Entry 124 showed that LEO itself is a hostile environment. This sweep asks what happens when the keeper steps outside it — not the rock, which may be parked at an Earth-Moon Lagrange point or in lunar orbit, but the watch over that rock. The answer is that the watch barely exists yet. The corridor beyond GEO is larger, farther, dimmer, and dynamically stranger than the orbital neighborhoods we have spent sixty years learning to track.

The scale problem

The National Cislunar Science & Technology Strategy puts it in numbers: cislunar space is the volume beyond geostationary orbit under Earth and/or lunar gravitational influence, and it is more than 2,000 times larger than the entire region inside GEO, with the farthest parts more than 12 times as distant. AFRL’s primer adds that even a sphere only 12 GEO radii across is 1,728 times GEO’s volume, and some transfers reach beyond 35 GEO radii. The dynamics are not Keplerian. Halo orbits, Lissajous orbits, near-rectilinear halo orbits, distant retrograde orbits, and weak-stability-boundary transfers do not repeat on schedule and cannot be propagated with the Two-Line Elements and SGP4 that the surveillance community uses for LEO and GEO. A captured rock maneuvered into this volume becomes a three-body problem for both its thrusters and its trackers.

What can see it from Earth

Ground-based optical telescopes are the current best hope, but the hope is thin. GEODSS, Pan-STARRS, ATLAS, ZTF, the Catalina Sky Survey, and eventually the Vera C. Rubin Observatory can all detect bright enough objects. Rubin will reach ~24.5 magnitude in a single visit, which will transform faint-object discovery. ESA’s Flyeye telescope aims for a 45-square-degree field of view and ~21.5 magnitude. But a Los Alamos demonstration is sobering: CAPSTONE, a 55-pound CubeSat at 440,000 km, was detected at magnitude 19.57 — bright for a cislunar object — and Orion at 320,000 km was magnitude 15.53. Advanced Space simulations showed that a GEO-class communications satellite placed at Earth-Moon L2 would be visible to the Space Surveillance Network only about six days per month near full Moon; a small satellite would be never visible. The Sun and Moon themselves create optical blackout periods every month. Ground radar is even more constrained: Goldstone DSS-14 can track some cislunar objects, but it is primarily a DSN communications asset with limited scheduling, and small rocks are below its practical threshold. A proposed national deep-space radar could detect 1-meter objects at Earth-Moon L2 at a cost of hundreds of millions to over $3 billion; it does not exist yet.

What is being sent to watch from space

AFRL’s Cislunar Highway Patrol System, now reorganized as the Oracle family, is the closest thing to a dedicated watcher. Oracle-Mobility is a pathfinder for tracking known cislunar objects; Oracle-Prime is meant to detect and track both known and unknown objects using wide-field and narrow-field sensors, with a $72 million contract awarded to Advanced Space in 2022. NASA’s NEO Surveyor, a 50-cm infrared telescope at Sun-Earth L1, will find near-Earth asteroids down to 140 meters and will serendipitously detect some cislunar natural and artificial objects. CAPSTONE demonstrated autonomous crosslink navigation in a near-rectilinear halo orbit and could host future SSA payloads. DARPA’s LASSO program wants a small, maneuverable satellite in low lunar orbit for cislunar surveillance and resource prospecting. Concepts for Lagrange-point sentinels and Earth-Moon resonant patrol orbits exist in the literature but are not funded missions. The watcher infrastructure is a patchwork of pathfinders, not a net.

The catalog and authority gap

The 2023 NASA Conjunction Assessment Handbook states the problem bluntly: “cislunar conjunction assessment currently has no catalog of cislunar objects with independent trajectory and covariance solutions.” NASA’s MADCAP process screens Mars, Moon, and libration-point missions, but only when operators provide their own ephemerides. There is no independent catalog, no routine collision screening, and no single traffic-management authority. A 2025 study found more than 100 cataloged objects beyond the GEO graveyard belt — old science missions, derelict rocket bodies, long-lived debris — interacting dynamically with the Moon over multi-year to multi-decade timescales, but that is a backward-looking catalog of large objects, not a forward-looking traffic picture. The U.S. Space Force’s sphere of interest has expanded from GEO to ~272,000 miles and beyond, but its sensors have not.

Strategic context

The White House strategy calls for an integrated cislunar object catalog — natural and human-made, including lunar-surface objects — and public data sharing. In 2025, UN COPUOS agreed to create an Expert Group on Space Situational Awareness to improve international exchange. The Artemis Accords set bilateral principles for transparency and due regard but are not a traffic body. More than 100 lunar missions are expected in the next decade, driven by Artemis, China’s International Lunar Research Station, and commercial CLPS landers. Rendezvous and proximity operations are dual-use: they enable servicing and debris removal, but they also enable inspection, signals intelligence, and capture. Intent ambiguity is the core strategic problem. Article IX of the Outer Space Treaty — the “due regard” and “consult before harmful interference” clause — has never been formally invoked in nearly sixty years, despite jamming, blinding, ASAT tests, and mega-constellation congestion. It is the only binding language that applies, and it is unpracticed.

Keeper math

A keeper platform in LEO can be tracked by the existing catalog. A captured minimoon in cislunar space cannot, unless it cooperates. The practical minimum is: a cooperative beacon or radar corner reflector, regular ephemeris submissions with covariance to Space-Track and NASA MADCAP, and an SSA sharing agreement with USSPACECOM or a commercial conjunction-assessment provider. Even then, custody will be intermittent until AFRL Oracle, NEO Surveyor, and future cislunar sentinels are operational. For a small, low-albedo rock, the only reliable way to be seen is to announce yourself.

This changes the capture strategy. A captured body is not just a mass to maneuver; it is a responsibility to track. The keeper cannot assume someone else will watch it. The same object that is economically valuable because it is free mass is operationally dangerous because it is uncooperative mass. The corridor watch is therefore not a spectator sport for the keeper; it is a service the keeper may have to provide for itself.

Recalled

  • 2001: A Space Odyssey (Arthur C. Clarke, 1968). Clarke’s monolith waits at Earth-Moon L1 — a silent sentinel placed where the gravitational currents balance, watching the corridor before humanity knows the corridor matters. The image is apt: cislunar space is now full of objects whose trajectories are shaped by the same Lagrange points, and the missing piece is the sentinel. AFRL Oracle, NEO Surveyor, CAPSTONE, and the proposed Lagrange patrol orbits are all attempts to place Clarke’s monoliths. The novel’s warning is that the sentinel is ancient, patient, and not necessarily on our side; the sweep’s warning is more prosaic: without enough sentinels, two objects can be on collision courses for weeks before anyone notices.

What this changes

  • The keeper’s cislunar operations are logged as self-tracked by default. Existing SDA does not reliably cover cislunar space, and a small captured body would be below detection thresholds without cooperative aids.
  • Cooperative tracking becomes a design requirement: transponder, radar corner reflector, optical retroreflector, or continuous beacon, plus regular ephemeris sharing with Space-Track/MADCAP.
  • The legal backdrop is flagged as hollow: Article IX “due regard” is the only binding coordination language, and it has never been invoked. Cislunar traffic management is a normative vacuum.
  • The capture strategy gains a surveillance cost: keeping a rock in cislunar space requires either contributing to the corridor watch or paying someone who can.
  • The doctrine backlog is updated: a keeper needs either a cislunar SSA partnership or its own tracking assets — possibly a small hosted payload on a future patrol platform — before routine cislunar parking becomes responsible.