A study analyzing the economic and safety impacts of different flight routing methods in the National Airspace System is presented. It compares filed flight routes, wind-optimal routes, and great-circle routes. Routing differences are measured by flight time, fuel burn, sector count, and number of conflicts. Wind-optimal routes exhibit on average approximately one percent less flight time and fuel burn than filed flight routes. In addition, they produce an average of 13 less conflicts in Class A airspace (18,000 feet and above). All three routing methods are qualitatively equivalent in terms of sector count distribution. These results agree with earlier studies, which investigated some combinations of these types of routes and metrics. The contribution of this paper is that it consistently compares the three routing methods across the United States using the four metrics.

Motivated by the need to select days with distinct traffic characteristics for evaluating novel air traffic management concepts and validating simulations, 517 days of delay data from the Federal Aviation Administration's Air Traffic Operations Network database are analyzed. The daily total time delay in minutes is then used as a distance metric within the K-Means algorithm to organize the 517 days into ten groups. Convergence characteristics of the K-Means algorithm and summary statistics of the groups are presented. The singlemetric K-Means algorithm is then extended to create a multiple-metric K-Means classifier. Two examples of multiple-metric classification are presented using two different sets of metrics to partition the 522 days, the original 517 days and five special days, of data into groups. Results show that this multiple-metric classifier is useful for creating sets of days that represent a variety of traffic conditions.