instead of piling up near the vent, as on Earth.

Starting about the time of the period of intense meteorite bombardment, the lunar mantle partially melted. The resulting magmas rose through the crust to the surface, ponding in low spots. These low spots were mainly the huge craters, called basins, that were left by impacts of the largest meteorites. Lunar volcanism continued for about 2 billion years.

Pronunciation: The Latin word mare is pronounced mar’-ay in English. The plural of mare is maria, which is pronounced mar’-ee-ah. Basalt is usually pronounced bah-salt’.

On Earth, volcanic rocks solidify from molten lava (magma). The most common type of volcanic rock is basalt. The ancient astronomers called the round, basalt-filled basins on the surface of the Moon seas because they were smooth dark areas surrounded by areas of higher elevation. The features were given Latin names like Mare Serenitatis for Sea of Serenity. We now know that the lunar maria are basalt flows, so we call the rocks of the maria mare basalts. Mare basalts are composed mainly, 50-70%, of pyroxene, but they all also contain 20-40% plagioclase, up to 20% ilmenite and related Ti-rich minerals, and 0-20% olivine. The maria are darker than the highlands because (1) mare basalts are rich in iron-bearing minerals, (2) iron-bearing minerals are dark colored, and (3) plagioclase is light colored. In contrast to the highlands, most of the rocks collected on maria by the Apollo astronauts are actual basalts, not breccias composed of fragments of basalt. This observation is one of several reasons why we know that the basalts mostly formed after the time of intense bombardment. Mare basalts cover about 17% of the surface of the Moon, but it is estimated that they account for only about 1% of the volume of the crust.

Apollo 11 basalt sample 10044 (left) and Apollo 15 basalt sample 15016. The Apollo 15 mare basalt is vesicular – it has holes that were once gas bubbles. Most Apollo basalts are not vesicular and, so far, none of the basaltic lunar meteorites are vesicular. On the left, the cube is 1 cm and on the right the cube is 1 inch on each side. Image credits: NASA

Because lunar meteorites are samples from randomly distributed locations on the surface of the Moon and because most of the lunar surface is feldspathic, most of the lunar meteorites are feldspathic breccias. Some are crystalline mare basalts, breccias composed of mare basalt, or breccias composed of both mare and highlands material (like QUE 94281, above). A few are dominated by noritic material of the Procellarum KREEP Terrane.

Lunar mare basalts, as well as basaltic meteorites from Mars, bear a strong resemblance to basalts from Earth. In the absence of a fusion crust, there is little about a lunar mare basalt that would provoke much interest in a geologist handed the rock by someone asking “what is this?” Careful examination under the microscope might reveal some suspicious features – the lack of certain minerals and abundance of others (ilmenite) or the low sodium content of the feldspar. The mineral grains would show signs of shock and fracturing from meteorite impacts. However, chemical tests would be required to prove a lunar or martian origin.

Types of rocks on the Moon: basalt, anorthosite, breccia, and  “soil” (regolith).

Fragmental and regolith breccias are the closest lunar analogs to terrestrial sedimentary rocks, and they bear some textural resemblance. However, there are numerous differences, nearly all associated with the lack of water and wind on the Moon. As noted above, lunar rocks do not contain carbonate minerals or abundant quartz, as do most terrestrial sedimentary rocks. There is no effective sorting mechanism on the Moon, so the lithic components of lunar breccias come in a wide variety of grain sizes, with no preferred size or orientation. Lunar breccias are largely fractal objects that look similar in cross section regardless of the scale at which they are viewed. (See ALHA 81005.) There is no known lunar rock that has any feature that resembles the layers that are characteristic of terrestrial sedimentary rocks. Terrestrial sedimentary rocks have layers because the Earth has gravity, so particles settle in water or in the atmosphere. The Moon has only weak gravity and no water or atmosphere.

If a rock is layered, then it is not from the Moon

Most small clasts in lunar breccias are fragments of plagioclase or anorthosite. It is rare for the aspect (length to width) ratio of a clast in a lunar breccia to exceed 3. Most clasts are angular, not rounded. (Exceptions: There are volcanic glass spherules in the lunar regolith (soil). Such spherules are sometimes found in regolith breccias, but they are <0.1 mm in diameter and not easily seen with the unaided eye. Impact-produced spherules occur and may be large, but they are not common compared to rock and mineral fragments. Impact-melt breccias may contain clasts that have been partially melted and which are consequently not angular.)

Brecciated lunar meteorites are sufficiently tough and cohesive that they survived the blast off the Moon and the hard landing on Earth. Many terrestrial sedimentary rocks break apart much easier. Unlike some terrestrial conglomerates, which resemble lunar breccias, the matrix of lunar breccias is as hard as the clasts. On broken or exterior surfaces of brecciated lunar meteorites, the clasts do not stand out in either negative or positive relief.


