Does Shackleton Crater Really Have Water Ice? What 2026 Data Shows

By James  ·  Published May 3, 2026  ·  Updated September 4, 2026

Shackleton Crater is the single most talked-about piece of ground on the Moon. China's Chang'e-7 named the crater's sunlit rim as its preferred landing area. NASA's 2022 shortlist of Artemis landing regions included a site called Peak Near Shackleton. Nearly every article you will read about the lunar south pole treats the reason as settled: there is water ice down there, and water ice becomes drinking water, breathable oxygen, and rocket propellant.

Here is the part those articles tend to skip. Water ice at the lunar south pole is confirmed — but it was confirmed in Cabeus, a different crater, in 2009. Inside Shackleton itself, the ice has never been directly confirmed at all. And in March 2026, an orbital survey using a camera built specifically to see into permanently dark craters came back without finding the abundant surface ice that a decade of mission planning had assumed was waiting.

That gap between what is confirmed and what is assumed is what this article is about. Below: where the ice has genuinely been measured, why Shackleton became the target anyway, what the newest data changed, how engineers plan to turn ice into fuel, and why two national programs pushing into the same small region have no dedicated mechanism for sorting out what happens if their operations overlap.

At a Glance

• Water ice is confirmed at ~5.6 wt% — but in Cabeus crater, measured directly in the LCROSS ejecta plume (Colaprete et al., Science, 2010)
• Shackleton's rim receives sunlight 80–90% of the year while its floor has likely seen none in a billion years. That lighting is what made it a target
• Inside Shackleton, substantial ice has never been confirmed. A 2025 reanalysis reads the crater's bright interior as rock, not ice sheets
• A March 2026 ShadowCam survey turned up no ice at concentrations the instrument could resolve — a floor of 20–30 wt% — only a few patches 20–50 m across. Low-concentration and buried ice are not ruled out
• NASA dropped Peak Near Shackleton from its Artemis shortlist in October 2024 — for reasons tied to lander performance, not to ice
• NASA's lunar reactor program now carries two figures at once: a 100 kW goal set in 2025, and a 20 kW flight floor set by an April 2026 White House directive
• China's Chang'e-7 — the mission built to settle the ice question — was postponed in August 2026 with no new date announced

What Is Confirmed: Water Ice in Cabeus Crater

Start with what is solid, because the solid part is genuinely remarkable. The Moon's rotational axis is tilted only about 1.54 degrees relative to the ecliptic — the plane of Earth's orbit around the Sun — compared with Earth's own 23.4-degree tilt, the one that gives us seasons. Because that axis sits almost upright, the Sun never climbs more than about a degree and a half above the horizon at the lunar poles. The floors of certain deep craters, screened by their own rims, therefore receive no direct sunlight — ever.

NASA's Lunar Prospector mission mapped hydrogen concentrations across these permanently shadowed regions (PSRs) — crater floors that, because of that small axial tilt, never receive direct sunlight — using neutron spectroscopy, identifying enhancements at both poles consistent with water ice trapped in the regolith (Feldman et al., Science, 1998). India's Chandrayaan-1 spacecraft, carrying NASA's Moon Mineralogy Mapper (M3) instrument, then provided near-infrared spectral signatures of surface ice on polar crater floors — a different detection method, yielding the same conclusion (Pieters et al., Science, 2009).

The most direct confirmation came from NASA's LCROSS mission, which deliberately crashed a rocket stage into Cabeus crater near the south pole in 2009 and analyzed the resulting ejecta plume. The plume showed water at concentrations of approximately 5.6 ± 2.9 weight-percent (Colaprete et al., Science, 2010). That is not a trace signal. It is enough ice mixed into the regolith to matter for extraction planning.

Cabeus crater — not Shackleton. This is where the LCROSS impact confirmed water at roughly 5.6 wt% in the ejecta plume. The permanently shadowed floor (dark blue) sits near −248 °C, cold enough to trap ice for billions of years. Credit: NASA LRO / LOLA

Two details in that record get flattened in most retellings. First, the ice sits in permanently shadowed regions at both the north and south poles, not exclusively the south (Hayne et al., Nature Astronomy, 2020). Second, and more consequentially for everything that follows, the direct confirmation came from Cabeus — a specific crater, not the region as a whole, and not the crater that ended up on every mission planner's map.

The temperatures explain why that ice has stayed put. NASA's LRO Diviner instrument records PSR floor temperatures as low as 25–40 K (roughly −233 °C to −248 °C) — among the coldest confirmed environments in the entire solar system (Paige et al., Science, 2010). At those temperatures, water molecules that migrated into shadow billions of years ago have never had enough energy to escape. Contrast that with the lunar equator, where daytime surface temperatures exceed 120 °C (Vasavada et al., Icarus, 2012). Ice there cannot remain stable over geologically meaningful timescales.

How much is down there in total? Estimates diverge sharply depending on the modeling approach. NASA's original 1998 Lunar Prospector analysis cited concentrations consistent with up to an estimated 6 billion metric tons. Later work, using different assumptions and larger survey datasets, produced a broader range — sometimes cited as 1 to 10 billion metric tons.

Every one of those figures carries substantial uncertainty, and all of them come from orbital survey modeling rather than ground-truth measurement. The ice is buried, unevenly distributed, and mixed with regolith in ways no in-situ survey has yet resolved. A spread of a full order of magnitude is not a footnote. It means the total has never been measured — only modeled.

So Does Shackleton Crater Have Water Ice? The Honest Answer Is: Not Confirmed

Shackleton became the focal point for mission planners because of geometry, not because of a confirmed ice measurement. Those are different things, and the distinction has blurred along the way.

The crater sits almost exactly at the lunar south pole, with the pole itself lying on its rim. It is roughly 19 to 21 kilometers wide and more than 4 kilometers deep, with steep inner walls and a bowl-shaped floor that LROC imaging confirms is in near-permanent shadow (Spudis et al., 2013). Lunar Reconnaissance Orbiter illumination data shows that certain elevated points on the rim catch sunlight more than 80 to 90 percent of the time over a lunar year. No single point is permanently lit, but the best ridgelines come close.

That is the pairing everyone wants. A rim in almost continuous sunlight sits directly above a floor that has not seen a ray of it in perhaps a billion years — solar power and a cold trap within a few kilometers of each other, the closest thing to an integrated power-and-water site anywhere on the lunar surface. On paper it is close to ideal.

Shackleton Crater. False-color elevation: red = sunlit rim, blue = permanently shadowed floor. The vertical contrast between those two zones is why the crater dominates south-pole planning — and it is a case built on lighting, not on measured ice. Credit: NASA SVS / LRO

On paper, though, is where the ice case stops. A 2025 analysis published in the Journal of Astronomy and Space Sciences examined ShadowCam and M3 data and concluded that the high reflectance seen inside Shackleton's shadowed interior fits plagioclase-rich rock and mass-wasting material better than it fits large sheets of water ice. That does not rule out ice dispersed through the regolith. It does mean the bright signal people had been reading as ice may never have been ice.

Then, in March 2026, a team led by Shuai Li at the University of Hawaiʻi at Mānoa published a broader survey in Science Advances. Their instrument was ShadowCam, roughly 200 times more sensitive to light than the orbital camera it was derived from, which is what lets it work inside craters the Sun never reaches. The team searched the south polar PSRs for two optical fingerprints at once: higher reflectance, and a distinctive forward-scattering behavior that rock does not produce.

Nothing turned up at the scale they were looking for. Across the shadowed regions surveyed, the data showed no ice at concentrations the instrument could resolve, and that floor sat at 20 to 30 weight-percent. What did turn up was small and scattered: a handful of bright exposures, 20 to 50 meters across, showing both signatures together. The team describes those few spots as consistent with concentrations above roughly 10 percent, while stopping short of calling them confirmed ice.

The result carries two limits worth stating up front, because both get lost in the headlines. Ice below the detection threshold is not ruled out; the authors say explicitly that widespread low-concentration ice remains possible. And ice buried beneath the uppermost regolith is outside what any camera can see at all. Their own conclusion is that settling the question will take a future instrument sensitive to concentrations under 1 percent.

So the survey does not rule out ice. It rules out ice in the abundant, near-surface form that a decade of mission planning had taken for granted.

Which leaves the strange situation this article is named after. The crater everyone associates with lunar water is not the crater where lunar water was confirmed. Cabeus is.

NASA's current shortlist reflects that split, though not for the reasons it might appear to. When the agency revised its candidate landing regions in October 2024, Peak Near Shackleton was among those dropped. But the stated drivers were vehicle performance and trajectory analysis tied to the lander design, not any judgment about ice. And the revision came a year before the reanalysis of Shackleton's interior, eighteen months before the ShadowCam survey.

So read it as a coincidence worth noticing, not as NASA conceding the point. The case for water at the lunar south pole rests on the whole network of PSRs across the region, with one directly measured site inside it. It does not rest on Shackleton's floor. Shackleton was chosen for its lighting, and the ice was assumed to follow.

None of this is a case against going. It is a case about what the first crews will be doing when they arrive. If the ice is buried and dilute rather than exposed and concentrated, extraction means drilling and excavation, not scraping a surface layer — a different machine, a different power budget, a different mission. Which raises the question of where that power comes from.

The Lunar Night Problem — and NASA's Two Reactor Numbers

The sunlit ridges are promising, but they do not solve everything. At equatorial latitudes, the lunar day-night cycle follows the synodic period of roughly 29.5 Earth days: about 14 days of daylight, then about 14 days of darkness. Near the poles, topography changes the arithmetic — the best ridges near Shackleton have far shorter dark intervals — but libration and seasonal shifts in the Sun's angle still produce multi-day blackouts. No location on the Moon is permanently sunlit.

During those blackouts, temperatures across most of the lunar surface drop to about −130 °C. Equipment that is not actively heated will fail. Batteries, electronics, seals, and mechanical components all have operational limits that lunar night routinely exceeds.

Riding out 10 to 14 or more days of darkness on stored energy alone means launching every kilogram of that storage from Earth. For early missions the mass budget is not there. For a permanent installation it is a structural problem batteries cannot solve at any launch mass that makes operational sense. Solar will certainly be part of any south-pole architecture; it simply cannot carry the full load.

NASA's answer is the Fission Surface Power (FSP) program, run with the U.S. Department of Energy. If you have read about this before, you probably encountered a specific number: 40 kilowatts of continuous electrical output for at least ten years, enough to run about thirty households. That figure is real, and NASA program pages still cite it — but it dates from the 2022 design contracts, and the effort has been reorganized around much larger ambitions since. If you go looking and find conflicting numbers, this is why.

The lunar south pole environment — target site for NASA's Fission Surface Power demonstration. Any base here has to survive extended darkness on stored or generated power. Solar alone cannot carry that load. Credit: NASA

In August 2025, then-acting NASA administrator Sean Duffy directed the agency to design, build, and deploy a lunar surface reactor delivering at least 100 kilowatts of electrical power, ready for launch by the first quarter of fiscal 2030, using a closed Brayton cycle for power conversion. A December 2025 executive order folded that goal into a broader space nuclear initiative. On January 13, 2026, NASA and the Department of Energy signed a memorandum of understanding formalizing the partnership to deliver a lunar surface reactor by 2030.

Then, in April 2026, the White House Office of Science and Technology Policy issued the National Initiative for American Space Nuclear Power, and a second figure appeared alongside the first. The directive tells NASA to develop a reactor of at least 20 kilowatts and fly a variant to the lunar surface by 2030, while favoring architectures that can scale to 100 kilowatts or more, and to narrow the field to no more than two designs within a year.

The two numbers are not a contradiction so much as a division of labor. The 100-kilowatt goal describes where the program is meant to end up. The 20-kilowatt floor describes what a first-of-its-kind reactor actually has to deliver on a 2030 flight.

The distance between them still matters, though, and it maps onto the ice question. Twenty kilowatts keeps a small outpost alive through the dark. A hundred begins to cover excavation and processing hardware. Full-scale propellant production sits higher still — NASA's own roadmap puts that in the megawatt class.

From Lunar Water Ice to Rocket Fuel: How ISRU Would Work

Ice matters to mission planners for reasons that go well beyond drinking water or radiation shielding. Water, split into its component elements and cooled to liquid form, becomes one of the highest-performing rocket propellant combinations known: liquid oxygen and liquid hydrogen. Lunar ice could in principle be refined into fuel for missions departing the Moon, or stored in cislunar depots supplying spacecraft that never carry propellant up from Earth's surface at all.

A NASA technical assessment on in-situ resource utilization (ISRU) — producing usable materials from local resources instead of launching them — lays out the production chain: extract water from icy regolith, purify it, then split it into hydrogen and oxygen by electrolysis. Two approaches are under study. Proton-exchange membrane systems need high-purity water going in. Solid-oxide electrolysis systems run hotter and tolerate dirtier feedstock, at a cost of roughly 38 kilowatt-hours of electricity per kilogram of hydrogen produced.

That purity trade-off is where the ice question comes back around. Ice dispersed through regolith at a few weight-percent, mined by drilling, does not arrive clean. The dirtier the input, the more the architecture leans toward the higher-temperature, higher-power option — and the more the whole enterprise depends on which reactor figure NASA lands on.

