[Kstars-devel] KDE/kdeedu/kstars/kstars/skyobjects

[Daniel-Marian Baboiu]

Akarsh Simha wrote:
> On Mon, Jun 28, 2010 at 07:17:57PM +0200, Alexey Khudyakov wrote:
>   
>> SVN commit 1143779 by khudyakov:
>>
>> Add functions to calculate refraction corrections to SkyPoint.
>>
>> Old code used lookup tables and linear interpolation to calculate
>> corrections. Now code just calculate correctionusing formula.
>> Optimizations could be added if needed.
>>
>> Correctted altitude is continous which means that gap will
>> not appear.
>>
>> Reverce corrections are calculated by solving equation
>>     
>>> h' = h + deltaH(h) iteratively
>>>       
> Wow! This is pretty interesting. Where did you get a continuous
> formula? Did you fit a smooth curve to the tables?
>   

In Meeus (Astronomical Algorithms, 1988) the refraction correction is
given as
R=1/tan(h0 + 7.31/(h0+4.4))
which appears to be accurate to 0.07 arcminute for all h0; largest error
is at 12 degrees; R is in arcminute, h0 is apparent altitude in degrees;
true altitude is h=h0-R. This gives residual -0.08" at zenith, so the
formula must be corrected by adding +0.0013515. This can be further
refined by calculating a correction to R (in arcmin)
-0.06*sin(14.7 R +13)
accurate to 0.015 arcmin for the whole range, but with larger error at
zenith (0.9").

The inverse formula (calculate effect of refraction when true altitude h
is known):
R=1.02/tan(h+10.3/(h+5.11)), consistent with the first formula within 4".

Both formulas need correction for pressure and temperature.

There are more complex formulas (I have a few papers) which also take
into account humidity and CO2 content, as well as wavelength dependence,
but I have to distill them). One of these models also takes into account
the change in atmosphere structure between troposphere and stratosphere.


DMB