[wwn_se]

GPS is not involved in this change really. Solar storms affect the magnetic field of earth to. Magnetic north is just a sensor going through software like everything else. Less risk... not more.

[rob74]

Er... I get it that it's about changing the "reference system" and not primarily about GPS. But the difference between "magnetic North" and "true North" depends on your position, and how are you going to get your position if not using GPS?

[jcrawfordor]

Using a wide variety of "traditional" techniques that range from dead reckoning (or "pilotage" which is dead reckoning corrected by landmarks) to politely asking a controller if they have you on radar. Airplanes operated under challenging conditions without falling out of the sky for a long time before GPS became a ubiquitous flight instrument (pretty recently, really).

[chrisseaton]

> GPS is not involved in this change really.

How do you determine true north without a GPS system?

[lisper]

How do you determine true north with a GPS system? GPS only gives you your position, not your orientation.

[chrisseaton]

> How do you determine true north with a GPS system?

  def true_bearing(magnetic_bearing, location)
    magnetic_bearing + gma(location)
  end

And that's the point - your true bearing is a function of your location.

[lisper]

Ah. Right. Duh. (I must be having an even worse day than I thought.)

[lxgr]

Wouldn't it be possible to broadcast local magnetic declination over ATIS or other automated broadcast systems?

As far as I know (and according to the article), modern navigation systems contain databases of the local magnetic declination anyway; instead of updating maps and navigational databases, we could just update these declination database instead every once in a while if I understand it correctly.

[chrisseaton]

How do you apply these things without knowing where you are?

True/magnetic deviation is localised. You need to know where you are to know what the localised value is.

[lxgr]

Wouldn't a very rough location suffice, i.e. something that could be either manually set (for general aviation and shorter distances), derived from a VOR station identifier or similar, or just estimated via dead reckoning/an INS?

In other words, if you don't even have a rough idea of where you are, what good will a magnetic heading do?

[chrisseaton]

Rough yeah - but I think it does change by integer degrees from actual map sheet to map sheet, so you could probably drift a couple of degrees without knowing it within an hour or so of flying.