Colour-colour diagram

John Greaves

2003 July 3

I experimented with colour - colour diagrams using TASS V-Ic and 2MASS J-Ks, and investigated where MisV1147 lived on this colour diagram.

I scythed off some TASS data, out of interest, with Welch-Stetson indices twixt 50 and 500. The upper cut off was because above that value it tends to be very small number of observations with one outlier, which affects the mean magnitudes in the two passbands, and not necessarily evenly, thus shifting the colours sometimes.

The lower limit is to force the chance of variability, though a somewhat lower limit could have been chosen, this was a first attempt, and I wanted some things at least to be fairly certain.

Then, I figured out the formatting style GATOR requires and formatted up a request list which I parsed through it to get 2MASS J-Ks for these objects. One arcsec matches, both datasets being good to said.

The graph shows the 1200 common objects.

It shows TASS V-Ic for the subset against 2MASS J-Ks.

A nice doglegged main sequence, with some objects well off the beaten track.

Now, these are mostly variables we are talking about, so obviously their magnitudes vary, and probably differently in different passbands, so they will vary in colour during variation too. But most variables do not drift all that much in colour during variation, and if they do, like in some EA stars, then often that event is relatively short lived upon the lightcurve. This is meant with respect to none catastrophic variables.

So, the mean colour (TASS is multi-epoch data averaged; 2MASS simultaneous measures) should be fairly representative.

There are a few stars identified on the graph. A couple of YSOs, a probable ZAND and a couple more 'more mundane' objects. I note also that some 'slow irregular variables' have turned out to be YSOs rather than LPVs on further examination.

The stars on the graph:

UX Ori YYOri or even proposed UXOR subclass
GT Ori semiregular - but is it...
GSC 0168 1700 HBH emission star and LS luminous OB star: B star for once!
DR Tau YYOri
StHA55 A possible ZAND, according to, more I know not (GSC 0128 0816)
EI Ori Lb - CGCS 1190
many CGCS 1697, TmzV296, 1FASTT 336, others? plenty caught it, none identified it

For the main body of the stars reddening won't amount to much in particular, as they are mostly variables, and probably mostly relatively nearby dwarfs. Some red giants will exist, and the ones in the Milky Way will suffer some reddening too, but most of the field giants will not suffer too badly.

Also, as you get closer to ever more infrared wavelengths, reddening reduces. The third table at shows absorption due to interstellar extinction at different passbands relative to Johnson V (data from 1989, ApJ 345, 245). As you can see, absorption is halved by Ic, nearly quartered by J, and tenthed by K.

So, all in all, the effect of reddening can be increasingly ignored as we go to the infrared. Be stars with circumstellar matter are usually checked for in the H-K and K-L plane, for example, ignoring reddening as a problem (see the table).

In practice, the reddening causes the main 'sequence' of stars in the V-Ic and J-Ks plane to be spread out a little in both axes, except for objects with obvious excess, which will lie away from this 'sequence'. The 'sequence' could be a much tighter line, though, if it wasn't for this.

However, for V-Ic ignoring reddening is not really acceptable, and it is only just acceptable for J-Ks.

Accordingly, when I plotted MisV1147 on the graph of V-Ic against J-Ks it appeared quite close to the known YSO (young stellar object) DR Tau, showing it had a significant excess in J-Ks. And despite being quite red, it is still somewhat bluer than it should be in V-Ic relative to J-Ks.

But I wanted to see where it really lay. Using the Av of 3.3 for MisV1147 (see and subsections) I was able to estimate new colours for it, using the Pejcha and Henden photometry to give an average maximum magnitude of around V 13.6, when V-Ic then averages out at about +1.3, all uncorrected for absorption/redenning.

Dereddening ('reblueing' actually) gave a value for V-Ic of around -0.13 (similar to what has been assessed beforeby Skiff from U-B, B-V and 'Q' and independently by West for V-Ic, via different but historically connected routes) and 2MASS J-Ks of around 1.18. Both being fairly sizeable reductions from +1.3 and +1.75 of the reddened V-Ic and J-Ks respectively.

The dereddened MisV1147 is also plotted on the graph followed by a subscript 0.

Then I thought I'd assess it against UX Ori, for I can assume that UX Ori is on the edges of the Orion OB1 association, and therefore use average table values for that association in terms of extinction and distance modulus to check the steps.

UX Ori is an A2ea giant star. This should have an absolute magnitude of roughly -0.2 and a V-Ic of about 0.0, as opposed to the no doubt reddened +0.24 from TASS data, which appears to be mostly about maximum when I compare it with quoted V maxima.

I find various sources pointing to E(B-V) of about +0.3 for Orion OB1, from which visual absorption, Av = 3.3 E(B-V), is 0.99, or 1.0.

The distance modulus to Orion OB1 is 8.5.

The TASS V magnitude is around 9.6, which is pretty much normal maximum.

So, we have V - Av - dist. mod. = 9.6 - 1.0 - 8.5 = 0.1. That is, the absolute magnitude of this object, should it lie at the distance of Orion OB1, and allowing for extinction/absorption, is Mv = 0.1... ...or fairly near what you would expect from an A2 type giant (A3/4 would be more likely, or the object lies slightly behind the generally averaged distance to Ori OB1, which itself has a three dimensional size which is bound to be relatively large for such a relatively nearby association. Relatively).

It seems fairly reasonable to accept the general Orion OB1 E(B-V) value of +0.3 for this star then, especially as it is not particularly involved with any of the nebulosity that criss crosses the region. A slightly higher value would fit expected Mv of -0.2 better, but for Av to be ideal it would need to be 1.3 (to get Mv -0.2 instead of +0.1), but that only leads to E(B-V) becoming +0.4, which given the uncertainty in the true distance, is a difference not worth worrying about.

So, the assumption of A_V = 1.0 is fair, which from the aforementioned tables of relative absorption leads to A_Ic of 0.48, A_J of 0.28 and A_Ks of 0.11 for this star, somewhat conveniently (where K and Ks are being treated as the same, although they aren't quite, but fortunately here the amount of absorption is quite small such that any subtle difference makes little real difference).

I've done so. UX Ori, suffixed with a zero, and is also plotted on the graph, being the dereddened estimate for that object. I say on the graph, in fact at V-Ic of -0.27, it falls off the bottom of the graph.

This is a little too blue for an A2 star, which is a bit of a problem. Possibly this may be connected to the fact that premain sequence stars usually hven't settled down onto the Main Sequence as of yet, and may be a little hotter than their Main Sequence counterparts. I'll have to read up on that.

By the way, on the point of relative absorption, notice that both MisV1147 and UX Ori are reddened more in V-Ic than in J-Ks , and that the overall reddening is less for the nearer UX Ori at the ~ 0.4 kpc of Orion OB1 than it is for MisV1147, stuck out there in Cepheus OB1 at ~ 3.3 kpc.

I personally am most interested by the point that MisV1147 and UX Ori are now a lot closer together on the graph. Their may be an 'island' zone on the graph where YSOs preferentially congregate. Given some more YSO candidates in other OB associations, come further V-Ic data, I'll have to look into it more.


TASS ( data is courtesy of Tom Droege's dataset, also see here for the same data (and newer):

2MASS is an IRSA/IPAC - University of Massachusetts Project

VizieR is part of the CDS Strasbourg,

Welch-Stetson index: AJ 105 1813 (1993)

DENIS 2nd data release is available via vizier

ASAS3 epoch photmetry is available via

Taichi Kato's file can be found at