NASA's ISRU maturation roadmap describes a scaling pathway from kilowatt-class demonstration units to megawatt-class industrial systems producing tens of kilograms of propellant per hour. The underlying logic is what engineers sometimes call breaking the tyranny of the rocket equation: the Moon's surface gravity is roughly one-sixth of Earth's, so propellant manufactured on the Moon and burned for lunar departure costs far less in energy terms than the same propellant lifted from Earth. Even a modest fraction of the estimated reserve, converted to fuel, would support a meaningful number of missions without resupply.

That is the shift that changes the Moon's role entirely — from a place you visit to a place you refuel at. A depot. A logistics hub. A forward platform. And it is why a crater's ice content stopped being an academic question and became a strategic one.

NASA vs. China at the South Pole: Overlapping Ground, Unresolved Law

The 13 original candidate landing regions near the lunar south pole, announced in 2022. The list was revised to nine in October 2024: five of the original regions carried over, four were new, and Peak Near Shackleton was not among them. Credit: NASA Scientific Visualization Studio

Between April 1 and 11, 2026, Artemis II carried four astronauts on a flyby of the Moon — the first crewed lunar mission since Apollo 17 in 1972. The landing timeline, however, had shifted two months earlier. On February 27, 2026, NASA announced a revised architecture: Artemis III was redesignated as a crewed test in low Earth orbit, rehearsing rendezvous and docking with commercial landers, targeted for 2027. The first crewed south-pole landing moved to Artemis IV, currently planned for 2028.

The candidate regions announced for Artemis III are expected to remain the target zones for that eventual landing. The 2022 round identified thirteen sites near the south pole, including Faustini Rim A, Peak Near Shackleton, Connecting Ridge, de Gerlache Rim 1 and 2, Haworth, Malapert Massif, Leibnitz Beta Plateau, Nobile Rim 1 and 2, and Amundsen Rim.

An October 2024 revision cut the list to nine: 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. Only five of the original thirteen carried over; four of the nine are new. Peak Near Shackleton is not among them.

China's Chang'e-7 was built to answer the ice question directly, and it never stopped pointing at Shackleton. The 8,200-kilogram stack carries an orbiter, a lander, a rover, and a mini flying probe designed to hop into a permanently shadowed crater floor with a water molecule analyzer aboard. Its stated preferred landing area is the illuminated rim of Shackleton, with published coordinates near 88.8°S, 123.4°E.

It has not launched. The spacecraft rolled out to the pad at Wenchang atop a Long March 5 on August 19, 2026, with a launch window opening August 24. Days later, with a tropical storm moving through the South China Sea, the China Manned Space Engineering Office announced that the probe did not meet launch conditions and that liftoff "cannot take place during the planned window this year".

The agency did not specify whether weather was the only factor, and as of early September 2026 no replacement date has been announced. Lunar transfer windows are unforgiving — the geometry does not come back around in a few days — and reporting since the announcement has pointed toward 2027, though that is press expectation rather than an official schedule.

Program	Target Area	Key Mission	Status (September 2026)
NASA Artemis	9 candidate regions near the south pole (2024 list). Shackleton-adjacent site dropped; a site near Cabeus B retained	Artemis III: crewed lander test in low Earth orbit (redesignated Feb 2026). Artemis IV: first crewed south-pole landing. Fission Surface Power demonstration	Artemis II flew April 2026. Artemis III targeted 2027; Artemis IV targeted 2028. Reactor launch-ready goal: 2030
China Chang'e-7	Illuminated rim of Shackleton crater (~88.8°S, 123.4°E)	South-pole water-ice survey; mini flying probe to sample a permanently shadowed floor directly	Postponed Aug 23, 2026. No replacement date announced; reporting points toward 2027

The legal framework meant to manage that proximity is contested. The Artemis Accords — 71 signatories as of August 31, 2026, when Türkiye became the most recent to sign — include provisions for safety zones, activity-specific areas around operations intended to prevent interference. The zones are explicitly framed as non-sovereign and temporary, drawing their legal basis from Article IX of the 1967 Outer Space Treaty, which requires states to conduct activities with "due regard" for other parties and to avoid interference that would harm them. The Accords themselves are not a treaty; they are a multilateral political agreement, what international lawyers classify as soft law. China has not signed them.

Some legal scholars argue the safety zones as described reach beyond what Article IX authorizes. Others argue they are a reasonable operationalization of an obligation that would otherwise stay abstract. The disagreement is unresolved. Article IX establishes the duty, but it supplies no procedure for adjudicating competing operations at the same lunar site — a gap with familiar parallels in how space debris liability works on Earth, where the rules were written for a far emptier sky.

