It seems to me that having the sensor hanging freely like a pendulum
(pointing down) would make it measure the vertical component.

"Eric Greenwell" wrote in message
...

OK, I'm persuaded! It now appears the sensor would need to be aimed up
or down rather accurately, or the at least the angle off vertical
measured accurately. Dang - that's harder. Perhaps the inertial system
would be a reasonable way to achieve this, or maybe differential GPS
system with antennas on the nose and tail.