Re: [guide-user] Digest Number 2936

Eduardo Vila-Echagüe Apr 3, 2016

Thanks, Bill.

Eduardo Vila Echague

Enviado desde Yahoo Mail para iPad

El domingo, abril 3, 2016, 5:05 a.m., <> escribió:

1 Message

Digest #2936
Re: Refraction formula used by Guide by "Bill Gray" feliks314159


Sat Apr 2, 2016 7:12 am (PDT) . Posted by:

"Bill Gray" feliks314159

Hi Eduardo,

Refraction is done using a method described in the _Explanatory
Supplement to the Astronomical Almanac_, pages 141-143. Details
are given in the source code at

The benefit of this is that it gives you about as accurate an
answer as you can get, by integrating the light bending as it
goes through a modelled atmosphere with corrections for
pressure, temperature, and humidity. (Which is why you can
set those quantities in Guide's 'Location' dialog. The method
can also correct for the wavelength of the light in question,
but I left that fixed at 574 nanometers.)

Truthfully, I don't know exactly how well the results match
reality. Seems to me that, especially at low altitudes, you
could get situations where the model of the atmosphere wouldn't
match the real atmosphere. But it is, to my knowledge, the best
method available.

For comparison, there are two other refraction methods, due
to Barrett and Saastamoinen, described a bit and with source code at

These are a _lot_ simpler, not requiring numerical integration,
and (as I recall) the results are rarely much different from the
integrated method.

Meeus writes of the simpler Barrett method (given in just a
few lines in the 'refraction()' function in refract.cpp) :
"Calculated in this way, the maximum error is stated to be only
0.015 arcminutes, or 0".9, for the whole range [zenith to horizon]."
So really, any of the three methods would do for most purposes.

-- Bill

On 2016-03-20 15:24, edovila@... [guide-user] wrote:
> Does anyone know how Guide computes refraction?
> Thanks
> Eduardo Vila Echague
> Santiago, Chile

To unsubscribe from this group, send an empty email to: