Reference Document on the Pakistan Crash

Of course condolences to all affected by this crash; and of course, as with nearly all such disasters, it will take a while to figure out what actually occurred. For roundup of first reactions, see this Atlantic Wire item. But as a basic reference document, it's worth being aware of the info shown below. It is a scan, from an aviator in Pakistan, of a Jeppesen approach plate* for the airport in Islamabad where the crash occurred. This is the approach -- ILS DME Runway 30 -- the plane was reportedly attempting when it crashed.


What this shows and why it matters:
- Nothing in and of itself. It's just a basic document for understanding the airport and one of its "instrument approaches," which planes would use to land in bad weather;

- It helps put in perspective one analysis in a Pakistani newspaper (also quoted in Atlantic Wire), which made it sound as if the pilots might have been doing something rash or unsafe with a "circling" landing. This analysis was:

If you are using runway 12** as it seems (the Murree road side) there is no ILS [instrument landing system]. The ILS is on runway 30 (the opposite side). So the procedure (which I never thought was safe) is you fly the ILS to 30 and then you break off and turn right and fly parallel to the Margalla hills and then turn back in and land on 12. It's not a circle but more of a race-track pattern.

You may ask why not have the aircraft turn left rather than right towards the Margalla Hills. The reason is that on the left of runway 30 is Dhamial Air Base, GHQ and so on and as far as I remember that is all so-called "Restricted" airspace. You cannot fly over it.

Final point, when you are flying parallel to the Margalla Hills, you are required to keep the airport on your left in sight. So I can visualize the captain in the left seat looking left. Maybe the [flight officer] was flying and craning his neck too. They just seem to have drifted into the Margalla Hills -- perhaps because of high winds. They lost what is called "Situational Awareness". Basically they did not know where they were.

Well, maybe. The bottom pane on the approach chart shows the "decision altitude," or "minimum descent altitude," for the airplane under different circumstances. Basically this is how close to the ground the airplane is allowed to get while relying strictly on instrument guidance. If the pilots cannot see the ground (or runway lights etc -- there are elaborate rules) by that point, they have to abort the approach and execute a "missed approach," also known as "going around," in which they climb away from the runway and then try again or head somewhere else.

In the best circumstances, with the plane headed for a "straight-in" landing in the same direction as the approach, this ILS procedure would allow them to get within 309 feet of the ground purely on instrument guidance. But for a "circle to land," approach, in which the plane would intend to land in the opposite direction from the approach, the minimum descent altitude could be as high as 842 feet above the ground. It would be illegal to go any lower than that if they couldn't see.

In a "circling" approach, the airplane would descend from the southeast -- the lower right of the chart shown above -- and then, once the pilots saw the runway, fly parallel alongside it to the other end (upper left of chart), at which point they would turn around and land in a northwest-to-southeast direction. You do this when the instrument approach guidance works only in one direction, but the winds favor a landing the other way.

On the one hand, this kind of landing is by definition more complicated than a "straight-in," with the runway right in front of you the whole time. On the other hand, the idea of flying alongside the runway in one direction ("downwind"), and then reversing course (turning to "base" and "final") to land in the opposite direction is a perfectly normal way to operate in good weather.*** And according to the rules for "circle to land," you can't attempt this kind of landing unless you have a clear view of the runway the whole time. Which makes it seem unlikely that the pilots would not have noticed if the plane were just drifting off to the side, into the hills. But for now that's just hypothesis too. For a discussion of other hypotheses, see the BBC here.

Bad weather is ultimately responsible for most of what goes wrong in aviation.**** Fog or clouds that keep pilots from seeing the runway (the Polish airline tragedy), strong or gusty winds that complicate the process of touching down, icing in below-freezing clouds (Colgan in Buffalo), severe turbulence from thunderstorms (probably the Air France flight over the Atlantic), etc. From first reports of this incident, weather appears to have played a big part, but this is all tentative. For the moment, if you read about a "circling approach" being involved, look to this chart to understand what that means.
* Jeppesen plates are proprietary and copyrighted, also, this is an out of date version of the plate -- thus my helpful photo-shopped addition of the "Not for Navigation" warnings!

** Runways have numbers according to their compass direction. Runway 12, in this case, has a compass direction of 120 degrees -- roughly southeast. The runway number is the compass direction divided by 10. But since nearly all runways can be used in both directions, the same strip of asphalt has two names, depending on which way you're heading. These opposite directions are of course separated by 180 compass degrees (half of a 360 degree circle). If you're headed southeast on the runway at Islamabad, you are on Runway 12. If you're on that exact same runway but headed the other way, the compass direction is 300 degrees (northwest), and you are on Runway 30. And if you are taxiing at the Islamabad airport, you would see markers pointing toward "Runway 12/30" -- the same strip of asphalt, with two different names.

*** This is known as "flying the pattern" and is how everyone learns to fly. Big airliners typically don't have to do this - though at Dulles airport, in particular, when the winds are from the north, I've often been on airliners that had to head far south ("downwind,") and then make two right turns ("base" and "final") before landing headed north.

**** For another time: the different genesis of commercial-airline versus private-pilot accidents, with flat-out pilot error playing a far larger part in the latter case. But there too weather makes a huge difference.