NavList:

Gary you said  Celestial navigation was the primary method of navigation for oceanic flights until the development of INS systems about 50 years ago. Even then, INS was very expensive and only available in some models of the Boeing 747 so all the DC-8s and Boeing 707s still continued to use celestial navigation through the 1980's. Our Air Force used it until about 15 years ago since they figured that the Ruskies would shut off their navaids in the event of a war, 

The Air force contemplated polar operations. Near the pole half of each year is in night allowing normal night time celestial fixes but during the other half it is daytime all the time. During these periods  a sunline is always available and using the moon and planets would allow for daytime fixes. More than half the time the moon would not be available, 1/2 the time it's declination was south and not observable from near the north pole, and even if it was in northerly declinations, part of each month it did not provide a good enough "cut" for accurate fixes so, if a planet were available, it would be used. If none of these objects were available then all they had available was running fixes, crossing advanced sunlines, and then the landfall procedure, positioning the plane on a sunline that passed through the target.

I think it’s fair to say that although celestial was used a lot by the USAF B52s and the RAF V Bombers, it was used primarily in the monitoring role.  In other words, the navigators were monitoring a navigational computer (the ‘system’) being fed from a number of inputs primarily heading from a number of different selectable sources, indicated and true airspeed, Doppler drift and groundspeed, and mapping radar when available and expedient to use.  Celestial position lines and fixes were plotted and compared with the ‘system’s’ position.  Depending upon the navigator-plotter’s feeling of how well the system was operating and the navigator-radar’s feeling for how good his celestial was, the system was ‘tweaked’ towards the celestial.  The formal procedure was known as resetting to a ‘most probable position or MPP’ for which there were laid down expected accuracies for each nav-aid.  Operating in an emergency mode without using mapping radar and Doppler drift and groundspeed were practiced in case it ever became necessary, but it wasn’t the expected method of operating.  It was invariably incorporated into navigation competitions to test the skills of the crews.  The only time I had to use just celestial over the Atlantic for real was when a hot air leak in the wing put several pieces of equipment out of action including the Doppler.  Unfortunately, the day just happened to coincide with an almost total eclipse of the Sun, which followed us west for some time at 15° longitude per hour, so it really wasn’t our day.  Fortunately, there was just enough Sun showing to get something out of it.  DaveP