[LEAPSECS] Equation of time

[Pierpaolo Bernardi]

On Fri, Feb 6, 2015 at 8:51 AM, Tom Van Baak <tvb at leapsecond.com> wrote:

> Most graphs of EoT appear accurate to the minute. Is it possible to predict
> the equation of time to the millisecond level? How complicated are the
> equations. What sorts of non-obvious corrections does one need to apply at
> the seconds level or milliseconds level?

The Astronomical Almanac gives a simple formula valid for 1 year,
which gives the value "to a precision of about 3 seconds".

I understand that to make better than this, a heavier computation is needed.