The current eclipse fraction uses the cylindrical shadow approximation, which treats the Sun as a point source and ignores the penumbra entirely. This overestimates full-darkness time by a small but real margin, and the error grows at higher altitudes. Cunningham (NASA, 1962) gives the conical formulation. The goal is to add it as an optional model and report the difference between the two so users can see how much the approximation costs them.
The current eclipse fraction uses the cylindrical shadow approximation, which treats the Sun as a point source and ignores the penumbra entirely. This overestimates full-darkness time by a small but real margin, and the error grows at higher altitudes. Cunningham (NASA, 1962) gives the conical formulation. The goal is to add it as an optional model and report the difference between the two so users can see how much the approximation costs them.