Procedural Content Generation for Games is the application of computers to generate game content such as landscapes, vegetation, or maps. This paper proposes a system (Atlas Chronicle) for the generation of world maps for games that feature story -driven locations. The system uses a combination of story abstraction, graphs, physics simulation, Perlin noise, interpolation, and climate mapping to generate 2D tile-based maps given a basic story structure. To measure the fitness of the levels generated, a user study was performed using established game enjoyment metrics. The study utilized three different content generation approaches for comparison. Eighteen participants (11 Male, 7 Female), aged between 18 and 24 years old were screened based on their prior gaming experience. Results show that levels generated by Atlas Chronicle resulted in statistically similar responses as for the manually designed and static levels.