Beta UMi "variability"

[nbrosch]

Bob, extinction close to the horizon does not behave as astronomy textbooks tell you it should do. In general, one uses the "regular" extinction correction: m(lambda)=m0(lambda)+k(lambda)*AM, where m(lambda) is the object magnitude, m0(lambda) is its magnitude outside the atmosphere, k(lambda) is the extinction coefficient at the wavelength of observation, and AM is the airmass of observation.

Textbook astronomy teaches that AM is sec(z), where z is the zenith angular distance, however when the altitude of the object is small the airmass needs to be corrected for the Earth curvature, for the change in atmospheric density with altitude, etc.

One of the useful corrections is by Hardie (1962, in Astronomical Techniques, ch. 8) and is an expansion:

AM=sec(z)-0.0018167*[sec(z)-1]-0.002875*[sec(z)-1]^2-0.0008083*[sec(z)]^3

Another useful formula is by Young & Irvine (1967, AJ 72, pp. 945-950):

AM=sec(z)*{1-0.0012*[sec(z)^2-1]}

You may also want to look at the paper dealing with low altitude observations of comets: http://cfa-www.harvard.edu/icq/ICQExtinct.html.

Finally, I wonder whether the differential refraction could play a role in your findings. Because of refraction, point objects that are not at the zenith present essentially short spectra that are perpendicular to the horizon. Obviously, the lower an object is, the 'longer' is the spectrum. The CONCAM magnitudes use a fixed aperture; this may miss part ofth light from a star when it is close tothe horizon.

Noah Brosch

[RJN]

Noah,

I split your post into its own thread. Undergraduate student Vic Muzzin has looked into this a bit, and graduate student Tilvi will likely make atmospheric extinction a chapter in his thesis on stellar variability with CONCAMs. Lior's WOLF program now maps the amount of extinction above 20 degrees with opacity maps. We were aware of some of the information you posted but not all. The info and great link you gave will be quite valuable.

Thinking again about this topic, it appears (again) that atmospheric extinction is both a blessing and a curse to the NSL project. It is a curse since we can never fully take it out, and it will cause errors that we cannot make just go away. Slight, monotonic "stellar variability" near the horizon might always remain indistinguishable from atmospheric extinction.

It is a blessing since the NSL project has a unique vantage point to characterize atmospheric extinction in real time in ways that have never been done before. The opacity maps are just the beginning, I think. Perhaps instead on INVOKING atmospheric extinction models, we can SOLVE for best fit parameters in them. For example, looking at stars of different colors near the horizon, we can perhaps compute a real time estimate of general aerosol extinction. Then, even though we have a broadband detector, the dimming of stars of different colors could allow us to solve for several extinction co-efficients as a function of wavelength.

Well, who knows. It's at least a dream.

- RJN

[nbrosch]

Bob,

"We were aware of some of the information you posted but not all. The info and great link you gave will be quite valuable."

Glad to help.

"Slight, monotonic "stellar variability" near the horizon might always remain indistinguishable from atmospheric extinction."

The solution is to look at nearby stars; it would be extremely unlikely to have stars at similar (alt, az) locations vary in phase. Some of the variation might be due to noctilucent clouds, or to the wavy brightening seen sometimes in CONCAM images; one should look for correlations with the location of the bright bands and the star being measured. Also, with sufficient measurements one might find a period. This would clinch its stellar origin.

"It is a blessing since the NSL project has a unique vantage point to characterize atmospheric extinction in real time in ways that have never been done before. The opacity maps are just the beginning, I think. Perhaps instead on INVOKING atmospheric extinction models, we can SOLVE for best fit parameters in them. For example, looking at stars of different colors near the horizon, we can perhaps compute a real time estimate of general aerosol extinction. Then, even though we have a broadband detector, the dimming of stars of different colors could allow us to solve for several extinction co-efficients as a function of wavelength.

Well, who knows. It's at least a dream."

I think that this could become reality. One of the reasons why I would like filters in the cameras discussed previously (Super-CONCAM, or CONCAM IV) is just so that the color dependence could be accounted for.

Cheers,
Noah