[Aavso-photometry] yet more on uncertainties

[Michael Koppelman]

To me, this plot shows why it is nice to present an error bar that 
takes into account both the comp star scatter and the SNR of the 
variable star:

http://www.lolife.com/astronomy/gsc3047-176/gsc3047-176_20040715.jpg

If you imagine that the comp star was not plotted, the error bars would 
give you all the information you needed to quantitatively figure out 
that some clouds must have gone through or something. An error bar that 
was just the comp star scatter would not show this.

Sorry that I never shut up about this but it is just my enthusiasm in 
(re)discovering how damn useful error bars are. I'm not at all an 
expert at this stuff but I do find it interesting.

I've also been experimenting with combining observations from more than 
one observer and it is ultra helpful to be able to do weighted 
averages. Two observers working together can drive the errors down to 
0.004 or better, even taking into account the comp star scatter. In 
fact, in regards to the upcoming transit, if you combined observations 
from a bunch of folks, provided that uncertainties were accurately and 
consistently represented, you could easily create a very high quality 
light curve in the milli-mag range. We should try it.

Cheers,
Michael Koppelman
http://www.lolife.com/astronomy/