Affiliation
American Association of Variable Star Observers (AAVSO)
Sat, 07/16/2016 - 04:02
Is there an agreed-upon definition of scintillation? I originally understood it to mean any atmospheric effect that causes rapid variation in star brightness. But I have seen a categorization of such variations into three distinct groups: variation induced at low alititudes, variation induced at middle altitudes, and variation induced at high altitudes. In this categorization, only the latter effect was considered scintillation.
Hi Tom,
The accepted definition for astronomical scintillation is the rapid intensity fluctuations from astronomical objects as their light passes through the Earth's atmosphere. Most people turn to the theoretical treatment by Andy Young from the 1960's, culminating in a 3-part series (Dravins, et al. 1997PASP..109..173D); a more recent paper is by Osborn, et al. (2015MNRAS.452.1707O). I haven't seen papers where they define scintillation to be only the upper-atmospheric part; I have seen papers where they discuss atmospheric turbulence, and split it into three parts as needed for proper adaptive optics correction. Perhaps you can give some references so that I can look into it further!
Arne
I may have been thrown off-track by the item below, which divides the causes of fluctuations into the three zones I mentioned. Only the high-altitude effect is explicitly called scintillation (see "The Optics of Turbulence" section).
http://www.handprint.com/ASTRO/seeing1.html
Below is an example scintillation light curve made with my photometer in V band. Scope aperture was 10 inches, airmass was 1.2. There are 400 consecutive 100msec samples; the RMS for a flat-line fit is 0.047. Peak-to-peak swing is about 1/4 magnitude. I was surprised at the large variation.
Hello Tom
It would be interesting to repeat this test, using exposures of say 1, 10, 20, & 30 seconds, using a corresponding fainter target to keep the SNR the same, and see the effects of sintilation.
Gary
Hi Tom,
Radu Corlan's great scintillation tables at
http://astro.corlan.net/gcx/scint.txt
indicate that for 50msec exposures, your RMS should be 0.047mag, so you are in the right ballpark. There will always be larger peak-peak excursions; that is just the nature of Poisson arrival times. Stars obviously have huge visual excursions when twinkling is present too.
The trick is longer exposures to average over this effect, of course. Typically the scintillation noise is a stronger function of exposure time than aperture, but both are important. That is why it is so easy to get millimag photometry with a multi-meter telescope. The "longer exposure" can be either a single long exposure (for CCD work) or averaging over many PEP integrations. If you averaged 100 integrations of 100msec, that is roughly the same as a single 10-second exposure, in which case the RMS uncertainty from scintillation itself should be more like 3mmag.
Also, remember that the brightness of the source can be important, as the RMS deviation is due to all noise sources, not just scintillation. How bright was the object you were observing? Was this photon-counting, or an SSP-3 type system?
Arne
The device was an SSP5, which is uses a PMT, but is still a DC photometer. The star in question was about 5th magnitude in V, with B-V near 0. I attach light curves that have 2, 5, and 10 second means (joined by line segements).
Well, the only formal treatment of atmospheric turbulence I've seen is that of Roggemann, et al (Rev.Mod. Phys. 69, 437-505, 1997). In it, the perturbation of the incoming wave front is analyzed by considering the entire turbulent region. And, in fact, adaptive optics (in which field those authors are pioneers) necessarilty integrates over the entire atmospheric path (well, almost all of it - to the Na layer). Having looked at the paper you linked to it seems they are attempting to give a phenomonelogical description of seeing, and of course it is valid to do so, as long as you realize the the seeing you get is the cumulative effect of all disturbances from the exoatmosphere to your eye or focal plane.
I've had a hard time slogging through the material I've found on scintillation - the parts I can understand, that is - but my understanding is that scintillation and seeing are considered two separable phenonmena: Seeing is a spatial modulation in the image, caused by varying tilt induced in the wavefront. Scintillation is an amplitude modulation, caused by varying curvature induced in the wavefront. The inducing of curvature supposedly happens at high altitude, while tilt effects are induced at mid and low altitude.
Tom
Whatever. I strongly recommend you look at the Rogemann et al paper, using Goodman (listed in the referrences thereto) to fill in your knowledge of Fourier optics.
My own experience with this topic is heavily steeped in the field of RF (radar in particular). In that regime it is sensible to separate out the amplitude and phase modulation components because the signals are often coherent and always narrowband. That is in strong contradistinction to the situation in astronomy; I think it ludicrous to attempt to separate the phenomena when dealing with inherently broadband, incoherent sources.
Thoughts, Arne?