Air transportation plays a crucial role for connecting people around the globe and has therefore received a considerable amount of interest in complex networks research. Existing network studies often consider only flight networks. However, another view on air transportation would present passenger flow from origin to destination. Since not all cities are linked by direct flights (some passengers have to connect at intermediate airports), both networks are different. For a comprehensive study of air transportation, flight supply and passenger demand should be taken into account. In this talk, I will discuss the framework of function-structure networks and how it provides a formal multi-layer model to describe such large-scale networked systems. Function-structure networks differ from other multi-layer networks in that they do not represent different systems (like gas and power supply networks), but rather provide a coherent perspective on a single system. An investigation of the global and local properties of the world-wide air transportation systems will demonstrate the approach. Of particular interest are directness metrics that quantify the relation between function and structure in such complex systems. I will discuss how function-structure networks can adapt existing network measures to better represent real world peculiarities. The approach might help explain anomalies found in previous investigations of air transportation systems.