National audienceMapping spatial movement implies to rethink cartographic modelling in addition to digital data issues. This opens research pathways about the representation of spatio-temporal interactions and interrelations. Existing approaches focus either on the map background itself (data analysis issues, to answer thematic questions) or on its “shape”, i.e. on the pure graphical aspects of visualization (edge bundling) and on the semiology rules behind it. This paper aims to further examine the role of the map background in the cartography of maritime flows. This essentially geographic approach differs from existing ones in that it approximates the representation of movement by modelling the mesh of the space of flows. Three possibilit...