Apollo Lunar Surface Journal Banner

Fairy Castles

Last revised 3 January 2006.

The following commentary on fairy castles was secured for inclusion in the Journal by Ron Wells, who discussed the problem with Bruce Hapke, one of the originators of the theory. Readers should note that I do not agree with the impression Hapke gives in the third paragraph that the returned Apollo cores confirm a porosity of about 70% to depths of several meters. The discussion of the surface materials by Carrier and others in the Lunar Sourcebook indicates that, while porosities of about 50% are typical in the top 15 cm and can be even higher right at the surface, below about 15 cm porosities are more typically about 43%. My sense of the subject is that fairy castles are strictly a phenomenon of the top few centimeters. Despite this technical disagreement, I am pleased to include Hapke's account because of the insights it gives into the contentious issue of the character of the lunar surface that bedeviled Apollo planners.

Commentary by Bruce Hapke - The History of the Fairy Castles

The complaint by Dave Scott (at 125:00:52 was based on an understandable, but unfortunate, confusion about different pre-Apollo models of the lunar surface. The fairy castle model of the lunar soil suggested by Hugh Van Horn and me, and the deep dust hypothesis of Thomas Gold. Our fairy castle model of the lunar regolith was published in Hapke and Van Horn (1963,J. Geophys. Res., 68, 4545). It was put forth as an explanation for the unusual, strongly backscattering nature of the way the lunar surface reflects light as a function of angle. It relies on the observation that there are adhesive forces between two particles caused by unsatisfied molecular surface bonds, and that if the particles are small enough these forces are larger than the weight of the grain. Hence, a small particle can be stably supported by one contact, rather than the three points required for a large particle. This makes any fine powder cohesive and have a high porosity, typically ~70-90%. If examined with a stereoscopic microscope so that the three-dimensional structure can be seen, the surface of a fine powder is seen to consist of towers leaning at crazy angles and connected by lacey bridges and flying buttresses. Hugh and I called these structures "Fairy Castles" to evoke an image of a mysterious landscape. This turns out to be just the kind of surface microrelief needed to explain the peculiar lunar photometric, polarization and thermal properties. I still think the name is apt.

At the time we published the paper, we considered it to be a significant breakthrough because it showed that the Moon had a fine powdery surface. Of course, after the Apollo landings everyone said that they knew all along that the lunar surface was covered with a fine-grained regolith (e.g., Wilhelms, To a Rocky Moon). However, actual accounts from that period tell a rather different story. Gene Shoemaker was quoted in an article in National Geographic (circa 1963) as saying that the surface was covered with cobbles (fist-sized chunks of rock). I still have a copy of the cover of the Houston telephone directory (circa 1964), which was a NASA publicity photo showing a spacesuited astronaut walking on the NASA geologists' best guess of the Lunar surface: volcanic ash consisting of centimeter-sized rocks. Following the Soviet Luna 9 landings, Gerard Kuiper, the preeminent planetary astronomer of his time, held a news conference in which he proclaimed that the surface was obviously volcanic 'aa' (rough scoriaceous) lava, adding that this would "tear an astronaut's boots to shreds". Even after the unmanned Surveyor landings on the moon, the NASA geologists continued to insist that the regolith was course-grained.

By contrast, in a paper from that time (Hapke, 1968, Planet. Spa Sci., 16, 101) I used the Fairy Castle model to predict that the Lunar regolith would have a mean grain size (weighted by mass) of between 10 and 100 micrometers. The actual mean sizes of the Apollo soil samples are all around 40-50 micrometers, although this value probably is too high because many grains are actually friable, weakly consolidated agglutinates of smaller particles counted as large particles. As I pointed out in that paper, by any definition of dust one cares to use, the lunar regolith may be accurately described as made of fine dust. Aldrin reported, "Kicking up dust", as the Eagle hovered above the surface. The dust covered everything including camera optics, almost making one of the Surveyor cameras useless. Astronauts coming in from Lunar EVA were described as looking like coal miners. Lunar spacesuits in museums are still badly soiled by fine dust in the fibers, in spite of repeated cleanings. Pre-Apollo radar data showed that the lunar surface has a porosity ~70% to a depth of at least several meters. By Occam' s razor, we argued that the regolith probably consists of Fairy Castles (i.e., fine-grained, cohesive powder) all the way down, and the Apollo cores confirmed this. Dave Scott was having so much trouble with the drill precisely because the soil is highly cohesive, as required by the Fairy Castle model. He would have had much less resistance if the regolith had consisted of the incohesive, course-grained sand the NASA geologists had predicted.

