Lunar-Vicinity Orbits

The parking-orbit question

Which Orbit Do You Park In? The Last 70,000 km of the Trip

The corridors page covers how to get to the Moon. This page covers where you stop when you arrive — and why different missions make different choices.

Every architecture that stages through a lunar orbit before landing pays three costs: the burn to enter the orbit (LOI), the station-keeping to stay in it while waiting, and the burn to descend from it. Apollo minimized descent Δv by parking in low lunar orbit. Artemis minimizes comms blackout, station-keeping, and polar dwell by parking in NRHO — and pays more to descend. Starship eliminates the parking orbit entirely. The orbit you choose reveals what you are optimizing for.

Low lunar orbit (LLO)

LLO: 100 km Circular — Apollo Heritage, Lowest Descent Cost, Worst Stability

The orbit Apollo used. The simplest geometry. The most expensive to maintain.

A circular orbit at roughly 100 km altitude has the lowest possible descent Δv to the surface — approximately 1.9 km/s for ascent and 2.1 km/s for descent (the asymmetry comes from the hovering and gravity-loss penalty on landing). Apollo flew in LLO at 110 km and descended from it. Every Soviet Luna sample-return mission used LLO. Chang’e-3, -4, -5, and -6 all staged through LLO. It is the most-flown lunar staging orbit by a wide margin. Observed

LLO has three problems that become acute for sustained operations:

Mascon instability. The Moon’s gravitational field is not uniform. Concentrations of mass (mascons) beneath the maria destabilize many otherwise circular LLO orbits. At 100 km altitude, some inclinations experience altitude changes of tens of kilometers within weeks without station-keeping burns. Apollo missions lasted days, not months — long enough to ignore mascons. A permanent station in LLO cannot.

Comms blackout. LLO passes behind the Moon every orbit — roughly 45 minutes of comms blackout per 2-hour period for a 100-km circular orbit. For crewed missions requiring continuous Earth contact, this is operationally unacceptable without a relay satellite.

No polar dwell. An equatorial LLO passes over the south pole for a few minutes per orbit. A polar LLO gives continuous polar ground track but still has mascon issues and a roughly 50% comms blackout. Neither provides the extended polar dwell time needed to support sustained surface operations at Shackleton.

Near-rectilinear halo orbit

NRHO: The Orbit Artemis Chose — And Why It Won Despite Higher Descent Δv

A 6.5-day orbit that dips to 3,500 km over the south pole, reaches 70,000 km on the far side, and costs almost nothing to maintain.

The Artemis 9:2 L2 southern NRHO has a perilune of approximately 3,366 km (minimum altitude ~1,450 km over the northern hemisphere) and an apolune of roughly 70,000 km (minimum altitude ~68,000 km over the southern hemisphere). The period is approximately 6.56 days. The 9:2 synodic resonance means 9 revolutions for every 2 synodic months — a ratio that keeps the orbit eclipse-free, eliminating the need for large batteries to survive extended solar outages. Annual station-keeping is approximately 15 mm/s — essentially zero. Observed

NASA selected NRHO over LLO, DRO, and L2 halo orbits after a systematic trade study. The NRHO won because it balances five properties that no other single orbit provides:

Near-continuous Earth communications. The NRHO never passes behind the Moon for more than a brief interval near perilune. Earth link availability exceeds 95% of the orbit — compared to roughly 50% for an equatorial LLO.

Polar dwell. The southern NRHO spends the slow, high-altitude portion of its orbit over the south pole, providing extended dwell time for surface-support comms and observation — exactly where Shackleton is.

Near-marginal stability. The NRHO is weakly unstable — stable enough to require almost no station-keeping, but unstable enough to provide useful departure manifolds for lunar-surface access and Earth-return trajectories.

Eclipse-free. The 9:2 resonance avoids geometry where the Moon blocks the Sun for extended periods. No long eclipses means smaller batteries, simpler thermal design, and continuous solar power.

Orion-compatible. Orion can perform LOI into NRHO and TEI (trans-Earth injection) from NRHO within its propellant budget. It cannot do the same for LLO — the additional LOI and plane-change burns exceed Orion’s Δv capacity. NRHO is not just NASA’s preference. It is the only staging orbit that Orion can reach and leave.

