A bit of an interest here as I recently covered this topic in my ATPL studies but it does not actually answer the question. So reading between the lines here is my attempt at a thread stopper:
My notes say the end of evening twighlight is when the sun is 6 degs below the horiz as previously stated.
Simplistically, the duration of twighlight would be 6/(15deg/hr*cos(latitude)). Therefore as the earth is tilted 23.5 degrees then from 66.5 lat to the poles the sun either doesn't set or doesn't rise and in the tropics, the sun will move at 15deg/hr (the speed of the rotation of the earth).
I think this is a red-herring but adds up nicely: Looking at some extracts the air almanac that provides the sunset/rise and twilight times as durations. The duration of twighlight in winter at 54deg lat is 45 mins (ish) and under 30 mins (ish) in summer. It just so happens that 6/(15*cos(54))= 45mins (ish) and that 6/(15*cos(54-23))=30mins (ish). The 23 is the tilt of 54 deg lat in mid- summer. However, I think I might have expected +/- 11.75 degs lat between winter and summer rather than the full-on 23.5 in one season??
The ANO's half an hour appears to be an crude safe value for the year. The 45 mins to darkness in mid-winter may answer Fuji's question. God knows why the FAA say night is one hour after sunset though.
Now, obvisouly the velocity of the earth's surface at the tropics is quick in relation to high latitudes so hence the faster onset of darkness...I think.
Well there you have the contents of my brain. Ugly ain't it?
HWD
