The good map – a benchmark generalized multi-scale map

There is a vast research literature for automating the production of topographic map series at 1:25k or 1:50k scales with map generalization. But these maps are being replaced by pan-scalar maps where map generalization is not used that much, partly due to a common belief that generalization is not necessary anymore. This paper addresses this common belief by presenting The Good Map, a benchmark pan-scalar map, mimicking the style of Google Maps, and enhanced by progressive generalization. Following new principles of progressiveness, we designed sequences of abstractions for each map layer. Roads, rivers, water areas, buildings and cities, railroads, and vegetation are generalized. As The Good Map is released as open data as a web map or as a Digital Cartographic Model, we hope that researchers will use it to foster the development of automated solutions for pan-scalar maps.