Air traffic congestion and management of air traffic flow is becoming
an increasingly important problem for travel in
the national airspace system (NAS).
One technique for analyzing the growth of complex systems is simulation
modeling. 
Previous models of the NAS have used a bucket approach where each sector 
in the NAS is modeled as a bucket; planes enter the bucket
when congestion is below a predefined tolerance level.
However, the NAS is partitioned
into sectors with clearly defined paths that the majority of air traffic
must negotiate. These paths, or jet routes, intersect at navigation
points or fixes. In this paper, we present a technique for modeling
jet routes and fixes for air traffic in the NAS. We use Floyd's algorithm
to compute the path from a source airport to a destination airport.
We begin by modeling normal air traffic and then extend
the model to include weather fronts and congestion.
Comparisons of our C++ simulator to previous work show
that our jet route approach is more efficient yet provides
a more detailed simulation.