For those who would like some stability in sunrise or sunset times, the following item that I posted on another forum some years ago might provide some interest:

At what latitudes could you be if there was a (roughly) 3-month period during which the sunrise time varies by no more than about 2 minutes? Or the same question for sunset?

Certainly not on the equator, where both times vary by about 31 minutes over the course of the year, thanks to the equation of time.

The following can be verified (for 2008) at

Sun or Moon Rise/Set Table for One Year ,

using Form B with universal time.

at 8.15 North, sunset is at 1809-1811 from 7 Feb to 15 May (99 days).
at 10.30 North, sunrise is at 0549-0551 from 29 July to 2 Nov (95 days).

at 8.45 South, sunrise is at 0602-0604 from 10 Feb to 12 May (93 days).
at 10.45 South, sunset is at 1755-1757 from 24 Jul to 27 Oct (96 days).

Of course, since the times are only given to 1-minute accuracy, the numbers involved are only approximate.

An example for this year:

In Recife, Brazil, the time of sunrise, rounded to minutes, has been in the range 5.22am – 5.24am from February 9th, and will stay that way until May 21st. Sunset however will have advanced from 5.46pm to 5.08pm by then.

The very early times (mean annual zenith about 11.20) are because Recife time is the same as for Rio de Janeiro, which at almost 45W is about 40 minutes behind Recife in sun terms.