The interaction of ice shelves with the ocean water underneath is one of the key processes for the future development of ice masses. Especially for the Antarctic ice sheet, mass loss through ice shelves is the dominant component of loss in the mass budget. Water masses entering the Filchner-Ronne Ice Shelf system are partly pre-conditioned in the coastal current of Dronning Maud Land, Antarctica. Given our very limited knowledge of true bathymetry underneath Antarctic ice shelves, especially smaller ones, it is obvious that all previous ocean models produced results which are based on limited knowledge of bathymetry and might therefore be considerably wrong. Coming along with our improved knowledge of the bathymetry, we now have the opportunity to assess the in[uence of errors and uncertainties in the bathymetry on ocean-modelling results, enabling us to predict basal melt rates more realistically than previously possible. Here we present three years of continuous ApRES (autonomous phase-sensitive radio-echo sounding) measurements underneath Ekströmisen, Antarctica, and four years of annual point measurements to determine the spatial and temporal pattern of basal melt. Results indicate that the temporal variability is very similar in years 2020 and 2021. It is strongest in month August to November, with melt rate peaks equivalent to up to 2.5 m/a and periodicity of roughly two weeks. Minimum melt rates are always higher than ~0.5 m/a September through November. From December to August, maximum melt rates are below an equivalent of 1 m/a and partly even no melting occurred or was below the detection limit, according to our analysis. The strong melting activity is accompanied by increased cryogenic noise in the same frequency bad. We embed out local results in the regional assessment of basal melt from satellite remote sensing and evaluate and discuss the overall comparability of observations as well as with available results from ocean models.