This interesting article on moonbase site selection article by Peter Kokh appeared in the publication SpaceViews. As it is a useful document, with his permission, it is here in the "Site Selection" section of the Artemis Data Book.
There are those so impatient to return to the moon that they disvalue any further robotic missions designed to reveal where the richest and most accessible resources lie as "money sink distractions." Yet, to reduce the chances of the first human outpost becoming a ghost town in unseemly short order, the careful selection of a site especially capable of supporting viable economic activity is hardly unimportant. Rather it is impatience that needs to be dismissed. Impatience always backfires. That's a cosmic law. There is no point in deliberately blindfolding ourselves and playing "pin the tail on the donkey" with a moon map as some apparently want to do.
At the same time, it is possible to argue that any good site will do to demonstrate the viability of a permanent human presence on the moon. The task of such a beachhead is to survive the days and nights, the heat and the cold, the radiation and solar storms and micrometeorite rain, the absence of a biosphere rich in organics and volatiles. Next the aim is to begin demonstrating an ability to use the resources that are common on the moon to provide some continuing support and a respectable part of the wherewithal to expand.
The moon's major resources (oxygen, silicon, iron, aluminum, calcium, titanium, and magnesium) are distributed rather homogeneously (relative to their very uneven redistribution on Earth). So, the argument goes, we can always pick a second more advantageous site to begin industrial settlement in earnest. Indeed, one might argue, the lessons learned in the initial demonstrator outpost might warrant a fresh start elsewhere, rather than expand upon the trial and error dawn base.
While there is certainly merit to this argument, it is also likely that whether those planning and going on to deploy the first base care or not, additional robotic resource-finding missions are likely to be flown before the first outpost can be erected. In that case, it would be foolish not to take into consideration the knowledge those probes supply.
Some general considerations can be made now. Both from a resource using and a tourist/film-making point of view, it would be stupid to locate the outpost either in the middle of crater-pocked highland terrain, or in the middle of the much flatter maria terrain -- when by picking a "coastal" site the mineral and scenic diversity of both (highland and "sea") are present. Happily, innumerable sites fit this requirement.
If early industrial activity beyond oxygen extraction is likely to center on iron as the easiest element to extract and produce, we already have fair evidence of extensive areas that fit the bill. We'd be suicidally foolish to locate elsewhere.
Another point of convergence is maximizing public interest and awareness. This should be important both to those who would like to see a government moonbase (in the mold of Antarctica's McMurdo Sound) and those who would like to see a civilian commercial outpost (like most every for real burg on Earth). One sure way of doing this is to locate the base in an area that can easily be identified by the trained naked eye, or at least in binoculars. Perhaps others in the habit of studying the moon with the naked eye might not concur, but the feature I find easiest to locate at all phases of the Moon visible in early to middle night hours is the Sea of Crises, Mare Crisium, to the north east of center. This oval Mare, the size of Wisconsin and Upper Michigan together, is clearly distinct from the "chain of seas" that run into each other: Fertility, Tranquility, Serenity, Rains, the Ocean of Storms, etc. I am aware of no one else who is partial to Crisium. Other proposed locations in Fertility, Tranquility, Serenity, Imbrium, the crater Alphonsus, etc. can be picked out by the trained eye easily enough in binoculars, but that makes them unidentifiable for the masses. Anyone can learn to spot Crisium immediately. Somewhere along the shore of Mare Crisium, along the highlands separating it from Mare Tranquilitatis or Mare Fecunditatis could make a fair site. Of course, this is only one consideration and must be weighed along with others. Daytime naked eye identifiability is not the only PR trick that promises to build public awareness. A nighttime beacon near the outpost, beaming enough lumens Earthward to be clearly picked out, would certainly command much more attention. This would suggest placing the outpost in a part of the Moon that is usually not illuminated when the moon is above the horizon in early evening hours -- in other words, well into the western hemisphere (coastal/shore areas of western Oceanus Procellarum, the Ocean of Storms, or in the Aristarchus area for example). In contrast, a beacon in any of the eastern seas (Tranquility, Fertility, Crisium, etc.) would not be visible until the waning (post full) moon that rises later in the evening and would be noticed by far fewer people.
On Earth we distinguish between improved and unimproved sites. The latter lack electrical and water utility access. But even unimproved sites on Earth have atmosphere and access to at least some rain. No site on the moon has as much, every lunar site being radically unimproved.
