Brightness estimation puzzle

[Demonperformer]

I have started examining the BD catalogue and have hit upon a bit of a puzzle: When comparing the given magnitudes with modern estimates, they seem to be a surpising amount off.

So far I have only examined zones +89 & +88 (a total of 181 stars), so the sample size is small. And when I remove all the given magnitudes of 9.5 (which was used for anything fainter than 9.4) I was left with 87 stars.

Of these 87, 8 (9%), have values that match the modern figure given in the SIMBAD database. A further 8 have values that are slightly fainter than the SIMBAD figure (within 0.29mag). 3 have values which are significantly fainter than SIMBAD (by 0.53, 0.68 & 1.03 mags). The remaining 71 (81%) have values that are brighter than the modern figure, 13 are slightly brighter (within 0.29mag) and the other 58 (67%) are quoted as being significantly brighter (0.31 to 1.07 mags) than SIMBAD.

Now I haven't done a lot in the way of estimating brightness with variables, but what I have read suggests that "with practice" it is "not difficult to estimate the brightness to 0.1 magnitude". If this is the case, and bearing in mind that these estimates were made by professionals, it seems surprising to me that so many of these estimates are off by so much, with a strong bias towards estimating brighter than the correct value.

I would be interested in any insights into why this should be so - particularly from any members who are experienced at making such estimates on variables.

Thanks.

[Astrobits]

Unfortunately I see too many oportunities for differences to be sure whether any one is prevalent. As the BD was prepared over 150 years ago I don't know if the observation methodology is recorded so we could check the data obtained then.

As the BD was prepared in the pre-photographic age it must have been visual only. It is certainly well known that individual observers of variable stars have "personal biases" which must be accounted for. Just look at a plot of many observations of any particular object. 10 people will NOT give you 10 answers all the same. In particular red stars appear brighter the longer you look at them and are a known problem for variable observers to be aware of.

Many stars are variables. There was no organisation like the AAVSO existing then to provide a list of reference stars and even today some of the current list of reference stars turn out to be variables that have escaped detection. Perhaps some of those stars that you identified as significantly different would bear longer term scrutiny. Have you checked them against the AAVSO records?

Bonn is at 37 deg north so the stars you have been looking at would never have risen more than 39 deg above the horizon from there. Atmospheric extinction would have been significant and probably variable, possibly varying with the colour of the star.

Just some immediate thoughts.

Nigel

[Demonperformer]

Thanks for your thoughts, Nigel.

I am noting those stars that are listed as variables (if listed on SIMBAD, but that includes the AAVSO list, also CSV & NSV).

I had not considered the atmospheric extinction angle. According to Norton, that altitude reduces brightness by between 0.1 & 0.2 magnitudes. But as most of the errors are towards thinking they were brighter than actually ... would he have been "over-compensating", I wonder?

I wasn't aware of this "personal bias" problem [my inexperience shining through - like the neighbour's security light!]. In my naivity, I would have thought that for any two observers, comparing a star at 9.2 against a 9.1 & 9.3 star would come up with 9.2.

But, maybe, this is getting at the root of the problem. Can I apply 21st century "we have millions of stars accurately measured to 3dp" thinking to this catalogue?

My understanding is that this was an early attempt to produce such a catalogue, and getting the brightnesses to 0.1 mags was a big step forward. As such, although the brighter stars were probably well-sorted, by the time he got down to 9th & 10th mag stars, he was probably having to more-or-less make it up as he went along. This line of thinking would imply that the brighter the stars are, the less inaccurate his estimate. I will have a look at that and see whether that fits.

It would be rather like a modern observer pointing a scope at a random part of the sky, looking through the eyepiece and trying to list the magnitudes of all the stars seen. Such an observer, if experienced, may be able to have a good stab at some of them, but many of the fainter ones would probably be a fair way off. Enter "personal bias" into the equation and voilà - maybe.

Now there's an experiment to ponder ...

Thanks.

[dph1nm]

I would be remarkably surprised if the old eye-balled magnitudes were anything like on the same magnitude system as modern CCD measurements taken through well-defined filters, unless someone has updated them. It just isn't likely.

NigelM

[neural]

If one was visual and the other photographic, isn't there an issue of spectral sensitivity? For example, are the most discrepant stars very blue or very red? I'd expect K/M stars to be brighter photographically than visually ... depending what filters they used etc.

[Demonperformer]

Yes, there is a significant difference between visual and photographic magnitudes of a lot of stars, but in this case, the catalogue was made visually and I am comparing them to the Vmag values from SIMBAD.

[Astrobits]

The Vmag in SIMBAD are almost certainly ccd + Visual filter and only approximate typical visual responses. ( In the AAVSO listings  "V" refers to ccd with a "visual" filter and visual is noted as "Vis". )

We do know that  vision responses are variable between individuals, a fair proportion of people ( mainly  males ) have some degree of colour blindness while some females are reported ( I am going om my memory here ) to have TWO red colour receptors  ( as opposed to the normal ONE ) and are particularly responsive to that end of the spectrum. I would have been surprised if the BD was totally consistent with modern values, but, as I mentioned previously, it might be an interesting programme to follow some of the most deviant ones on a long term basis.

Nigel

[Demonperformer]

Yes, Nigel, you are right.  SIMBAD (mostly) uses the figures from the Tycho-2 catalogue, which used the appropriate filter. Would we expect such a combination to be different from any one individual's visual appraisal to be relational (linearly? logarithmically? not sure).

I am attaching a list of the 13 stars which are listed in BD (+89 & +88) as brighter than 9.3 and have >0.5mag difference between that figure and SIMBAD in case (1) it suggests any link between them that might shed more light on the situation, and/or (2) you (or anyone else) wish to keep an eye on them in case they are doing anything interesting.

I can add further examples as I get further through BD, if that would be of interest.

Thanks

#sgl1.txt

[Astrobits]

Most of those magnitudes are below my current limit with 50mm binoculars ( telescope awaiting my building of an observatory ) but they could be ideal for DSLR Photometry ( which I don't do---yet ).  However, over long periods of time, starting now, I think that those all-sky surveys that will become  operational over the next few years will be able to track any changes far better than individuals with DSLR's.

I suspect that the reasons for most of the discrepancies lie in the past rather that the present.

Nigel