For years the speculative literature on lunar preservation has favored the equatorial lava tubes — wide basalt galleries beneath Mare Tranquillitatis and the Marius Hills, attractive for their natural radiation shielding and structurally stable ceilings. The Lunar Ark codex took the same starting position. After working through the actual site-selection problem in the L1-CRY, L1-PWR, and L1-COM decompositions, that position no longer holds. The rational site for a thousand-year ark is not an equatorial cave. It is the rim of Shackleton Crater at the lunar south pole.
What Changed: Water, Power, and Coverage
Three independent developments between 2022 and 2025 shifted the calculus. LROC and Diviner mapped the permanently shadowed regions (PSRs) inside south-polar craters and confirmed surface temperatures stable below 40 K — cold enough to hold water ice and volatiles for billions of years without active refrigeration. NASA's VIPER site-selection studies and the Intuitive Machines IM-2 / PRIME-1 lander (February–March 2025) returned the first direct measurements of polar ice chemistry. And in November 2025, ESA's Moonlight programme published its south-pole PNT (positioning, navigation, timing) coverage plan in coordination with NASA's LunaNet architecture — the south pole is where the comms and navigation infrastructure is going to be.
"Crater-rim peaks near Shackleton receive >85% annual illumination while their adjacent PSR floors remain below 50 K. No other lunar terrain offers continuous power and free cryogenic cold within a single rover traverse."
— Synthesized from NASA Artemis III Science Definition Team Report & LROC south-pole illumination studies
Why Shackleton's Edge for the Ark
The codex now treats the south-pole rim as the site, on four grounds the lava tubes cannot match:
- 1 Free Cryogenic Cold: PSR floors near Shackleton sit at 25–50 K year-round. Passive cooling against this sink lets the seed vault and digital substrates (see L3-CRY-VLT, L3-CRY-PSV) run at storage temperature without burning power on refrigeration.
- 2 Near-Continuous Solar: Rim peaks on the Connecting Ridge and de Gerlache massif clear 85–90% sunlit hours annually. Vertical solar arrays (L3-PWR-SOL-VERT) co-located there feed the vault below with kilometer-scale cabling — no thousand-hour battery bank, no nuclear-only dependency.
- 3 In-situ Resources: South-polar regolith contains accessible water ice (oxygen, hydrogen for propellant and life support) and the standard suite of mineable oxides. ISRU excavation nodes (L3-ISR-EXC-DRL, L3-ISR-EXC-SCP) cost an order of magnitude less mass to deliver than equatorial counterparts that must ship volatiles from Earth.
- 4 PNT & Comms Coverage: ESA Moonlight and NASA LunaNet are building their first-generation relay and navigation constellations around south-polar coverage. The mesh comms nodes (L3-COM-MESH-PNT, L3-COM-MESH-RLY) plug into infrastructure that is going to exist regardless of the Ark.
What About Shielding?
The lava tubes' one decisive advantage was radiation protection from overhead basalt. The codex's answer at the south pole is buried regolith berms over modular vault structures — 3–5 meters of compacted local regolith provides equivalent shielding (≈10–20 g/cm² overburden) without requiring a confirmed, intact, accessible cave. The trade is a few hundred tonnes of moved regolith versus a decade of cave-validation missions before any vault hardware can be staged. The codex takes the regolith.
The lava tubes are not abandoned — they remain candidate secondary sites in the L1-MMD (Mission Multi-Domain) decomposition, useful for redundancy and for any future expansion that requires the structural volume only a cave provides. But the primary Ark, as the codex now describes it, sits on the rim of Shackleton.