The South Pole Crater Question — Shackleton, Chang'e 7, and Artemis III

Document 3 of the Moon set. The number of viable landing sites at the lunar south pole is small, perhaps a dozen, with the best concentrated in a few square kilometres. Multiple programmes are targeting the same area in 2026-2028 under different legal frameworks with no working coordination mechanism. Five scenarios for what happens next, the publication's reading of which are likely, and what proactive coordination would look like.

This piece is about a single small piece of the lunar surface and what is going to happen at it.

The lunar south pole — specifically, a few peaks on the rim of Shackleton crater and the floor of the crater itself, together with a handful of similarly favourable locations within roughly 200 kilometres — is the most contested piece of off-Earth real estate in human history. Both NASA's Artemis III mission and China's Chang'e 7 mission have selected overlapping landing zones in this area. The Indian Chandrayaan-3 successor mission is targeting nearby sites. The Russian Luna programme is targeting nearby sites. The European Space Agency's Argonaut lander is targeting nearby sites.

The contention is not a feature of geopolitical rhetoric. It is a feature of physics. The Moon is tilted on its axis by only 1.5 degrees (compared with Earth's 23.5). The Sun therefore skims the horizon at the poles all year and never rises far above it. A small number of peaks at the south pole are illuminated for 90% or more of the year. Adjacent to those peaks, the floors of certain craters are in permanent shadow — they have not seen sunlight in billions of years and their temperatures are colder than the surface of Pluto.

The combination matters. The illuminated peaks provide continuous solar power. The shadowed floors hold water ice. A base at the right combination of peak-and-crater has power, water, propellant feedstock, and a stable thermal environment for instrumentation that needs cryogenic temperatures naturally.

There are perhaps a dozen such locations. The best are concentrated in a much smaller number — a handful, by some assessments. The peer-reviewed analyses converge on the same short list. The Chinese mission planners and the NASA mission planners have read the same papers and reached the same conclusions about which sites are best. This piece is about what happens next.

The geography

The principal candidate locations at the lunar south pole, as of 2026:

Connecting Ridge / Peak Near Shackleton. A ridge on the rim of Shackleton crater that receives sunlight for approximately 90% of the lunar year. Adjacent to the permanently shadowed floor of Shackleton, where temperatures below 90 Kelvin are sustained and water ice has been inferred from neutron-spectrometer data. NASA's Artemis III mission has identified this as a candidate site. China's Chang'e 7 has identified its target as "the illuminated rim of Shackleton crater" — overlapping with the NASA candidate region.

de Gerlache Rim and crater interior. Approximately 70 kilometres from Shackleton. Similar peak-and-shadow combination. Also on NASA's Artemis III candidate list.

Haworth Crater. One of the largest permanently shadowed regions in the south polar area. Less illumination on adjacent ridges than Shackleton or de Gerlache; more shadow surface area.

Faustini, Cabeus, Shoemaker, Sverdrup. Other permanently shadowed craters. Each has been observed to contain volatile compounds, including water ice. Cabeus is where the LCROSS impact in 2009 confirmed water ice presence.

Malapert Massif. A large mountain near the south pole with extensive sunlit area. The leading candidate for a longer-term base location rather than a single-mission landing, on some assessments.

"Peaks of eternal light." The mythological phrase for permanently sunlit locations is a slight overstatement; no peak is sunlit 100% of the year due to lunar libration and small-scale topography. But there are at least three points on Shackleton's rim that are collectively sunlit for more than 90% of the year, with the dark intervals being short enough to be bridged by battery storage. These are the most valuable real estate.

The total surface area of "best" sites — combining the illuminated-peak constraint with adjacent permanently-shadowed-region access and reasonable terrain for landing — is, on most analyses, in the low square kilometres. Not the low hundreds. Not the low tens. Low single digits.

What is being planned, by whom, when

NASA Artemis III. Scheduled crewed landing late 2027. Candidate landing sites are publicly documented; the leading candidates include Connecting Ridge near Shackleton and de Gerlache Rim. Final site selection is operationally driven and depends on Artemis II results, lander readiness (SpaceX Starship HLS), and ongoing surface characterisation by precursor missions including the Chinese-launched orbital data, NASA's Lunar Reconnaissance Orbiter, and (where available) joint scientific data.

