[vsnet-chat 535] Red variables and high B-V values

[Berto Monard <lagmonar@csir.co.za>]

Re: B-V=3

Hi chatters,

As a photometrist/pyrometrist (and a lot of other ..ists) in the
fields of UV, VIS and IR, I deal on a daily basis with spectral and
'brightness' measuring matters but in completely different terms
and units than the astronomers do, although I try to ally with them
out of pure astro* interests.  And  (between us) astronomy is
more interesting than those other fields (ssst!.. don't tell them
here). 
Anyway, considering my engineering background, you must
admit I did come halfway allready!

The phenomenon of high B-V values is a straightforward
consequence of the measurement concept combined with the
actual 'Planckian' radiation spectra of blackbody (BB) targets, as
we can approximate the stars to be.
An incandescent lamp is another approximate BB, not as good as
a star but something fit for comparison here. Such a lamp usually,
if switched on at least, has a yellowish colour. Its apparent
temperature (colour temperature) is around 2850K but the actual
temperature of the filament is about 2770 K. Therefore the lamp
looks hotter than it actually is (spectral emissivity phenomenon)! 
If you were to increase the voltage supplied to the lamp, you 'll
notice the lamp gets much brighter and also more whitish in
colour. Further increases in the voltage and the lamp gets very
bright and bluish until it ends in a 'novalike' or 'cataclysmic' climax
of short duration (a couple of seconds). It does not necessarily
involve an explosion of the bulb, although the lamp will look
different afterwards: a piece of art. 
This is the interesting experiment. 
The other experiment is if you start off with the same lamp but you
decrease the voltage supply.  I advice to do this experiment first
and you can do it at home, but what happens here is not so
interesting, more Miralike if you want...  Decrease the voltage and
the lamp gets more reddish, a further voltage decrease and the
lamp gets brown-reddish and very weak and you can stare at the
filament (preferably for not too long, it's not free of radiation). It
looks as if the lamp is 'off', although it is not switched off ! At this
moment IRAS would still pick up lots of signal from this lamp,
infrared (IR) radiation and the actual filament temperature might
still be about 700 K (about 400 centigrade / Celsius). This is
paradoxal to what I wrote before, ie that the lamp looks hotter than
it is!  
Anyway, this anti-climax experiment is not unlike the behaviour of
LPVs (SRs and Miras). Red variables also radiate more at the long
wavelength side of the visible spectrum. They are around 2500 to
3500K and look red, more red than a lamp would at the same actual
temperature (again for reasons related to the spectral emissivity
and I do you a favour by not elaborating on this topic). 
In addition to their pulsations and consequent volumetric
expansions the Mira stars show colour shifts similar as in the 
lamp. This originates from temperature changes on the surface,
another consequence of the dynamics in such stars. 
When the star surface cools, we see a fainter star, partly due to
the lower bolometric (over the total spectrum) radiance (though
the effective radiating surface area might be larger) but mainly due
to its spectrum shifting towards longer wavelengths and outside
our visibility range.  Further temperature decreases will cause
dramatic drops in visual magnitude, but IRAS would only notice
small drops in its signal. At that moment one can imagine (and
certainly prove using radiation laws) the blue part (B) of the
spectrum to be extremely faint as compared to the red (R) and
even the 'white' (V) part. In magnitude terms R<V<<B.
B-V=3 (magnitudes) corresponds to a brightness ratio of about 15
between the white (photometric V) and blue (B) radiation levels. A
large B-V factor (>3) is very sensitive to further small temperature
changes and therefore it could go down to values of 4, 5..  without
IRAS noticing much change in its detection! But you will, visual
observer! 
And here is one of the origins for the large discrepancies in
magnitude estimates between individual observers and even
between the different 'photometric' systems. 
Spectral sensitivities differ from person to person, and even from
eye to eye. The effects are most noticeable near the cut-off
wavelengths of the photometric (or scotopic) function. The impact
on the accuracies of visual magnitude estimates  is therefore
minimal for whitish stars, but becomes gradually larger as the star
colours get more pronounced in red or blue. 

All the above implications apply to photometric (read: radiometric)
measurements in the different 'photometric' systems and
wavelength bands, and to photographic plates, using different
emulsions (read: of different makes and from different batches and
ages). 

The 'new' variable near M27: could it be such a case? 

In view of this it can be readily deduced that sky plates taken at
different 'wavelengths' have to be different, sometimes to an
extent that they are not recognised as signatures of the same part
of the sky.


Regards,

Berto Monard,
Pretoria 

PS: Another phenomenon occurs when we are staring at a red
star. The star seems to grow in brightness, Is this change in
brightness sensation due to a kind of 'servo-mechanism' in  our
neural network, causing a sudden and temporary shift in our
'scotopic' vision (from rods to cones maybe ...  just speculation):
it's another cause for mismatches in magnitude estimates and it's
known that staring at red stars (during scotopic vision) must be
avoided. 
I wonder if a similar effect occurs when staring at blue stars before
dark adaption: must find out..

VSNET Home Page

Return to Daisaku Nogami

---

vsnet-adm@kusastro.kyoto-u.ac.jp