Re: Wrong times for sunrise/sunset?

[Bob Proulx]

Marcin Borkowski wrote:
> C-u M-x sunrise-sunset (and today's date) says:
> 
> --8<---------------cut here---------------start------------->8---
> Sat, Jan 19, 2019: Sunrise 7:54am (CET), sunset 4:13pm (CET) at Poznań,
> Poland (8:19 hrs daylight)
> --8<---------------cut here---------------end--------------->8---

I am a little further south and get 9:40 of daylight.

> Here are my settings:
> 
> --8<---------------cut here---------------start------------->8---
> (setq calendar-latitude 52.4)
> (setq calendar-longitude 16.917)
> (setq calendar-location-name "Poznań, Poland")
> --8<---------------cut here---------------end--------------->8---

How did you decide upon that longitude and latitude?

> (BTW, if anyone is ever near here, please drop me a line - we might be
> able to meet in person;-)).
> 
> And C-u M-: calendar-time-zone says (correctly) 60.
> 
> However, https://www.timeanddate.com/astronomy/poland/poznan (for today)
> says:
> 
> --8<---------------cut here---------------start------------->8---
> Sunrise Today: 07:51↑ 124° Southeast
> Sunset Today: 16:14↑ 237° Southwest
> --8<---------------cut here---------------end--------------->8---

I looked but could not see that page listing a longitude and latitude
for that location.

> I also noticed that other online services give yet other results.
> 
> Anybody knows why the difference(s)?

I do not know but I will guess.  Here are some ideas.

The longitude and latitude of the two calculations were different
enough to produce that difference.  If the two locations are not
identical then the calculations will produce a different result.

The times given were to the nearest minute.  Errors due to rounding or
truncation may cause them to be closer together or further apart in
result.

The models used to calculate sunrise and sunset may be different
between the two methods.  I didn't investigate but there are different
approximations for the non-spherical shape of the earth.  The planet
is somewhat pear shaped.

If the two methods were different then they would produce slightly
different results.  I would trust a calculation based upon the
https://en.wikipedia.org/wiki/Nautical_almanac ( now known as The
Astronomical Almanac ) for your location and altitude as
authoritative.  (It has been a while since I have done the
calculations myself however.  I would need a refresher.)  If that were
known then we would know which model was more accurate for your
location.

Do you know if timeanddate.com uses civil twilight?  Or nautical?  Or
astronomical?  Wikipedia has a good graphic for the differences.

  https://en.wikipedia.org/wiki/File:Twilight_subcategories.svg

One might wonder, what's the difference between them?  I am
undoubtedly going to describe this wrong.  But hopefully it will be
good enough to explain the concepts.

If the upper limb of the sun (the upper limb is the top of the circle
that is the sun) is below the horizon between 0 and 6 degress then it
is civil twilight.  However it is still quite bright out due to
refraction of the sun due to the atmosphere.  It is that refraction
that requires the sun to be below the horizon 6 degrees before it is
lost from visibility.  It is still too bright to see stars.  That
makes it too bright for navigation by star sights.  But the sun is
below the horizon.

If the upper limb is between 6 and 12 degrees below the horizon, a
number that as far as I have been able to determine is a practical
number from observation and usefulness and not from any intrinsic
constant of the universe, then it is dark enough that the bright stars
are visible and also the horizon.  There are 57 bright stars typically
used for celestial navigation and the brightest of those are visible
when the sun is below 6 degrees of the horizon.  And it is also bright
enough to see a clear horizon in order to observe by sextant the angle
of the star above the horizon.  This is nautical twilight and is the
period of time when celestial navigation by star sights are taken.
When one can see both the horizon and one of the bright 57
navigational stars in order to observe their altitudes.

If the upper limb is more than 18 degrees below the horizon then it is
too dark to see the horizon line.  It is not possible to observe by
sextant the altitude angle of a celestial object.  But that is when
the dim celestial objects can be observed.  Astronomers need the sun
to have set or not yet risen in order to have good "seeing".

And so we see that even a seemingly simple thing as defining sunset
depends upon what we need to know it for!  Do we need to know if the
car driver should have lights on?  Or if we need to take star sights
with a sextant?  Or if we are going to be able to see interesting
astronomical objects with a telescope?

I will guess the difference is due to some combination of the above
along with the possibilty of it being something else. :-)

Bob

P.S. Trivial: Ask random people what is the most important navigation
star and most will pick Polaris the North Star.  However that is not
one of the 57 bright stars usually used for celestial navigation.  It
is not the brightest of stars.  Also in equatorial latitudes it is
hard to see low stars through the haze.  It might not be possible to
observe it during nautical twilight.  It isn't visible in the southern
hemisphere at all.  While the North Star is by its position a useful
star it arguably is not "the most important" by a lot.  Yet it has the
best marketing department! :-)
https://en.wikipedia.org/wiki/List_of_selected_stars_for_navigation