Archaeoastronomy and Refraction Near the Earth’s Surface

Abstract

Empirical studies of refraction based on observations of sunset or sunrise over the ocean or other level¹ surface have indicated that refraction is too large and variable to support precise archaeoastronomical observations. On the other hand, ethnoastronomical studies of Puebloan sun-watching over elevated horizon markers have revealed observations near the solstices consistent to 1.2 days or 2–6 arcminutes in azimuth. A comparison of the different geometrical configurations of these two types of observations, of the interactions of these geometries with the strong and changing temperature gradients near the surface and with the specific local topography resolves this paradox by demonstrating that observations over a level horizon obtain near-surface refraction one to two orders of magnitude greater than refraction over elevated horizon markers.