Why China and the US Are Racing to Shackleton Crater at the Moon’s South Pole

Near the bottom of the Moon there is a crater about 21 kilometers (13 miles) wide whose floor has probably not seen sunlight in billions of years. A few kilometers away, on its rim, the Sun almost never sets. That odd pairing, endless night beside near-endless day, has made Shackleton crater one of the most sought-after spots in the solar system. As of October 2026, the United States and China are both building toward it.

Neither country is planting a flag there soon. China’s Chang’e 7 robotic lander, which was expected to aim for Shackleton’s rim, was postponed on Aug. 23, 2026, the day before launch, and is now expected in 2027. NASA reshuffled Artemis this year so that Artemis III, in 2027, stays in Earth orbit, and the first crewed landing near the south pole moves to Artemis IV in 2028. Below we cover why this one corner of the Moon matters so much, what each side is actually flying, and why “first” counts for more than bragging rights.

A crater on the axis of the Moon

The Moon’s spin axis is tilted only about 1.5 degrees relative to the plane of its orbit around the Sun (Earth’s is about 23.5 degrees). The Moon has no real seasons, so near the poles the Sun just circles low along the horizon all year. The ground there is lit from the side, all the time.

That geometry produces two kinds of land. Deep craters have floors the Sun never reaches; these are the permanently shadowed regions, or PSRs. NASA’s Lunar Reconnaissance Orbiter (LRO) has measured temperatures in some of them below about 40 kelvin (−233°C), which ranks them among the coldest places measured anywhere in the solar system. High ground works the opposite way. Ridges and crater rims stick up into the sunlight and stay lit for most of the year. Planetary scientists call these spots the “peaks of near-eternal light.” There are no spots with truly eternal light, but some points near the south pole are lit for well over 80 percent of the year in LRO-based illumination models.

Shackleton is about as close as you can get to having both in one place. It is roughly 4 kilometers deep, its interior sits in permanent shadow, and the lunar south pole itself lies on its rim. A ridge that connects Shackleton’s rim to the nearby de Gerlache crater holds some of the best-lit ground near the pole. If you stand on the rim, the solar panels are in sunlight and the ice is a short drive away.

The ice: what we actually know

People suspected for decades that the poles held ice. The direct evidence came in the 2000s and 2010s, from several missions:

• Chandrayaan-1 (India, 2008–09). NASA’s Moon Mineralogy Mapper flew on India’s first lunar orbiter and detected hydration signatures across the surface. A 2018 reanalysis of its data in PNAS reported direct evidence of exposed water ice inside shadowed areas at both poles.
• LCROSS (NASA, 2009). On Oct. 9, 2009, the Lunar Crater Observation and Sensing Satellite crashed a spent rocket stage into the shadowed Cabeus crater near the south pole and then flew through the debris. The plume contained water. The science team’s estimate was roughly 5.6 percent water ice by mass in the soil that was thrown up, plus other volatiles.
• LRO (NASA, 2009–present). LRO is still mapping the poles. In 2012, a team using its laser altimeter reported in Nature that Shackleton’s floor is brighter than its surroundings, consistent with ice making up perhaps as much as 22 percent of the surface material. The authors were careful to say other explanations are possible.

That last caveat matters. We know there is water in the polar cold traps. We do not yet know how much there is, how deep it goes, or whether it is pure enough to be worth digging out. Nobody has landed inside a PSR and checked. Chang’e 7 was built for exactly that job, and NASA’s VIPER ice-prospecting rover could fly as soon as 2027, according to the Planetary Society.

Why ice is worth a space race

Launching mass from Earth is very expensive, and water is heavy. A crew of four uses a lot of it for drinking, cooking, hygiene, and cooling, and every liter that doesn’t have to be launched is a liter’s worth of payload freed up for something else. Water is also two other things:

• Air. Electrolysis splits water into hydrogen and oxygen, and the oxygen can be breathed.
• Rocket fuel. Liquid hydrogen and liquid oxygen are the same propellants that power the core stage of NASA’s Space Launch System. A polar ice mine with an electrolysis plant would amount to a fuel depot sitting a fraction of Earth’s gravity well away from where the fuel gets used.

This is what “in-situ resource utilization” (ISRU) means in practice: living off the land. It is the gap between a camping trip and a settlement. China’s Chang’e 8 is explicitly an ISRU test mission, and NASA’s long-term Artemis plans depend on the same idea.

What about helium-3?