China Chang'e 7. Launch scheduled for 2026 on a Long March 5. The target is the illuminated rim of Shackleton crater. The mission carries a lander, a rover, a relay satellite (Queqiao 2), and — uniquely — a "mobile hopper" designed to leap between sunlit ridge locations and permanently shadowed floor locations carrying a molecular analyser to search for water ice in situ. The mission is technically ambitious and operationally focused on the same surface area Artemis III is targeting.

China Chang'e 8. Scheduled 2028. In-situ resource utilisation (ISRU) demonstration. Building blocks for the International Lunar Research Station to follow in the 2030s.

International Lunar Research Station (ILRS). A China-Russia-led programme for permanent presence at the lunar south pole. Phased construction through the 2030s. Currently has agreements with Russia, Belarus, Pakistan, Venezuela, South Africa, Egypt, and others. The ILRS coalition is broadly the non-Artemis coalition; the two groupings do not formally compete but do not coordinate.

India Chandrayaan-4. Lunar sample return targeting south polar regions. India is an Artemis Accords signatory (2023). Site selection coordination with NASA is in progress; there has been less public conflict here than between the Artemis coalition and the Chinese programme.

Japan, UAE, Korea. Smaller missions, generally either coordinated under Artemis or in parallel without conflict.

The aggregate picture: at least three crewed or robotic mission programmes are targeting the same small surface area in 2026-2028, with the China-Russia-led ILRS planning permanent presence at the same general area from the 2030s.

What makes this politically combustible

The technical similarity of the missions is, paradoxically, the source of the political risk. Different missions targeting different sites would be mostly uncontroversial. Multiple missions targeting the same site, in proximate timing, under different legal frameworks, with no working coordination mechanism, is the configuration that produces friction.

The "safety zone" question becomes immediate. The Artemis Accords' Section 11 commits signatories to notify each other of activities and to coordinate around safety zones. China is not a signatory. If Chang'e 7's lander operates on the illuminated rim of Shackleton in 2026, and Artemis III's HLS targets a nearby location in 2027, the Artemis mission cannot impose a safety zone on Chang'e 7 hardware and Chang'e 7 cannot impose one on Artemis III. Both can do what they want under the OST. The OST does not have a coordination provision.

The "first arrival" question is real. The first operator to establish working infrastructure at a particular site sets practical facts. The second operator's options shrink. This is not about legal title to lunar surface — under the OST, no one can have legal title. It is about the operational reality that a working power-and-communications relay at a peak, with attached water-extraction infrastructure in an adjacent crater, occupies the location in the way a kitchen occupies a kitchen. The OST's free-access principle does not address whether the kitchen has to share.

The "harmful interference" question is undefined. OST Article IX commits states to "due regard" for the interests of other states and to avoid "harmful contamination" and "harmful interference." Both terms are undefined. Does landing 200 metres from another state's operating equipment constitute harmful interference? Does a rocket exhaust plume that distributes regolith across another state's solar panels constitute harmful contamination? Does sampling water ice from a permanently shadowed region that another state has surveyed but not yet operated in constitute interference with their planned activities?

None of these has a binding answer. The Artemis Accords offer non-binding guidance. The OST offers the principle without the operationalisation. The result is that decisions about what constitutes acceptable behaviour will be made by the operators themselves, in real time, with diplomatic protests as the only feedback mechanism.

The "heritage" question is latent. The Apollo 11 landing site is at the Sea of Tranquility, far from the south pole. But the south pole region contains the Soviet Luna landing sites and will, after the next few years, contain hardware from Chinese Chang'e missions and NASA Artemis missions. Disturbance of these sites — by exhaust plume, by surface mobility, by surface mining — is not directly prohibited by any binding rule. The Artemis Accords' Section 9 commits signatories to preservation; it does not bind non-signatories and does not specify what preservation means.

