Hi all,
I completed Denis Conti's excellent AAVSO Exo planet course last October and as a result I have joined AAVSO. I have managed 4 exoplanet observations which I am working on with AIJ. It's great to see a exoplanet light curve drop out.
It would seem that the selection of comparison stars and "playing with the parameters" can produce a range of possible outputs which can make comparison of mid transit times, duration of transit times and depth of transit vary. This makes comparison with other observers data "tricky" particularly when looking for or monitoring changes in these and other parameters.
Regarding comparison stars;
Considering, both the case of filtered and non-filtered observations.
Do we select by stars of a similar colour and bright enough to target object or do we select purely by brighter stars?
I expect to have further questions considering selection of fixing parameters once I settle my mind on comparison stars. I'm trying to adopt a consistant approach in my own work that remains consistant with that of others.
Many thanks,
Eric (UK)
Here is a link to a forum page where I previously discussed much of what you are inquiring about.
https://www.aavso.org/photometric-accuracyprecision-perspective
As you already recognize, consistency of how you observe and process is probably the most important element.
Per Aruda Ad Astra,
Tim Crawford
Hi Tim,
thanks for the comments and link which re-addresses the problem better than I put my question.
For asteroid/comet photometry the advice would be to try and select comparison stars similar in colour to the asteroid or comet when using filters and when un-filtered to try and select stars that match the max QE of the asteroid or comet.
When using AIJ I tend to use flux rather then magnitudes directly.
Indeed the errors introduced by each observer's resulting error more than likely lies within the overall error of multiple observers for the data point. I shall try and adopt a consistant method for myself.
Thanks,
Eric
It would be interesting to see how your selection of comparison stars affected your results for a specific transit in specific filter bands. Why don't you post some examples.
In general comp stars close to the same color are best. among other things there is less variation as airmass changes. When you model your light curve, you correct it using the OOT tails of the curve to correct for extinction changes using a linear or quadratic correction to get the OOT part of the curve as close to a horizontal straight line as possible. The less color difference between comp(s) and target the less correction required and in general the better the correction will be. Differences in the correction can shift the mid transir point slightly one way or the other due to tilt of the trend line and can affect the symmetry of the curve.
In general comp stars with about the same brightness are best. They allow you to maximize SNR unless you are exposure duration limited. By that I mean your time resolution will be too low if you increase your exposure time to increase signal to noise. If your are exposure duration limited, you can reduce random error by using comps brighter than your target. However, After a certain point brighter is not better if it makes your exposures so short that scintillation starts to become the limiting factor for photometric precision. There was a reference in the course that provided information of the size of scintillation effects for different telescope apertures and different exposure durations.
Separation between comp and target is also a factor. Differential seeing effects, differences in focus and flat fielding error and probably a host of other effects that are not included in the CCD error equation all tend to be less as you reduce separation between target and comp.
I am divided as to whether using an ensemble of a few comp stars actually gives better results than one good comp star. I think it depends on the specifics of each field. However, measuring all the comps in a field allows you to determine what combination gives you the most consistent results. and also allows you to determine if there is any micro-variability in any of the comps. Once you determine what gives the most consistent results for you stick with it for that field.
One thing to keep in mind is that the shape and depth of the transit light curve changes depending on the filter band you use due to limb darkening.
Brad Walter
I use ensemble comp stars for my exoplanet photometry simply because it beats down the noise. I suppose the best of both worlds would be to use an ensemble of comp stars of similar color to the program star. Boxcar averaging in 3-5 image groups will also lessen the noise; at the expense of some precision of the ingress and egress time details.
Donn Starkey
Hi Brad,
thanks for your comments which mirror my thoughts at the moment and I believe I will opt for similar coloured stars of similar or slightly brighter magnitude when possible. My fov is relatively small.
Together with a small group of other observer's I'm trying to construct a research programme based on Exoplanets following our success with SdB and circumbinary planets which led to publications including A&A.
Best wishes,
Eric