I am trying to calculate the geographical direction one needs to take to get from one point (Lat1/Lon1) to another (Lat/Lon2). Mathematical instructions are given at http://www.movable-type.co.uk/scripts/latlong.html.
For my purposes I would like to use the 'constant bearing' or 'rhumb line method' rather than the initial/final bearing method.
I have tried to construct the formulae as needed in Excel, along with the formulae for initial/final bearing and their average (row 27) for control purposes. One would expect that the constant bearing is always between the initial and final bearing and actually more or less close to their average.
In my speadsheet this works pretty well as long as the starting point has LAT1=0 (columns E, F and G). The further LAT1 deviates from 0 the worse the result. Why? I presume this has to do with the way Excel calculates ATAN2. Note that the difference between constant bearing and average bearing increases with increasing $dPHI. I have maximised $dPHI in cell H19, any further increase results in a NUM error.
I considered this in my row 26, but it does not resolve the issue.
Can you provide a solution? Thanks so much!