In the preface to my book, Theory of Reflectance and Emittance Spectroscopy, I give an anecdotal account of the discovery of Fairy Castles.

Now let me turn to Gold' s deep dust model. If one actually reads what Gold said in his paper (Gold, 1955, Mon. Not. Roy. Astr. Soc., 115, 585), rather than what some NASA geologists say he said, one finds that it is a carefully reasoned, logical model that was fully consistent with the observations known at that time. Gold argued that (1) the Moon had always been a cold body, (2) the craters were of impact origin, (3) the maria were deposits of material eroded from the surrounding highlands and carried into depressions by an electrostatic transport process, and (4) there was a darkening process operating on the lunar surface that lowered the albedo of material exposed at the surface This model was considered to be a less likely but, nevertheless, credible possibility by most persons outside the NASA Astrogeology group, including Harold Urey, a Nobel Prize winning scientist and the father of modern planetary and meteorite chemistry.

The notion that the maria deposits were so unconsolidated that a spacecraft would sink out of sight in them was not a part of the original model, nor was it essential to it. Unfortunately, Gold had a penchant for the dramatic, and over-emphasized this possibility prior to the Apollo missions. This is the part that was not taken seriously by most persons. Even Gold considered it unlikely, as he admitted to me in private conversations. However, since he was a member of President Kennedy's President's Science Advisory Council, NASA was forced to take his suggestions seriously, much to the chagrin of some of the Astrogeologists.

However, the deep dust model remained viable up until the Apollo landings because there was little direct evidence for volcanism in the pre-Apollo data. On several occasions NASA Astrogeologists identified features in telescope or remote spacecraft images as volcanic and as "proving that the moon has had a long and complicated history of volcanism", only to have them turn out to be something else. The best example of this kind of gaffe was the Apollo 16 landing site, which was selected by the NASA Astrogeologists because it was clearly an ancient lava flow and possibly part of the original crust of the moon. However, when the astronauts got there they found that they were walking around on part of the ejecta blanket from Mare Imbrium. It was only after petrologic examination of the Apollo 11 rocks that it became clear that the maria were volcanic flood basalts. Unfortunately, Gold was never able to accept this result.

It is relevant to point out that we are having a similar debate today about the smooth plains on Mercury. Ironically, it is now Don Wilhelms, one of the NASA Astrogeologists, who has cautioned against prematurely interpreting them as lava flows.

Why was there this personal animosity on the part of some NASA Astrogeologists towards Gold's model and the confusion of the Fairy Castle model with it? There appear to be two main reasons. First, one must realize that there were two groups of scientists investigating the moon in the decade before Apollo. One group consisted of persons with a background in astrophysics, who studied the moon by quantitative remote sensing methods. Their optical, thermal and radar data were all consistent with Gold's model in its less extreme form. Most members of this group probably never had a course in geology. By contrast, the other group was the NASA Astrogeology Branch put together by Shoemaker. These persons were trained in classical geology and most of them probably never had a course in physics. The two groups did not speak the same language and did not understand each other's reasoning. Consequently, the Astrogeologists tended to ignore the astrophysicists' arguments.

Second, Gold came from the British School of debating, in which opponents take great delight in sarcastically skewering each other with rapier wit. While this makes for great entertainment at scientific meetings, it does not enhance communication among persons holding opposing points of view who are not used to this debating style. The NASA Astrogeologists were often the butt of Gold's sarcasm, and I think it got to the point that they either ignored or deprecated anything he said. This attitude apparently rubbed off on to anyone associated with him. Since I worked in the Cornell University Center for Radio Physics and Space Research at the time Gold was the director, my work apparently tended to be mixed up with his and was also ignored by them.

Finally, it should be emphasized that many aspects of Gold's model are correct, after all. The craters are of impact origin. While the maria did turn out to be lava flows, the flow surfaces are buried under several meters of dust. There are, indeed, erosional and depositional processes operating on the lunar surface, including electrostatic levitation, although these processes are not as efficient as Gold hypothesized. And there is ample evidence for his predicted darkening process, which today is called "space weathering". I think it is time for the devil to be given his due.]