Underway - Pacific Crossing Notes: A Sailor's Guide to the Coconut Milk Run - page 124

timator, good for situations where we cannot make assumptions about the statistical prop-
erties of our measurements - precisely what we need.
With this in mind, the approach we propose is straightforward:
Step 1 : Identify two bodies (stars, planets, moon) that will be visible at twilight, with their
azimuth angles separated by 45° to 135° (90°±45°). A difference in azimuths outside this
range will lead to an LOP crossing angle that is either too acute or too obtuse, which may
amplify any error in the altitude measurement when later plotting the fix.
For example, one June evening found us sailing between Fiji's Lau islands (see Figure 4,
Picking the right stars ). Looking for suitable candidates among the fifty-seven navigation-
al stars, we selected bright stars that would be visible above the horizon. This left us with
the eight stars shown in Figure 4. Of these eight, only four were not obscured by sails, rig-
ging, and gear off the stern, and only Arcturus and Antares provided a robust crossing
angle - another very good reason to focus on only two bodies.