Moon-race stories often mention a second prize: helium-3, a light isotope of helium that is extremely rare on Earth. Because the Moon has no thick atmosphere or global magnetic field, the solar wind has spent billions of years implanting helium-3 into the grains of lunar soil, especially titanium-rich ilmenite, as Space.com explains. That happens across the whole sunlit surface. Helium-3 is spread thinly through the regolith almost everywhere and is not concentrated at Shackleton, so getting it out means digging up and heating very large amounts of soil.

The appeal is fusion. In principle, deuterium–helium-3 fusion releases most of its energy as charged particles rather than the neutrons that damage reactor walls in deuterium–tritium designs. Side reactions still produce some neutrons, though, and nobody has yet built a working helium-3 fusion reactor. Most fusion projects use deuterium and tritium.

China has talked about helium-3 for a long time. Ouyang Ziyuan, the founding chief scientist of the Chang’e program, told the BBC in 2013 that lunar helium-3 could meet human energy demand “for around 10,000 years at least.” After Chang’e 5 returned its samples, the Beijing Research Institute of Uranium Geology measured their helium-3 content and the best extraction temperatures, China News Service reported in September 2022.

In the U.S., the interest is commercial. Seattle startup Interlune won a $6.9 million NASA contract in May 2026 to build a payload that measures and extracts solar-wind gases, aimed at a 2028 mission. GeekWire reports its near-term buyers want helium-3 for quantum-computing cryogenics, not for fusion.

The upshot: ice is the reason to go to Shackleton now. Helium-3 is a long-term bet that depends on fusion technology that doesn’t exist yet.

Power and a view of home

Most of the Moon goes through roughly 14 Earth days of sunlight followed by 14 days of darkness. Solar-powered hardware has to survive two weeks of night at temperatures far below freezing, which is why many landers have died at their first sunset. On a well-chosen polar peak, the dark periods are much shorter. A solar array mounted vertically, turning to follow a Sun that stays near the horizon, can collect power almost continuously.

Communications are the other half. From much of the south polar region, Earth hangs low on the horizon. High ground near Shackleton keeps a direct radio line to Earth more of the time, while sites deeper in the terrain may need a relay satellite. China already has one: Queqiao-2, launched in 2024, which the Planetary Society notes will support Chang’e 7.

Prime real estate is scarce

All of this adds up to a short list of places that work. A good polar landing site needs gentle slopes, a decent amount of sunlight, a line of sight to Earth, and shadowed terrain within reach. Very few spots meet all four. When NASA narrowed its options in 2024, it named nine candidate regions near the south pole: Peak near Cabeus B, Haworth, Malapert Massif, Mons Mouton Plateau, Mons Mouton, Nobile Rim 1, Nobile Rim 2, de Gerlache Rim 2, and Slater Plain. Those were chosen for the landing then planned for Artemis III. NASA’s Artemis IV page now points to the same nine regions, and NASA says the final site depends on the launch date, because lighting and trajectories change through the year.

Shackleton’s own rim is not on NASA’s list, but de Gerlache Rim 2 sits at the other end of the same sunlit ridge. In practice, American and Chinese plans aim at the same small patch of the Moon.

Nobody can own it, but you can get there first

Legally, the Moon is off limits to ownership. Article II of the 1967 Outer Space Treaty, which both the United States and China have ratified, says outer space, including the Moon, “is not subject to national appropriation by claim of sovereignty, by means of use or occupation, or by any other means.” Article IX also requires states to act with “due regard” for one another’s activities.

The U.S.-led Artemis Accords add a practical layer on top. Section 11 lets signatories announce “safety zones” around their operations, areas where others should coordinate before coming close, so that a rocket’s exhaust doesn’t sandblast a neighbor’s equipment. The Accords say these zones are temporary, sized to the activity, and not a sovereignty claim. San Marino became the 76th signatory in September 2026.

Critics, including China, which has not signed, argue that the effect is still first come, first served. Whoever lands first on a sunlit ridge with a view of a cold trap gets to set the terms for everyone who comes after. There may be only a handful of such ridges, so a single well-placed base could, in practice, shape access to a large share of the best ground. That’s why the timeline matters.

The US side: Artemis, reshuffled

NASA’s year started with a success. Artemis II launched on April 1, 2026, and carried Reid Wiseman, Victor Glover, Christina Koch, and Canada’s Jeremy Hansen around the Moon. The crew reached 252,756 miles from Earth, a new human distance record, and splashed down off San Diego on April 10.

Before that flight, NASA had already reworked the plan. Artemis III, now set for 2027, will not land. It will test systems in low Earth orbit, including rendezvous and docking with one or both commercial landers, life support, and the new spacesuits, ahead of a landing on Artemis IV in 2028. NASA said it would then aim for at least one surface landing every year. In June, NASA named the Artemis III crew: NASA astronauts Randy Bresnik, Frank Rubio, and Andre Douglas, plus ESA’s Luca Parmitano. NASA also said a Blue Origin lander pathfinder would launch first and wait in orbit for Orion. On Oct. 13, NASA reported that booster stacking was complete and all four RS-25 engines were installed on the core stage.

