Apollo 11 had proved the ability of the LM to land on the Moon, but the fact that it came down off target was frustrating. The ability to land within about 1,000 feet of a specific point was a prerequisite to being able to undertake a planned geological traverse. After the flight dynamics team had devised a simple method to correct for the perturbations of the mascons, they were so confident that they reduced the size of the target ellipse. In addition, it was decided to cut the number of backup sites from two to one. There were five prime sites on the short-list for the first landing. The easterly ALS-1 and ALS-2 sites in the Sea of Tranquility had been backed up by ALS-3 in the Meridian Bay, with ALS-4 and ALS-5 in the Ocean of Storms in reserve against a major launch delay. It would have been natural to send Apollo 12 to one of these sites, but the conservative constraints had resulted in the choice of ‘open’ sites, and the geologists were eager to sample the ejecta of a sizeable crater. In fact, even before Apollo 11, the site selectors had re-examined sites rejected due to the inconvenient proximity of a crater, and listed them for a later mission. In the end, however, in order to convincingly demonstrate the ability to address a ‘pin-point’ target it was decided to land alongside an unmanned probe. The relaxation of the operational constraints allowed the reinstatement of the Surveyor 1 site (ALS-6) in the Ocean of Storms. However, because this was so far west that it did not permit a backup, it was decided instead to visit Surveyor 3 in the eastern Ocean of Storms. Originally designated 3P-9, this site became ALS-7. Pete Conrad and Al Bean landed their LM, ‘Intrepid’, within 600 feet of Surveyor 3 on 19 November 1969. On their first excursion they deployed the deferred ALSEP, and during a 3-hour traverse the next day they ranged 1,200 feet from home, collected samples, and cut parts off the Surveyor as trophies. Meanwhile, in orbit, Dick Gordon photographed the site being considered for the next mission.

A pre-mission investigation of the morphology of the craters at the Apollo 12 site predicted that the regolith would not exceed 6 feet in thickness, and that the large impacts would have excavated bedrock. Whereas breccias and basalts were represented equally by number in the Apollo 11 samples, just two of the 34 rocks returned by Apollo 12 were breccias. The crystalline rocks were coarser and more texturally diverse. In view of the fact that they contained less titanium, it appeared, on reflection, that the basalt of the Sea of Tranquility was unusually enriched in this element. This chemical variation confirmed that the dark plains were not from a single source. Indeed, the fact that four kinds of basalt were identified at the Apollo 12 site meant that there had been several distinct flows in this local area.[52] However, the crystallisation dates clustered within a fairly narrow window, which suggested that the extrusions were the result of partial melting of pockets of rock at shallow depth. The initial results were confusing, but it was immediately recognised that something profound had been discovered concerning early lunar history. The first measurement yielded an age of 2.7 (±0.2) billion years, which meant a billion years had elapsed between the extrusions in the Sea of Tranquility and the Ocean of Storms. The next result pushed this up to 3.4 billion years, but as the analyses continued the dates converged on 3.2 billion years. This 500-million-year span in ages for the lavas at the two landing sites indicated that the driving process had been persistent. Geochemist Paul W. Gast made a surprising discovery in the basalts, in the form of an abundance of potassium, phosphorus and some of the ‘rare earth’ elements. By linking their chemical symbols, Gast coined the label ‘KREEP’. On trying to isolate this material, he realised that it was not present as a mineral. The term is an adjective, and it is more correct to describe the Ocean of Storms basalts as being KREEPy. By way of an ‘instant science’ explanation for the media, Gast suggested this chemical additive might have been picked up from the ancient crust that some scientists believed formed the ‘basement’ of the dark plains, and he even speculated that it might be associated with the putative light-toned basalt believed (by some) to be prevalent in the highlands, but when the material proved to be rich in radioactive elements, in particular thorium and uranium, it was realised that this could not be typical of the crust because the heat of radioactive decay would have prevented the crust from solidifying. This KREEPy additive became a mystery for a subsequent mission to resolve.

After being discarded, the ascent stage of the LM was deliberately de-orbited, and the ALSEP seismometer recorded the crust ‘ringing’ for nearly an hour with a signature quite unlike a terrestrial signal. At the Lunar Science Conference, Gary V. Latham, the principal investigator for the seismic instruments, noted it had been difficult to tell the difference between a moonquake and an impact until this strike had provided a point of reference, whereupon it was found that surprisingly few of the 150 seismic events on record were internal quakes. It seemed that the crust was brecciated to a depth of about 18 nautical miles, indicating that, after the crust had solidified, further impacts had churned this up to a considerable depth, forming a

‘megaregolith’. In order to probe to greater depths, it was decided that on future missions the spent S-IVB should be made to impact the Moon.