If an airport increases its noise charges to a certain level, it may force airlines to land somewhere else, change to direct non-stop flights, or even change the decision of choosing this airport as a hub. If we consider air routes, types of aircraft and flight frequencies being airline operators' decision variables, subjected to airlines' customer demand, this problem is a traditional airline network design problem. In an airline network, there are a set of airports (nodes) and a set of links (routes). The goal in an airline network design is to determine “the weekly frequency of flights served by different aircraft along route”.
Airlines have to consider the network design to minimize their noise cost as well as other operating costs when they fly to those highly noise sensitive airports. For example, there is a trade-off between using large aircrafts and small aircrafts. Large aircraft may generate higher noise costs per landing but can carry more passengers per flight, with fewer frequencies to a noise sensitive airport. Small aircrafts may have lower noise costs, but require more flights resulting in higher operating costs. When the penalty cost of flying to a busy airport is too high, airlines might even cancel the decision to land. How airlines should react to airport noise policies would be an interesting issue for airline operators and for airport authorities.
The externality of the environment, and its impact on both airline operating costs and network structure, has not received the attention it deserves. Therefore, Hsu and Lin (2005) formulated an airline network design model with an empirical example demonstrating the phenomena discussed above. In this network design model, the objective function minimizes air carrier operating costs, and it describes the notion of flow-economies on the link between hub airports. The model has to satisfy the transportation capacities on each link, the passenger demand on each node, and an equal number of take-off and landing operations occurring at each airport, within the network.