llondel,
From memory, there's something like 43dB of processing gain in GPS. As you say, an upwards looking antenna on an aeroplane is not going to necessarily see so much from a ground based jammer. In the days when I was involved, there were some cunning ideas to mitigate jamming. Direct Sequence Spread Spectrum is best jammed (in terms of total power) by a single carrier, but reasonable cunning in the signal processing can get rid of that. There were some clever correlation techniques designed (but I haven't a clue what happened to them) that meant that jamming had to be with a very GPS like signal. But it's over 15 years now since I was involved with that, having moved on to other things. I seem to remember suggestions of it being possible to get a jam/signal ratio of 60 or 70dB before things fell over.