Chang'e-7's delay buys some time on that question. It does not resolve it. Both programs are still working the same small stretch of polar terrain for the same reason — maximum rim illumination above maximum inferred ice potential — even if their shortlists no longer name the same crater. The polar geography that produced that convergence will not rearrange itself to accommodate two separate sets of national plans. And there is a certain irony in where things now stand: the one mission designed to settle whether the ice is there is the one sitting on the ground.

I grew up looking at the Moon the way most people did — with wonder, not strategy. The shift from one to the other happened quietly, somewhere between Apollo and Artemis, and I'm not sure any generation was given a clear moment to notice it. The hope, however naive, is that the next generation still gets to look up and see something that belongs to everyone. Whether that remains possible is not a scientific question. It is a political one, and it is still open.

Related on thesecom.net

• Space Debris Hit a Florida Home: What the Law Actually Says

About the Author James is a science writer covering astrophysics, space policy, and the history of scientific discovery. He writes for thesecom.net, where the goal is to explain what scientists actually found — not just what the headlines said they found. His work draws on peer-reviewed sources and primary mission documentation throughout. Editorial profile.

Sources & References

• Colaprete, A. et al. (2010). Detection of Water in the LCROSS Ejecta Plume. Science, 330(6003), 463–468. doi:10.1126/science.1186986
• Li, S. et al. (2026). Searching for surficial water ice in lunar permanently shaded regions (PSRs) with ShadowCam. Science Advances, 12. doi:10.1126/sciadv.aec8211
• Pieters, C. M. et al. (2009). Character and Spatial Distribution of OH/H₂O on the Surface of the Moon Seen by M3 on Chandrayaan-1. Science, 326(5952), 568–572. doi:10.1126/science.1178658
• Paige, D. A. et al. (2010). Diviner Lunar Radiometer Observations of Cold Traps in the Moon's South Polar Region. Science, 330(6003), 479–482. doi:10.1126/science.1187726
• Feldman, W. C. et al. (1998). Fluxes of Fast and Epithermal Neutrons from Lunar Prospector. Science, 281(5382), 1496–1500. doi:10.1126/science.281.5382.1496
• Hayne, P. O. et al. (2020). Micro cold traps on the Moon. Nature Astronomy, 5, 169–175. doi:10.1038/s41550-020-1198-9
• Vasavada, A. R. et al. (2012). Lunar equatorial surface temperatures and regolith properties from the Diviner Lunar Radiometer Experiment. Journal of Geophysical Research: Planets, 117, E00H18. doi:10.1029/2011JE003987
• Spudis, P. D. et al. (2013). Evidence for water ice on the Moon: Results for anomalous polar craters from the LRO Mini-RF imaging radar. Journal of Geophysical Research: Planets, 118(10), 2016–2029. doi:10.1002/jgre.20156
• JASS (2025) — Shackleton ShadowCam & M3 reflectance analysis: janss.kr
• NASA — Candidate regions for landing the next Americans on the Moon (2022 list): nasa.gov
• NASA — Update on Artemis III Moon landing regions (2024 nine-region list): nasa.gov
• NASA — Adds Mission to Artemis Lunar Program, Updates Architecture (February 2026): nasa.gov
• NASA — Fission Surface Power program overview: nasa.gov
• NASA / U.S. Department of Energy — Lunar Surface Reactor by 2030, memorandum of understanding (January 13, 2026): nasa.gov
• NASA Glenn Research Center — industry request for information, 100 kWe closed Brayton fission surface power system: nasa.gov
• Aviation Week — National Initiative for American Space Nuclear Power, OSTP directive (April 2026): aviationweek.com
• SpaceNews — China delays Chang'e-7 lunar south pole landing mission launch (August 2026): spacenews.com
• The Planetary Society — Chang'e-7 mission profile: planetary.org
• NASA — Artemis Accords, signatories and full text: nasa.gov
• U.S. Department of State — Artemis Accords signatory list: state.gov
• Outer Space Treaty (1967), full text: unoosa.org
• NASA Technical Reports Server — lunar ISRU assessments: ntrs.nasa.gov
• LROC — Shackleton crater imaging and illumination data: lroc.im-ldi.com

About this article: Written by James for thesecom.net. Published May 3, 2026; substantially revised and re-verified September 4, 2026. Research drew on NASA and Department of Energy technical publications, peer-reviewed journal sources with DOI links, and primary mission documentation. Scientific understanding of lunar resources continues to evolve as new mission data arrives; readers are encouraged to consult the primary sources above for the most current position.

Disclaimer: This article is provided for educational and informational purposes only. It summarizes publicly available research and the author's own reading of it at time of writing. Nothing in this article is intended as professional advice of any kind.