The Limits of Water Maser Kinematics: Insights from High-Mass Protostar AFGL 5142-MM1

Multi-epoch VLBI observations measure 3D water maser motions in protostellar outflows, enabling analysis of inclination and velocity. However, these analyses assume that water masers and shock surfaces within outflows are co-propagating. We compared VLBI data on maser-traced bowshocks in high-mass protostar AFGL 5142-MM1, from seven epochs of archival data from the VLBI Exploration of Radio Astrometry (VERA), obtained from April 2014 to May 2015, and our newly-conducted data from the KVN and VERA Array (KaVA), obtained in March 2016. We find an inconsistency between the expected displacement of the bowshocks and the motions of individual masers. The separation between two opposing bowshocks in AFGL 5142-MM1 was determined to be $337.17\pm0.07~\rm{mas}$ in the KaVA data, which is less than an expected value of $342.1\pm0.7~\rm{mas}$ based on extrapolation of the proper motions of individual maser features measured by VERA. Our measurements imply that the bowshock propagates at a velocity of $24\pm3~\rm{km~s^{-1}}$, while the individual masing gas clumps move at an average velocity of $55\pm5~\rm{km~s^{-1}}$, i.e. the water masers are moving in the outflow direction at double the speed at which the bowshocks are propagating. Our results emphasise that investigations of individual maser features are best approached using short-term high-cadence VLBI monitoring, while long-term monitoring on timescales comparable to the lifetimes of maser features, are better suited to tracing the overall evolution of shock surfaces. Observers should be aware that masers and shock surfaces can move relative to each other, and that this can affect the interpretation of protostellar outflows.