NavList:

   

Reply

Rodney Myrvaagnes [mailto:rodneym@XXX.XXX] said:
> This is more a problem in error prediction than absolute time.
...
> If you can estimate the error range for each of your
> inputs, you could do the calculations for the outer edges
> and get an idea of your final range of error.
>
> On Thu, 3 Aug 2000 09:03:36 -0400, Richard B. Emerson wrote:
>>
>>I want to ask a question that was posed by this incident: while
>>acknowledging that there is dilution of accuracy in advancing (or
>>retiring) LOP's as they age (that is, the older an LOP is, the less
>>confidence a navigator should have in advancing that LOP down a DR
>>track), it's been my impression that any LOP within the current
>>navigation day (i.e., noon to noon) is fair game for advancing or
>>retiring.  This implies that, with the understanding of loss of
>>accuracy with time, LOP's as much as 23 hours 59 minutes old can still
>>be used and that LOP's over 12 hours old are certainly still usable.
How far (in time) you'd be willing to advance an LOP might hinge on
whether it is oriented in a forgiving angle with respect to your
honest estimate of the error in your DR.
Try plotting the following simplified example:
 -- Pick a starting point and label it 1200 FIX.
 -- Over the next 5 hours we estimate that our course made good is
    090d and our distance made good is 25 miles. This is your
    estimated position, labeled 1700 EP.
 -- In fact, our actual course made good was 085d, distance made
    good was 25 miles. Plot this as the 1700 ACTUAL.
 -- At 1700 we get a single LOP, bearing 315d. Plot this through
    the 1700 ACTUAL.
 -- If we had an LOP bearing 010d at our 1200 FIX, we could advance
    it to our 1700 EP (25mi @ 090d), cross it with the 1700 LOP (315d)
    and the resulting RFIX is pretty close to the 1700 ACTUAL because
    the 010d LOP is nearly parallel to our actual error.
 -- If we had an LOP bearing 080d at our 1200 FIX, we could advance
    it to our 1700 EP (25mi @ 090d), cross it with the 1700 LOP (315d)
    and the resulting RFIX is farther from the 1700 ACTUAL because
    the 080d LOP is nearly perpendicular to our actual error.
Thus, how old an LOP you'd be willing to advance for an RFIX is a
function of how much error you think you have accumulated in your
EP and how sensitive the advanced LOP will be to that error. Currents
are always a problem. When the currents are minimal or are
well-understood and predictable, you may have a better estimate of the
distance made good then the course made good. A well-calibrated log
is within most sailor's reach, but estimating the net effect of
steering error and leeway is at best an art. Thus, more of your DR
error is likely to be lateral. The running fix will be more sensitive
to DR error the more the LOP being advanced parallels your track.
An advanced LOP running across your track is forgiving of lateral DR
error and sensitive to distance error, while an advanced LOP running
along your track is sensitive to lateral error and forgiving of
distance error. As Rodney noted, you use what you've got, but
understanding how the errors accumulate helps you decide how much
confidence each input deserves.
 -- Peter <peter_smith@XXX.XXX>

   

Reply