NavList:

Ken M wrote:

"You can see that the spread of lunars covers a full degree but his final
position was pretty close."

 

It might be interesting to find out how much error there was in the almanac data for those dates. Circa 1801, that was still a major source of error in longitudes derived from lunar distances.

 

And wrote:

"He took this latter step because he was intending to map
everywhere he travelled, but he clearly put more faith in his celestial
observations than did many of the ocean navigators that Frank has written
about."

 

Under good conditions, I think it goes without saying that lunars are more reliable when observed on land than at sea. The platform is stable. One can shoot dozens of lunars to make a set instead of a handful. You can wait for the weather to change. And so on... But something that might not be so obvious is that dead reckoning is simpler at sea since it consisted of long legs on constant headings. At least there aren't any currents on land!