In regards to other posts that mention built in terrain models, this is true, and really this is the E in EGPWS. It uses a Geometric Altitude, based on GPS altitude, coupled with a terrain model lookup feature, that computes a pseudo-barometric altitude. This is designed to reduce or eliminate altitude errors resulting from temperature extremes, non-standard pressure altitude conditions, and altimeter setting mistakes.
The method for the 'lookup' feature varies. Some systems use a grid lookup to the model (typical), while others use the lat/long location.
As far as how 'extensive' the terrain model is, I have seen little other than the WGS84 ellipsoid, with added/subtracted known issues...the FMS database is relatively limited in data size, (as in very little storage capability)
Some systems correlate with procedure leg. As an example, if you are on the intermediate leg, and have less that a 500' ROC, it will alert...
Other systems correlate with distance from the runway, for example tapered from threshold to a 400' ROC from 5nm to 12nm, tapering to 7000' ROC at 12nm. (from this, one can see that if you have a FAF at 5nm, one system will have a ROC at 500, while the other 400, thus different alert potential)
The look ahead includes the precipitous terrain feature, which uses the radalt input. This is why when you have terrain that suddenly rises, especially on final, the system extrapolates that as a continuous rise and will alert.
EDIT: catia.. I am not seeing how a non-gps A320 could have EGWPS...Typical GPS systems are connected to the IRU, not to the FMS. The incident you mention, with lack of response from the GPWS system, it seems the forward radar look ahead would have seen the obstacle...
Last edited by underfire; 1st Jul 2014 at 09:51.