I recently flew commercial to Scotland and was tracking our progress on my iPad using Foreflight. I made a direct-to line that followed the great circle route, but noticed that we meandered off of it quite a bit.
The only explanation I can think of is that the crew was trying to take advantage of favorable tailwinds by following the jet stream. Is this assumption correct?
And if so, can an autopilot be programmed to follow a route dictated by a winds aloft forecast?
Most flights across the Atlantic don't fly direct great circle routes; they use the North Atlantic Track system, which is an array of "lanes" you might say, usually 60Nm apart, like one of those slot car racing places with 6 or 8 lanes of cars. Just about everybody that can make the trip non-stop will be in the tracks at various altitudes. Off each coast are an array of fixed entry point fixes called Oceanic Entry Points as jumping on or jumping off points.
The north/south position and the precise routing of the tracks is set up each day, with the focus on "a westbound flow departing Europe in the morning, and an eastbound flow departing North America in the evening." (ICAO Doc 007) Tracks are moved up and down the take advantage of winds or weather, especially jet streams.
Jet streams flow along air mass boundaries at the top of the troposphere and change constantly as air masses move. The tracks are set up to try to exploit or avoid them depending on whether they are favourable or not (as well as other weather systems).
can an autopilot be programmed to follow a route dictated by a winds aloft forecast?
Yes, but the question is how it is done, and these days it's better to think of it as not just the autopilot, but a Flight Management System (FMS).
In the case of the North Atlantic Track system explained by John K, the waypoints that define the route an airplane is to take are defined by latitude and longitude.
In my day (old fart-retired in 1999) the waypoint lat/long values were read to us over the radio, and we entered them into inertial navigation system system (actually 3 of them) which then controlled the autopilot—they voted so to speak.
These days there are a better ways of getting the assigned route into the FMS, and which hopefully someone familiar with them can explain.
The point is, though, that the FMS is not looking at the winds aloft forecast and deciding the route, it's simply flying a route given to it by some means that has been ordered by a ground controlling facility to take advantage of the winds.
Now, could an FMS be programmed to receive the winds aloft forecast and decide from that the best route. Probably so, but when you add in the complications of separation for all the other aircraft doing the same thing, I don't think we're there yet. In other words, automated ATC is still down the road a ways to say the least.
The North Atlantic (NAT-HLA) High Level Airspace is special use airspace where a series of optimized routes called the OTS, Organized Track System, are published every day, twice a day. This airspace is unique for having a largely unidirectional flow of traffic, and the waypoints to be followed over the Ocean are not fixed. Your assumption "following the jet stream" is part of the answer and the following 2 extracts from ICAO NAT-HLA Doc. 007 should effectively serve as a more complete answer to that question.
2.1.1 As a result of passenger demand, time zone differences and airport noise restrictions, much of the North Atlantic (NAT) air traffic contributes to two major alternating flows: a westbound flow departing Europe in the morning, and an eastbound flow departing North America in the evening [...]
2.2.1 The appropriate OACC constructs the OTS after determination of basic minimum time tracks; with due consideration of airlines' preferred routes and taking into account airspace restrictions such as danger areas and military airspace reservations. The night-time OTS is produced by Gander OACC and the day-time OTS by Shanwick OACC (Prestwick) [...]
The routes actually flown are created by the company Flight Dispatcher and provided to flight crew as an "Operational Flight Plan" or OFP in paper or digital format. This plan is sometimes called a Fuel plan because it determines the legal minimum fuel required. Most of the time it represents the most fuel efficient routing but it can also be set up for time efficient routing or routing manually strung up by the Dispatcher etc. Overflight charges (fees) are also factored into the cost optimization. For longer flights especially, a 'route search' is carried out as there maybe many route options available and an effective choice can result in much cost saving.
The Flight Planned route and other details are fed to the on board FMC (Flight management Computer) either manually by the pilots in case of a paper flight plan or, the plan can be up-linked by the Dispatch office using electronic means, and checked, accepted and activated by the pilots in the cockpit with a few button presses. Capability exists today for the pilots to accept route change suggestions by Flight Dispatch whilst enroute in real-time thanks to realtime monitoring of the flight progress and weather/wind conditions and subject to ATC clearance.
At the current state-of-the art, and with increasingly reliable datalink capabilities, the grunt work of optimal flight planning is left to a specialist Flight Dispatcher and sophisticated flight planning software. Comparatively, the on board FMC can do rudimentary optimization with respect to winds aloft as it only knows the limited wind forecast fed in by the pilot or uplinked from Flight Dispatch. The Autopilot/FMC combo can fly a given route, but it cannot create that route.
And Yes, the auto-flight system's normal mode is to couple the autopilot to the flight plan loaded into the FMC, infact this is the way most of the navigation is done in today's airplanes.
Finally, some points on GC flying -
