High-latitude ionospheric convection is a useful diagnostic of solar wind-magnetosphere interactions and nightside activity in the magnetotail. For decades, the high-latitude convection pattern has been mapped using the Super Dual Auroral Radar Network (SuperDARN), a distribution of ground-based radars which are capable of measuring line-of-sight (l-o-s) ionospheric flows. From the l-o-s measurements an estimate of the global convection can be obtained. As the SuperDARN coverage is not truly global, it is necessary to constrain the maps when the map fitting is performed. The lower latitude boundary of the convection, known as the Heppner-Maynard boundary (HMB), provides one such constraint. In the standard SuperDARN fitting, the HMB location is determined directly from the data, but data gaps can make this challenging. In this study we evaluate if the HMB placement can be improved using data from the Active Magnetosphere and Planetary Electrodynamics Response Experiment (AMPERE), in particular for active time periods when the HMB moves to latitudes below 55°. We find that the boundary as defined by SuperDARN and AMPERE are not always co-located. SuperDARN performs better when the AMPERE currents are very weak (e.g. during non-active times) and AMPERE can provide a boundary when there is no SuperDARN scatter. Using three geomagnetic storm events, we show that there is agreement between the SuperDARN and AMPERE boundaries but the SuperDARN-derived convection boundary mostly lies ~3° equatorward of the AMPERE-derived boundary. We find that disagreements primarily arise due to geometrical factors and a time lag in expansions and contractions of the patterns.