Yet some sites have assets, beyond minerals, that other sites do not, such as appreciable part-time (and rarely, full time) shade. This can be important in planning thermal equilibrium maintenance with the placement of heat rejecting radiators, etc. Rille walls and crater walls and escarpments all provide part time shade depending on the local path of the Sun across the sky. In general such minimally improved sites are scattered everywhere, but are more densely located nearer to the poles where the maximum elevation of the Sun over the equatorward horizon is lower. This would seem to directly compete with the landing/take off economy of equatorial sites. But keep in mind, with the moon's lethargic rate of rotation, the touted desirability of equatorial sites is grossly exaggerated.
More significant an asset than shade is true "lee" vacuum, where there are surfaces never exposed to the lunar sky, and thus always protected from cosmic radiation, solar ultra-violet, solar storms, and the micrometeorite rain as well as wild day-night sunshine-shade temperature swings. Such areas will be ideal for warehousing and garage space and unpressurized industrial operations. They exist underground.
The moon has no limestone caves made by running and dripping water. But it does have lava tubes on the order of many tens of meters wide and high, many tens of kilometers long. These substantial lee voids are currently known only from indirect, yet indisputable evidence. Winding valleys, aka sinuous rilles, are a related feature, made from rivers of very fluid lava. Many rille valleys have bridged sections that suggests the visible valleys are near-surface lava tubes with collapsed roofs and that the "bridges" are intact tube sections. Elsewhere we see winding chains of rimless craters that can only be collapse pits where parts of a largely intact lava tube below have fallen in. The inference is that elsewhere, there are wholly uncollapsed lava tubes. As the mare basin-filling lava sheets were laid down in distinct episodes with lava tubes likely forming in each layer, there may be many intact lava tubes well below the surface layers in some lunar seas.
Where are these lava tubes and their "lee vacuum" to be found? In the maria, mostly near coastal areas! While we are a long way from identifying all such features, we can locate a base in a coastal region with partially collapsed rilles in the likelihood of finding usable intact tube sections nearby.
Quite a different consideration is ease of surface transport between a base site and other major areas of the moon with comparable assets. A coastal site on one of the "chain" of mare basins on the Nearside seems best (leaving out Crisium and other highland-locked seas). On Earth, some locations seem born to host major towns. To give a few examples, straights and narrows (Singapore, Detroit), major river confluences (St. Louis, Pittsburgh, Wuhan), major harbors (San Francisco, Rio de Janeiro, New York), lakeheads (Duluth), lake ends (Buffalo) and interlake constrictions (Niagara Falls), places where water routes and overland routes converge (Chicago), etc.
Now on the moon, we have no bodies of water or waterways. But we do have a chain of Seas or maria across which the going is easier -- at least in general. Here and there on the maria, lava flow front scarps and rilles and occasional craters block arrow-straight travel, forcing bypasses. Places where such obstacle negotiation becomes easier recommend themselves as strategic sites. Coastal highland promontories and headlands (e.g. the cusps of Sinus Iridium in NW Mare Imbrium) are also route narrowing spots. There are "straights" and channels through which traffic must move, like the Alpine Valley that links northern Mare Imbrium and Mare Frigoris.
Craters with breached walls have interior assets more easily accessible than those of others. Passes through scarps and mountain chains also lure the town planners. Similarly, if highland-locked seas seem to include otherwise desirable townsites, there will be spots along their coasts from which a route through adjoining highland areas may be relatively easier to negotiate. Such spots too would claim attention.
One site is not "just as good as" any other -- except to one grossly unfamiliar with the moon. We could know more than we know today before making a final selection. But if we carefully weigh all we now know about variously advantaged locations, we can pick a good site viable long term, even if the main thrust of industrial lunar activity occurs elsewhere.
It is perhaps decades too early to tell whether some twist in lunar development will add a strong tilt to this or that location, perhaps even despite a low score on the test points above. A "go" for lunar solar power arrays on the east and west equatorial limbs would be one such all-bets-off eventuality.
Eventually, there are sure to be a good plurality of developing settlement sites around the moon. Our point, then, is not to be sure to pick the very best site for the first outpost, but just to pick a site good enough to remain actively occupied well into the future. We need to take the task seriously, but not so seriously that we lose sleep agonizing over it.