Presentation #302.11 in the session Computation, Data Handling, Image Analysis — iPoster Session.
For most locations on the Earth, the Sun appears to rise and set daily, a reassuring, regular pattern often taken for granted. Assuming a reliable, high-accuracy astrometric algorithm, the differences between observed and computed sunrise times should be attributable to refraction, dip, or terrain. Between 2017 June to 2021 September, Urban carefully timed 50 sunrises from Ocean City, Maryland so that we could assess the effectiveness of current refraction models. Observing sunrises over the Atlantic Ocean provides a water horizon, which eliminates the complications of terrain. As part of her 2018 investigation of refraction at the horizon, Wilson demonstrated that the available models produced sunrise or sunset times that inadequately matched observed values even at mid-latitudes. However, preliminary analysis of these mid-latitude events shows reasonable agreement between the observed and computed times although in at least one case the resulting difference was as large as 5 minutes. Bartlett et al (2020) found larger discrepancies for events observed in the Arctic. Their results are consistent with earlier reports from polar regions that, when the Sun rises or sets at a shallow angle, variations in refraction can affect the observed time by minutes, hours, or days. A comprehensive investigation of how refraction affects observations on the horizon requires a substantial data set of phenomenon times with corresponding meteorological data collected at multiple sites over a range of latitudes. Once thoroughly analyzed, the data will suggest improved models that will lead to enhanced processing of images captured at large zenith distances and better planning for twilight activities.