NavList:

Richard,

I cannot understand your point.

The heading of the LOP and the calculated altitude are derived on the basis of the EP (or AP depending on which method is used) so no change takes place at all if Hs varies.

The position of each LOP with respect to the EP (or AP) is determined by comparing the observed altitude and the calculated altitude.

If I use an index correction of +2’ instead of -2’, all my altitudes will be 4’ greater and I will have to move the LOPs 4’ (or 4 miles) towards the GP of the star.

This has nothing to do with the obvious consideration that 2’ of error in altitude is not 2’ of error in fix position: 2’ of error in altitude simply move the LOP 2’ (or 2 miles) towards or away from the GP of the star. In fact in this example the error doesn’t seem to affect the fix at all, because I used two couples of LOPs and the fix was calculated by finding the intersection of bisectors of opposite LOPs.

Federico

Er, "advance all the LOP's by 4'" because of the 4.0' difference between on and off arc?  Unless I've totally missed something, that just doesn't work.  All you can adjust is the Hs reading to change from a reading 2.0' the "wrong" way to 2.0' the "right way" (that is, add 2.0' to or subtract 2.0' from, as appropriate, the Hs value).  The LOP's, and the heading to the LOP from the AP in particular, cannot be moved to just accommodate the Ic error in in the Hs data. Put another way, 2.0' of error in altitude reading is not 2.0' of error in a fix's position.  Try reducing the sights with the "wrong" and "right" Ic's and you should see why moving the LOP's alone simply isn't right.  If it happens that the LOP's do move by 4.0', it's only coincidence. 

Rick Emerson
S/V One With The Wind

P.S.  I'd be much obliged if someone would send me, >off-list, please<, the original excise data. 

Federico Rossi wrote:

I agree with Jeremy that knowing the exact nature of the index error in practical navigation is crucial.

I’ve also tried to advance all the LOPs by 4’ (considering thus a -2’.0 index error instead of + 2’.0) and found that, apart from getting a tighter group of LOPs, the fix (intersection of bisectors between opposite LOPs) is almost the same, in fact bisectors tend to eliminate systematic errors like this one and are more robust from a statistical point of view than single LOPs.

Federico.