[Aavso-photometry] how cloudy is to cloudy??

[Arne Henden]

mlfleenor at charter.net wrote:
> Being new at this I am tempted to shoot even with thick cirrus clouds and "sucker-holes" should they open up later on. Right now its clouds and the forecast says partly cloudy with fog for the a.m. 
> 
> I have read that differential photometry is well suited for less than perfect skies but I am curious how cloudy is too cloudy??
> 
I could write a few pages of text, but will refer this kind of question
to my upcoming book if you want a lot of detail.
   There are three major effects of clouds:  reduction in received flux
from the object in question; increase in the sky background, especially
for urban sites or moonlit skies; and differential extinction across the
frame due to the non-uniform nature of clouds.  The decrease in flux
can be accounted for by increasing the exposure time, though with varying
clouds you may go from properly exposed to saturated and back again.  I usually
have to turn off automation and manually expose, watching the real-time
guide star to set the exposures.  The background increase means you
generally can't work as faint, and the background may have stronger
color effects (sodium illumination, etc.) than normal.  Again, this
isn't a big problem for brighter objects.  In fact, I often reserve
photometry of bright objects until I have cirrusy skies just to decrease
the received flux, and since I am working with really bright objects,
the increased background is still negligible.
   The bigger problem for really accurate photometry is the variable
extinction.  You can open/close while a cloud is halfway across your
field, an effect that worsens the larger the field size, the shorter
the exposures, and the more structure in the clouds.  Your photometric
quality will always suffer under any kind of cloud condition (but still
far better than all-sky photometry under the same conditions).  You can
mitigate this somewhat by choosing comparison stars close to the target
star and using multiple comparisons that spatially surround the target.
   Bottom line: no reason to halt photometry under light cloud condition,
but just don't expect really high quality data.  I've worked through
5 magnitudes of clouds with decent results for a particular program.
Arne