The interaction of partitioning the airspace and delaying flights in the presence of convective weather is explored to study how re-partitioning the airspace can help reduce congestion and delay. Three approaches with varying complexities are employed to compute the ground delays. In the first approach, an airspace partition of 335 high-altitude sectors that is based on clear weather day traffic is used. Routes are then created to avoid regions of convective weather. With traffic flow management, this approach establishes the baseline with per-flight delay of 8.4 minutes. In the second approach, traffic flow management is used to select routes and assign departure delays such that only the airport capacity constraints are met. This results in 6.7 minutes of average departure delay. The airspace is then partitioned with a specified capacity. It is shown that airspace-capacity-induced delay can be reduced to zero at a cost of 20 percent more sectors for the examined scenario. While the first two approaches investigate the upper and lower bounds in terms of delay and number of sectors, the third approach investigates the tradeoff between the number of sectors and the delay by re-applying the traffic flow management using the re-partitioned sectors. In this approach, the weather constraints are reflected in the sector partitions, and the delay is shared between airspace and airports. The solutions discovered by this approach are 6.9 minutes of average delay with a 312 sector configuration and 8.1 minutes of delay with a 253 sector configuration. Results show that a sector design that is tailored to the traffic and weather pattern can reduce delay while reducing the number of sectors at the same time. However, airspace partitioning can only address the delays caused by airspace congestion. Even in the presence of convective weather, the airport capacity constraint causes the majority of the delay.
The different versions of the original document can be found in:
DOIS: 10.1109/dasc.2011.6096211 10.1109/dasc.2011.6096055