Sawn faces of Apollo 16 sample 60019 on the left and lunar meteorite MAC (MacAlpine Hills) 88105 on the right. Both are coherent regolith breccias. There are several similarities. Clasts are lighter in color than the matrix. Clast sizes are about the same (1 cm cube) and vary in size. Fractures do not deviate around the clasts – they pass through them as if they were not there. Unlike as in some terrestrial sedimentary rock, the clasts are not “sorted” and there is no preferred orientation of the clasts.

Metal and Magnetism

Meteorite collectors know that most meteorites attract a cheap magnet because they contain iron-nickel metal. The most common type of meteorites, the ordinary chondrites, do indeed contain metal as, of course, do iron meteorites. Lunar mare basalts and the original rocks of the lunar highlands contain essentially no iron metal (much, much less than 1%). Brecciated lunar meteorites, however, contain some metal from the asteroidal meteorites that have bombarded the Moon. Among lunar meteorites, Dhofar 1527 contains the most metal, about 1.7%; most contain much less. In other words, lunar meteorites do not attract magnets, as do most other kinds of meteorites.


Because of the simplicity of lunar mineralogy, lunar rocks have predictable chemical compositions. Nearly all the aluminum is in plagioclase and nearly all the iron and magnesium are in pyroxene, olivine, and ilmenite. Thus, on the plot of aluminum concentration (Al2O3 in figure below) versus the concentrations of iron (FeO) plus magnesium (MgO), lunar meteorites (and nearly all Apollo lunar rocks) plot along a line connecting the composition of plagioclase and the average composition of the three iron-bearing minerals because these are the only four major minerals in the rock. If the composition of a rock does not plot along this line, the rock is almost certainly not a lunar rock.

Many people contact me about rocks that they hope or think to be meteorites. If they are adamant, I suggest that they get a chemical analysis of the rock. All of the “terrestrial rocks” of this plot represent rocks that people have had analyzed. Most lie below the lunar-meteorite line because they contain quartz or calcite, which plot in the lower left corner of the plot. A few of the terrestrial rocks plot on or near the lunar trend. All of these are igneous rocks like basalts that are dominated by pyroxene, olivine, and plagioclase. The same process forms basalts on Earth, Moon, and Mars, so they all have the same basic mineralogy and composition. Other tests are required to distinguish Earth basalts from planetary basalts. The three green points represent the 3  types of ordinary chondrites (H, L, and LL), which consist mainly of olivine, pyroxene, and iron-nickel metal. As a consequence of the metal, they plot at high FeO(+MgO). (For geochemists: “FeO” is total Fe as FeO.)

On Earth, the silica (SiO2) concentration of igneous rocks is used as a first-order chemical classification parameter because it varies widely among different kinds of rocks. On the Moon (1) there are no rocks rich in quartz or other silica polymorphs*, (2) in a given rock, particularly breccias, the average concentration of silica in the three main minerals, plagioclase, pyroxene, and olivine, are all about the same, and (3) in highlands rocks ilmenite is usually present only in small amounts (<3%), so silica concentrations of common lunar rocks vary by only a small amount. In lunar meteorites, SiO2 concentrations span the narrow range from 43% to 47%. Because aluminum varies by more than a factor of 3, however, aluminum is more useful as a chemical classification parameter. (Titanium is used in mare basalts.) Similarly, among nearly all common lunar rocks calcium concentrations vary by only a factor of 2, from 10% to 20% as calcium oxide (CaO). This is much less than the range in terrestrial rocks. A rock with silica or calcium oxide concentrations substantially outside these ranges is almost certainly not a lunar rock.

* Some lunar mare basalts contain up to 5% cristobalite, a silica mineral. There are some rare and small lunar samples with 50-70% SiO2 because they contain tridymite, quartz, or silica glass. These include felsites, granites, and related silica-rich rocks like quartz monzodiorite. There are also rocks that contain <10% CaO because they contain little plagioclase. These include some ultramafic rocks like dunite and some picritic volcanic glasses.

In Earth rocks, iron occurs in both the 2+ and 3+ oxidation states. On the Moon, iron occurs in the 0 (metal) and 2+ oxidation states, although in lunar igneous rocks almost all of the iron is in the 2+ oxidation state (in olivine, pyroxene, and ilmenite). On the Moon all manganese is also in the 2+ oxidation state. Because Fe(II) and Mn(II) have very similar chemical behaviors, iron does not fractionate from manganese during lunar geochemical processes, as it does on Earth. As a result, the ratio of iron to manganese in lunar rocks is nearly constant at 70, regardless of whether the rocks are from the maria (high Fe and Mn) or from the highlands (low Fe and Mn). Nonlunar meteorites have different FeO/MnO ratios than Moon rocks. Earth rocks have a huge range of FeO/MnO ratios, but for average terrestrial crust the ratio is a bit lower than on the Moon.