The cost is higher descent Δv. A lander departing from NRHO perilune to the south-polar surface pays roughly 2.5 km/s total (including NRHO-to-LLO transfer plus descent), compared to ~2.1 km/s from LLO. That ~400 m/s penalty translates to a larger, heavier lander — and is one reason Starship HLS is so large. But when all mission costs are summed — LOI into the staging orbit, station-keeping during loiter, descent, ascent, TEI — the NRHO total and the LLO total are comparable. NRHO pays more on descent and less on everything else. Observed

Distant retrograde orbit

DRO: The Warehouse Orbit — Stable, Distant, and Wrong for Staging

Dynamically stable for centuries. No useful departure manifold. No polar coverage. A warehouse, not a staging orbit.

A DRO loops retrograde around the Moon in the rotating Earth-Moon frame. It can be stable over an enormous range of sizes — from a few thousand to tens of thousands of kilometers from the Moon — with station-keeping approaching zero. As the libration points page covers, DRO is a warehouse: excellent for parking propellant reserves, spares, and strategic inventory that should not drift. Observed

DRO is not a staging orbit for surface access. It has no inherent polar coverage. Descent Δv is higher than from NRHO. And because it is stable, there is no manifold waiting to carry a vehicle away cheaply — every departure is a deliberate, propulsive event. NASA evaluated DRO as an Artemis staging orbit and rejected it for these reasons.

Frozen orbits

Frozen Lunar Orbits: The Relay and Navigation Orbits

Mascon-resistant at specific inclinations. Not staging orbits — infrastructure orbits.

At inclinations near 27°, 50°, 76°, and 86°, certain elliptical orbits are “frozen” — their eccentricity and argument of periapsis remain nearly constant despite mascon perturbations. These orbits are predictable over months and years without large station-keeping budgets. They are the natural home for communications relay constellations, navigation beacons, and surface-mapping vehicles.

ESA’s Moonlight navigation constellation, NASA’s LunaNet concept, and China’s Queqiao relay architecture all use frozen or near-frozen orbits. Frozen orbits serve the base from above. They are not the orbits a lander descends from. Building

Direct descent

Direct Descent: Starship’s Approach — Skip the Parking Orbit Entirely

No LOI burn. No loiter. No station-keeping. Just descend from the transfer trajectory and land.

SpaceX’s cargo Starship architecture eliminates the parking orbit. The vehicle departs LEO on a trans-lunar injection burn, coasts to the Moon, and performs a single powered descent directly from the transfer trajectory to the surface. No LOI, no loiter in NRHO or LLO, no waiting for a launch window from orbit to the landing site.

The Δv cost is roughly the same as LOI-plus-descent from LLO — about 4.0–4.5 km/s total from the transfer trajectory to touchdown — but the propellant is burned in one continuous sequence rather than split between an orbit-insertion burn and a later descent burn. Starship’s enormous propellant capacity (1,200+ tonnes at departure) makes this single-burn architecture feasible where smaller vehicles cannot carry the margin.

The trade-off: no parking orbit means no loiter capability. If the landing site has a problem — lighting, terrain hazard, comms outage — the vehicle cannot wait in orbit for conditions to improve. It either lands on this pass or performs an expensive abort. For cargo that lands permanently and becomes infrastructure, this trade is straightforward. For crew, it is not. That is why Artemis stages crew through NRHO and cargo through direct descent — two architectures optimizing for different things. Building

Comparison table

Lunar-Vicinity Orbit Comparison

Six options, six different optimization targets.

The real trade

Why “Lowest Δv to the Surface” Is the Wrong Optimization

Every orbit trades descent cost against something else. The question is which “something else” you can afford to lose.

LLO minimizes descent Δv. It loses comms, stability, and polar dwell. NRHO gives up 400 m/s of descent performance and gains near-continuous Earth link, eclipse-free power, and negligible station-keeping for years. DRO gives up everything useful for staging and gains permanent stability. Direct descent gives up the ability to wait and gains the elimination of LOI entirely.

The NASA trade study that selected NRHO for Artemis found that when all costs were summed end-to-end — TLI from LEO, LOI into staging orbit, station-keeping during loiter, descent, surface stay, ascent, loiter for TEI window, TEI, and Earth entry — the NRHO and LLO architectures produced comparable total Δv. NRHO paid more on descent and less on insertion, station-keeping, and abort trajectories. The total was roughly a wash; the operational advantages were not.

The deeper lesson applies beyond the Moon. The right staging orbit is not the one closest to the surface. It is the one whose total cost — including the cost of operating there for months while waiting — is lowest when the entire mission is priced. Apollo could ignore this because the lunar module descended within hours of LOI. Artemis cannot — Gateway was designed to loiter for months between crew visits, and the HLS may wait weeks in NRHO before descending. At those timescales, station-keeping, comms availability, and eclipse avoidance dominate the trade. Descent Δv is a line item, not the answer. Observed

Record

Provenance and Data Sources