NavList:

I have been reading about the early history of the U.S. Coast Survey which was chartered by Thomas Jefferson in 1807 to create accurate charts of the coast, rivers and near hydrographic features of the United States.

To determine the latitude of key land survey points and coastal features, a transit or zenith telescope was used. Longitude, a function of time, made use of transits and the telegraph but that’s another story.

At the time of the survey, refraction tables were not believed to be accurate enough for precise bench mark work. A work-around was devised to minimalize the effects of refraction when making celestial observations and determining the latitude. The technique was as follows. :

- two stars are selected, they had to relatively high in the sky at meridian passage (it was known that the high stars suffered less from atmospheric refraction).
- one star must pass north of the observer’s zenith, the other south.
- both stars to be observed had to culminate within short time of each other (to equlize the effects of atmospheric distorition in both observations)
- find their zenith distances (z and z’) and consult the almanac for their declinations (d and d’)
- Latitude = 1/2(d+d') + 1/2(z-z ')

Method was known as Talcott-Horrebow Method and claimed results accurate to 0.1 second of arc, or ten feet.

I don't know if this level of accuracy is achievable but the method is quite interesting whereas one observation's refraction cancels the other.

Ed

http://adsabs.harvard.edu/full/1911AN....187..337S
https://searchworks.stanford.edu/view/2321695