Despite the many satellite that have contributed to the exploration of the Martian system since the iconic flyby of Mariner 4 in 1965, the origin of two Martian moons, Phobos and Deimos, remains unknown. One of the most accredited theories is that they may have formed in-situ following a gigantic collision between two large planetesimals. However, the spectral features of Phobos and Deimos remain too different from spectra of the Martian soil, suggesting a different origin. Another theory is that Phobos and Deimos are passing asteroids captured by the Mars’s gravity field that may provide key evidence on the history and evolution of the early solar system. In particular, how water could have been delivered to Earth following migration and bombardment of asteroids and comets over the years. Unfortunately, the remarkably circular and equatorial orbits of the Martian moons seem to challenge this hypothesis.
 
The Martian Moons eXploration (MMX) mission, planned for launch in 2026, aims to shed new light on the origin of the Martian moons by collecting samples from the surface of Phobos and bringing them back to Earth by 2031. As part of an extended observation and characterization campaign, MMX will spend several months in the vicinity of the Martian moon ahead of the crucial sample retrieval operations. During this proximity phase, the spacecraft will be placed on quasi-satellite orbits (QSOs) around Phobos with different geometries and heights from its surface. Accurate knowledge of the spacecraft’s trajectory is therefore required in order to navigate the chaotic Phobos environment, perform and contextualize scientific observations, as well as refine the moon’s shape and gravity field ahead of surface operations. Strong perturbations govern the spacecraft’s dynamics at low altitudes owing to the strong influence of the nearby Mars and the moon’s irregular shape, whose orbit-attitude coupled motion can help uncover new information regarding its internal structure.
 
We present a covariance analysis study of MMX during its proximity phase around Phobos in order to quantify uncertainties as well as the maximum degree and order of the moon’s gravity field that can be estimated throughout the different mission phases. Building on the original work by Ciccarelli et al., 2024, our analysis exploits lidar and optical data to improve upon the knowledge of the spacecraft state. Synthetic images have been generated in Blender using the Orochi camera’s specifications and CORTO, the Celestial Object Rendering TOol, using surface parameters that have been optimized based on real heritage imagery. The study will also include radiometric tracking data from Earth, thereby investigating how the navigation performance vary with and without range and Doppler data. Results will show how the covariance envelopes of MMX with respect to the surface of Phobos can vary across the different mission phases as well as how different orbital configurations can support the geodetic investigations and scientific return of the mission.