What could happen in 2026-2030

Five scenarios, presented in rough order of likelihood:

Scenario one — uneventful coexistence. Chang'e 7 lands at Shackleton in 2026, gathers data, leaves. Artemis III lands at a nearby but different site in 2027 or 2028, gathers data, returns crew. Multiple operations occur in the same general area but do not collide. Each programme learns from the other's data; informal scientific coordination is real even where formal political coordination is not. This is the optimistic scenario. It is genuinely possible if both programmes execute conservatively and the political climate does not deteriorate. The publication's view: this scenario is plausible but not assured.

Scenario two — practical coordination emerges. A working "scientific exchange" mechanism develops between the Artemis coalition and the China-Russia programme — perhaps initially through third parties (UN COPUOS, the International Astronautical Federation, the International Astronomical Union) and then through direct bilateral exchanges. Site coordination becomes routine. The legal framework remains contested but operational practice fills the gap. This has analogues in international science cooperation through politically difficult periods (the US-USSR scientific cooperation during the Cold War; the ISS-era cooperation through episodes of strained US-Russia relations).

Scenario three — minor incident, diplomatic management. An Artemis-coalition mission and a Chinese mission have a near-miss at a south polar landing site. Or a regolith-plume event damages instrumentation. Or a safety-zone notification is ignored and operations occur within the notified area. The diplomatic response is intense, the operational response is to share more data and to schedule more carefully, and the legal framework is amended at the margins (perhaps through a new UN General Assembly resolution that nominally applies to all states without being binding). Nothing fundamental changes.

Scenario four — serious incident, framework crisis. A collision, a fatality (worst case, on a crewed Artemis mission), or a deliberate disruption produces an actual crisis. The legal framework is found inadequate. Either a formal treaty negotiation begins — possibly through a revived COPUOS process — or the bifurcation between Artemis and ILRS hardens into mutual non-recognition of each other's operational rights. The latter is the more likely outcome on current political trends; the former would require a propitious moment that does not currently exist.

Scenario five — strategic competition becomes overt. The competition at the south pole is openly framed as a contest, with the framing being that whichever programme establishes operating infrastructure first will set the rules. Military doctrine begins to include space-domain awareness of cislunar space as a contested area. Dual-use infrastructure is built to specifications that include the option of denial of access to adversary operators. The OST's "peaceful purposes" clause is interpreted more broadly to permit such infrastructure. This is the worst scenario in long-run terms; it is not the most likely on current evidence but it is not implausible.

The publication does not predict which scenario will dominate. It notes that scenarios one and two require deliberate effort from both sides; scenarios three and four are what the current trajectory produces with no intervention; scenario five is what happens if either side decides the strategic stakes are high enough to justify the framing.

What the publication would do, if asked

The publication does not advocate. But on this specific question, it notes that the case for proactive coordination is unusually strong:

The cost of a coordination failure is high — measured in mission cost, scientific cost, and political cost. The cost of coordination is low — measured in mission cost, scientific cost, and political cost. The asymmetry is, by the publication's read, the largest on any open question in current space policy.

The mechanism for coordination does not yet exist. The OST's principles are right; the Artemis Accords' operationalisation is one step in the right direction; the absence of any formal coordination mechanism that binds China and the Artemis coalition is the gap. A track-1 government-to-government channel for south-pole operations specifically — perhaps administered through a neutral third party — would be small, cheap, and high-value. It does not need a new treaty; it needs a working group with operating authority on a specific question.

The publication's frame, named openly: scenarios one and two are achievable. The default trajectory is towards scenarios three, four, and five. Closing the gap requires people to do something about it, and currently no one is.

What this piece does not conclude

It does not say which site is best. It does not predict which programme will land first or land most successfully. It does not say that China's programme is more or less legitimate than the US programme. It does not say that the Accords are sufficient or insufficient. It does not predict timeline failures or successes.

It does say that the south polar landing sites are a finite resource, that the legal framework for sharing them is inadequate, and that the consequences of the current configuration are foreseeable. What is foreseeable is, in matters of state behaviour, sometimes avoidable. The publication takes no view on whether it will be avoided in this case.

For the legal framework that underlies all of this, see the companion piece on the Treaty Framework. For the broader question of what the lunar resource economy actually is, see the piece on Helium-3 and the Fusion Argument. For the public-brief overview of the whole question, see the Moon public brief.