The landers are the hard part. NASA’s Human Landing System program has two contractors. SpaceX is building a lunar version of Starship, which needs a series of tanker flights to refuel it in orbit before it can leave for the Moon. Blue Origin is building Blue Moon, including a smaller robotic Mark 1 cargo lander. On Oct. 5, Artemis III astronaut Andre Douglas told Reuters the mission could still launch next summer but that “anything can happen.” Reuters noted that some industry sources doubt the landers will be ready. Blue Origin also lost a New Glenn rocket and damaged its Florida launch pad in a static-fire accident in May, which Ars Technica reported as a setback for Blue Moon’s first flight.

The China side: Chang’e, Lanyue, and a 2030 deadline

China’s robotic program has a strong record. Chang’e 4 made the first landing on the far side in 2019, Chang’e 5 returned near-side samples in 2020, and Chang’e 6 brought back the first far-side samples in 2024. Chang’e 7 was supposed to extend that record to the pole. It has an orbiter, a lander, a rover, and a hopper designed to jump into shadowed craters and sniff for water. It carries 18 science payloads, including instruments from Egypt, Bahrain, Italy, Russia, Switzerland, and Thailand. The Planetary Society says China hasn’t announced a final landing site, but the preferred candidate is Shackleton’s rim, with a targeted landing accuracy better than 100 meters.

On Aug. 23, 2026, Chinese authorities said Chang’e 7 “does not meet the conditions for launch” during this year’s window. Reporting from Space.com and Ars Technica pointed to a tropical storm near the Wenchang launch site. As of mid-October, China has not announced a new date. Unofficial chatter cited by Ars points to February or March 2027, which would be followed by months of reconnaissance from lunar orbit before any landing.

Chang’e 8 comes next. China’s space agency says it will launch around 2029 and test resource-use technologies as a building block for the International Lunar Research Station (ILRS). China leads the ILRS with Russia as its main partner, and the two have discussed a lunar nuclear power plant by 2035.

The crewed program is moving too. China’s stated goal is to land astronauts on the Moon before 2030. On Feb. 11, 2026, it flew a prototype Long March 10 rocket and carried out an in-flight abort test of the new Mengzhou crew capsule at maximum aerodynamic pressure. Earlier tests included a landing-and-takeoff test of the Lanyue lunar lander. The mission plan uses two Long March 10 launches, one for Lanyue and one for Mengzhou. The two spacecraft dock in lunar orbit, and two astronauts ride Lanyue down to the surface. China has not named a crewed landing site.

Video: “What Is Actually Happening With China’s Moon Base” (The Space Race on YouTube, August 2026), a walk-through of Shackleton, its connecting ridge, and the Chang’e 7 and NASA landing targets.

So who gets there first?

That depends on what counts as “there.” No one has yet landed a spacecraft right at a lunar pole. India’s Chandrayaan-3 touched down at about 69 degrees south in 2023, and Intuitive Machines’ Athena reached the Mons Mouton area in 2025 but tipped over on landing. The first robotic landing near Shackleton itself is now likely in 2027, and Chang’e 7 and Blue Moon Mark 1 could end up arriving within months of each other. For people on the surface, NASA’s target is 2028 and China’s is before 2030. NASA’s own reshuffle is a reminder that both dates depend on landers that haven’t flown yet.

The bigger contest is over who sets the norms. The Artemis Accords and the ILRS are competing ideas about how countries should behave on the Moon, and each bloc is signing up partners. Smaller space nations that put an instrument on Chang’e 7 or sign the Accords are choosing a side, at least in part.

Other countries are heading south too. India’s Chandrayaan-4 sample-return mission is planned to target the south polar region, and the joint India–Japan Chandrayaan-5/LUPEX rover is aimed at no earlier than 2028 to study water near shadowed regions. Both India and Japan have signed the Artemis Accords.

The bottom line

Shackleton matters because it combines, within a few kilometers, the things a long-term lunar base needs: water you might mine, sunlight you can count on, and a view of Earth. No treaty lets anyone own it. Whoever arrives first, though, gets to choose where to build and to declare the safety zones around it. That’s why a launch postponed by a storm in Hainan and a lander docking test planned for Earth orbit are both being watched as moves in the same race. It’s also part of a larger shift in American spaceflight toward commercial hardware, the same shift behind NASA’s push to replace the ISS with commercial space stations.

Further reading