Sundial Accuracy

The author has dreamed about building the ultimately accurate sundial for some years. However there are a few challenges to overcome before a sundial could be consistently accurate at the level of the second.

By providing some way for observers to correct for Equation of Time, sundial makers regularly achieve +/- 1 minute accuracy in their dials when compared to UT. What are the factors that limit sundial accuracy?

Width of the sun's disk.
The sun's disk covers about .5 of a degree in the earth's sky which causes the blurred boundary line of shadows. A blurred shadow line makes it harder to read a sundial. This problem can be partially solved using either:
- A rod or point gnomon where both sides of the shadow can be seen. It is easier to judge the middle of a symmetrical shape than to see it's boundary.
- A focusing sundial which projects a focused picture of the sun onto a scale.
Errors in physical construction of the sundial.
Setting out the sundial's gnomon and dialface can be done to high accuracy, but always with some margin of error.

Fundamental problem

However the most fundamental problem for accuracy of the sundial is the:

Irregularity of the motion of the sun from year to year.
During one leap year cycle of 4 years, the key solar parameters of Equation of Time and Declination vary significantly! This is shown in the following graph.

The graph shows the difference between Equation of Time during 2002-2005 and long term average values for each day of the year. You can see that the variation is +/-22 seconds of time.

The long term average values for Equation of Time were chosen to give optimal values (i.e. minimum error) over the 21st century.

As you can see from the graph, if a sundial's equation of time correction will give 'good' results for over the long term, the sundial will have errors up to +/- 22 seconds of time during 2002-2005.

Calculate it!

The Sun API provides the sdxEOT function for calculating Equation of Time.