Whole-rock FeO/MnO ratios in lunar meteorites and comparison to Earth rocks, martian meteorites and HED meteorites (howardite, eucrite, diogenite). FeO/MnO alone can distinguish lunar meteorites from other achondrites but not always from terrestrial rocks.

The element chromium is in greater concentration in lunar rocks than most Earth rocks (bottom plot here). Chromium concentrations in mare basalts range from 0.14% to 0.44% (as Cr). Even the feldspathic lunar meteorites, with 0.05-0.09% Cr, are considerably richer in chromium than is the average terrestrial crust (~0.01%).

These two figures are similar to the figures above, but here the five unfilled green triangles represent five samples of alleged lunar rock discussed in the first paragraph above (none are lunar). Left: Lunar meteorites from the highlands (unfilled blue squares) have a constant ratio of thorium to samarium (represented by the diagonal blue line). Lunar meteorites from the maria (filled blue squares) tend to have lower, but similar, ratios. Some terrestrial samples have similar ratios of incompatible elements as the lunar highlands ratio but some do not. Right : All lunar samples have very low concentrations of arsenic compared to terrestrial rocks and meteorites. Except for rare felsites, all lunar rocks also have low concentrations of potassium compared to terrestrial rocks.

Concentrations of the alkali elements (potassium, sodium, rubidium, and cesium) are 10 to 100 times lower in lunar rocks than in terrestrial rocks. Terrestrial sedimentary rocks often contain sulfide minerals like pyrite. Sulfide minerals are rare in lunar rocks and elements such as copper, zinc, arsenic, selenium, silver, mercury, and lead, which are often found in sulfide minerals, occur in very low abundances in lunar rocks. Low concentrations of alkali elements and sulfide-loving (chalcophile) elements are one of the most characteristic features of lunar rocks.

Odd Rocks

As noted above, there are known exceptions to the generalizations, and we lunatics certainly hope that we haven’t discovered all the minerals and rock types that occur on the Moon. However, known samples of unusual composition and mineralogy are rare and usually occur only as small (<1 gram) clasts in breccias or in the soil. We have no reason to suspect, based on data obtained from orbit on the Clementine and Lunar Prospector missions, that any region of the Moon is rich in types of rocks significantly different from those we know about or postulate might exist. Most ore-forming processes on Earth involve water, so we would not expect any hidden ore deposits on the Moon. Keep in mind that if more than 400 lunar meteorites have been blasted off the Moon and found on Earth, then at any given point on the lunar surface there can be rocks from any other point. For this reason, the fact that the lunar surface was “poorly sampled” by the Apollo and Luna missions is in itself not a good reason to suspect that rocks vastly different from those that we have studied exist at unsampled points on the Moon. Tens of thousands of lunar rocks and rocklets have been studied since the Apollo missions. It is highly unlikely that any yet-unfound lunar meteorite will differ substantially in the minerals it contains or in its geochemical character from the Apollo lunar rocks and lunar meteorites.

They Were Faked

Any geoscientist (and there have been thousands from all over the world) who has studied lunar samples knows that anyone who thinks the Apollo lunar samples were created on Earth as part of government conspiracy does not know much about rocks. The Apollo samples are just too good. They tell a self-consistent story with a complexly interwoven plot that is better than any story any conspirator could have conceived. I have studied lunar rocks and soils for 50+ years and I could not “make” even a poor imitation of a lunar breccia, lunar soil, or a mare basalt in the lab. And with all due respect to my clever colleagues in government labs, no one in “the Government” could do it either, even now that we know what lunar rocks are like. Lunar samples show evidence of formation in an extremely dry environment with essentially no free oxygen and little gravity. Some have impact craters on the surface and many display evidence for a suite of unanticipated and complicated effects associated with large and small meteorite impacts. Lunar rocks and soil contain gases (hydrogen, helium, nitrogen, neon, argon, krypton, and xenon) derived from the solar wind with isotope ratios different than Earth samples of the same gases. They contain crystal damage from cosmic rays. Lunar igneous rocks have crystallization ages, determined by techniques involving radioisotopes, that are older than any known Earth rocks. (Anyone who figures out how to fake that is worthy of a Nobel Prize.) It was easier and cheaper to go to the Moon and bring back some rocks than it would have been to create all these fascinating features on Earth.