Fig. 3.

Hierarchical organization of the air-transportation network. (A) Global-level affinity matrix and hierarchical tree (the representation is the same used in Fig. 2). (B) Top-level modules. Each dot represents a city and different colors represent different modules. Note that the top level in the hierarchy corresponds to major geopolitical units. (C) The “Eurasian” module (which is composed of the majority of European countries, ex-Soviet Union countries, Middle-Eastern countries, India, and countries in Northern half of Africa) splits for levels ℓ = 2 into five submodules. (D) The “Near and Middle East” submodule further splits into seven submodules for ℓ = 3 (D). Note that the air-transportation network is large and very dense (Table 1), and thus the organization of the network is not at all apparent (SI Fig. 11). Remarkably, the modules our method detects show a clear agreement with geopolitical units.