Computing fuel cost of a specific route is not the same as finding an optimal route. Route finding in the jetstream is like a continuous and constantly changing traveling salesman problem in 3D.
Right. I stand corrected. It could be thought of as a vector space of wind velocities. Simple gradient decent would not yield optimal solutions in many cases.
I wonder if the jetstream is constant enough to consider it as a static field for the flight duration. Probably not. Wind likes to change.
