AI-generated image A conceptual visualization of multi-station ground tracking: three DSN complexes triangulate a spacecraft's position using Doppler shift and round-trip signal timing. 384,400 km Earth-Moon distance 20,200 km GPS satellite altitude 100M× GPS signal weakening at lunar distance ~1 m DSN position accuracy at the Moon 1.3 sec One-way Earth-Moon signal delay <10 m LunaNet target accuracy Why GPS Stops Working Before You Even Leave Earth Orbit GPS works because 31 satellites in medium Earth orbit continuously broadcast precise timing signals. Your phone, your car, a precision guided munition: all of them compute position by measuring tiny differences in arrival time from multiple satellites simultaneously. The system is elegant and it covers the entire globe. But GPS satellites orbit at roughly 20,200 km altitude. The Moon sits 384,400 km from Earth, nearly 19 times farther. Radio signal strength follows an inverse-square law: double the distance, quarter the power. At lunar distances, GPS signals are around 100 million times weaker than at Earth's surface. The signals are technically there, but they arrive pointing the wrong direction. GPS antennas broadcast downward toward Earth's surface. To catch those spillover signals at the Moon requires a sensitive receiver and favorable geometry. NASA proved this isn't entirely impossible. The GNSS Receiver Experiment (LUGRE), flown aboard Firefly Aerospace's Blue Ghost lander in early 2026, successfully detected GPS and Galileo signals at lunar distances during transit. That's a genuine milestone. But "detectable under controlled experimental conditions" is a long way from "useful for precision navigation." LUGRE data will help engineers understand what's feasible, not deploy a system tomorrow. AI-generated image A 70-meter Deep Space Network antenna can detect signals from spacecraft billions of kilometers away. At lunar distances, DSN achieves roughly 1-meter position accuracy and 0.1 mm/s velocity precision. The Deep Space Network: Earth's Eyes on the Solar System Before any spacecraft can navigate precisely at lunar distances, someone on the ground needs to know where it is. That job falls to NASA's Deep Space Network, one of the most capable radio antenna systems ever built. The DSN operates three complexes placed roughly 120 degrees apart in longitude: Goldstone in California's Mojave Desert, the Madrid complex in Spain, and Canberra in Australia. That geographic spread means at least one complex can always see any spacecraft in deep space, regardless of Earth's rotation. Each complex hosts multiple dishes, with the largest measuring 70 meters across. How DSN ranging works: A ground station transmits a signal to the spacecraft, which immediately retransmits it back. The round-trip travel time, multiplied by the speed of light and divided by two, gives distance. Simultaneously, the Doppler shift of the received signal (the same effect that changes a siren's pitch as an ambulance passes) gives radial velocity to better than 0.1 mm/s. Combined with observations from multiple stations as Earth rotates, DSN can pin down a spacecraft's position to roughly one meter at lunar distances. DSN accuracy improved substantially after the GRAIL mission in 2012. GRAIL sent twin spacecraft into tight lunar orbit, and their gravitational interactions mapped the Moon's internal mass distribution to unprecedented detail. That data revealed lunar mascons, concentrations of dense material left by ancient asteroid impacts. Mascons create local gravitational anomalies strong enough to perturb a spacecraft's orbit by hundreds of meters over days. Without a gravity map, those perturbations look like navigation errors. With GRAIL's data, engineers can model them out. The limitation of DSN navigation is that it requires Earth-based infrastructure. A spacecraft can't self-navigate using DSN; ground controllers must perform the computation and uplink corrections. With 1.3 seconds of one-way light travel time between Earth and Moon, a round-trip command exchange takes at least 2.6 seconds minimum. For a spacecraft descending toward the lunar surface at several hundred meters per second, that latency is fatal for any guidance loop that depends on Earth. DSN Complex Location Largest Dish Longitude Goldstone Mojave Desert, California, USA 70 m 243° E Madrid Robledo de Chavela, Spain 70 m 356° E Canberra Tidbinbilla, Australia 70 m 149° E How Apollo Got There: Stars, Gyroscopes, and Ground Control The Apollo program solved the navigation problem with a combination of onboard inertial measurement and ground-based tracking. The spacecraft carried an Inertial Measurement Unit (IMU), a set of precisely calibrated gyroscopes and accelerometers that tracked every rotation and acceleration from the moment of launch. Integrated over time, those measurements predicted position and velocity without any external signals at all. IMUs drift. Tiny manufacturing imperfections and thermal effects cause gyroscopes to accumulate error over hours and days. Apollo corrected for this using star trackers: optical telescopes that identified specific stars by position and brightness. A star fix would reveal any discrepancy between where the IMU thought the spacecraft was pointing and where it actually was, allowing a correction. Astronauts performed these sightings manually, using a sextant-like device built into the spacecraft. DSN provided the third leg of the system: independent position and velocity measurements from Earth, which flight controllers could compare against the onboard state. When they disagreed, engineers determined which to trust and uplinked corrections. The system worked, but it required constant human involvement, a team of specialists at Mission Control, and direct voice communication with the crew. The critical gap: Apollo had no way to automatically compare what its cameras saw on the lunar surface with any pre-stored map. Powered descent and landing relied on radar altimeters for altitude and the astronauts' own eyes for hazard avoidance. Neil Armstrong manually flew the Eagle away from a boulder field during the final 100 meters of the Apollo 11 descent. There was no computer system that could have done that automatically. That limitation defined what Apollo could do. Landings targeted broad, relatively flat areas. Precision placement, landing within meters of a specific science target or a pre-positioned supply cache, wasn't achievable. The technology simply didn't exist. Seeing the Surface: Terrain-Relative Navigation Changes Everything AI-generated image Terrain-relative navigation overlays a real-time camera feed onto pre-loaded digital elevation models. When craters and ridgelines match, the spacecraft knows exactly where it is on the surface, without any signal from Earth. The navigation revolution for lunar landing came from combining two things that didn't exist in Apollo's era: affordable high-resolution digital cameras and onboard computers fast enough to do image processing in real time. Terrain-relative navigation (TRN) works by comparing what a downward-pointing camera actually sees with a pre-stored digital elevation model (DEM) of the lunar surface. As a spacecraft descends, the system identifies distinctive features, craters, ridgelines, boulder fields, and matches them to the map. That match gives an absolute position fix independent of any ground-based system or GPS equivalent. The difference in landing accuracy is dramatic. DSN-only navigation delivers a spacecraft to within roughly 100 km of its target. With TRN, that shrinks to 100 meters or better. Some systems under development aim for single-digit meter precision, which would allow landing adjacent to pre-positioned equipment or within a specific geology target area. • Firefly Blue Ghost (Feb 2026): Landed on Mare Crisium using terrain-relative navigation. First commercial lunar lander to successfully demonstrate TRN at operational scale. • Chan