After making that picture of the brightest stars by their luminosity or quantity of light (as opposed to their “magnitude”), I realized that I could have arranged it more instructively in two columns like this:
The stars in the left column are all in the southern hemisphere of the sky; those on the right are the brightest stars of the north. (The figures after each star’s name are its magnitude and its light as a fraction of Sirius’s.)
Fred Schaaf had remarked in his Sky & Telescope article that it would take the combined light from Arcturus, Vega, and Capella – the three brightest north-hemisphere stars – to match that of Sirius. (We found that it would actually take more.) But those three do not even rank next after Sirius; there are two other south-hemisphere stars above them, Canopus and Alpha Centauri.
So a part of the point is that the southern hemisphere happens to be richer in very bright stars than the northern. Since our culture developed in the north, we don’t even have traditional names for some of those deep-south luminaries, Alpha and Beta Centauri and their neighbors Alpha and Beta Crucis. (Names have been concocted for them, Rigilkent, Hadar, Acrux, and Mimosa, but we can scarcely remember those and we keep referring to them by the Greek-letter designations. Many people have heard that Alpha Centauri is the nearest star to us.)
I’ve made my program add up the luminosities of the stars in terms of Sirius’s. The 12 southern stars (including Sirius) add up to 2.9 Siriuses of light, while the 11 northern stars amount to only 1.595 Siriuses. (Sirii, I suppose I should say.)
Of course the statistics depend on where we make the cut for our list of “brightest” stars, and I’ve used the convenient cutting-point of magnitude 1.6, giving a list of 23. The southern hemisphere happens to be richer in these few very bright stars, but that is just because of the chance of where the very few very brightest happen to be. If we were to include dozens more of the fainter stars, or hundreds, or thousands, their individual contributions of light would be increasingly tiny, but presumably the totals of starlight for the hemispheres would come to be about equal. The only overall inequality in amount of starlight is between the Milky Way and the rest of the sky.
Instead of columns, why not arrange these luminosity symbols on a map?
Because of the range of sizes of our symbols, some fall on top of others, especially in the bunches in the Orion region and the Crux region, both of which are in or near the Milky Way.
I thought this might be an alternative way of symbolizing stars for star maps in general, but it wouldn’t really work.
If for instance you wanted to draw a map of an area including stars down to magnitude 8 (like Neptune) and also including Sirius, you’d have to make the squares for the faintest stars at least say 1 square millimeter in size; and then the square for Sirius would have to be about 6000 square millimeters – that is, 77 millimeters (3 inches) on a side – inconveniently huge. This illustrates why we use our logarithmic scale of magnitudes instead.
I have one more comment, which probably will not surprise Guy, given my obsession with observational rarities such as simultaneous perihelic oppositions of Jupiter and Mars, and the essence of the comment is “how bright was (or will be) the brightest nighttime star in the past (or future)?”: Sirius is visual magnitude -1.46, and according to the Wikipedia entry on Eltanin, in 1.5 million years Eltanin will be the brightest nighttime star, “nearly as bright as Sirius [is now].” Another source suggests that Aldebaran was the brightest nighttime star around 300,000 years ago, at magnitude -1.4, almost as bright as Sirius. I can’t resist asking then, has anyone projected the night sky backward or forward far enough to identify a time when the brightest nighttime star was much brighter than Sirius is now? Perhaps approaching Jupiter or Venus magnitude? As awesome as such a sight would be, Fred Schaaf points out an even more remarkable coincidence in his book on bright stars. He writes that sometime around maybe 400,000 years ago, Aldebaran and Capella were very close together in the night sky, almost at our north celestial pole! A very bright double pole star, how cool would that be?
The answers to this would be in that survey of star positions at other eras that I referred to, based on the Hipparcos data. There was a Sky & Telescope article about it, and I don’t seem to have a reference to that article, so I’d be glad if anyone knows.
I love this kind of “accounting applied to astronomy” for some reason. So I couldn’t help to notice a couple of things: The first is that you write that Sirius, Canopus, and Alpha Centauri are brighter than the brightest northern hemisphere star, Arcturus, but your table data says otherwise (Arcturus -0.04, Alpha Centauri -0.01). Is it possible that your data for Alpha Centauri only shows the A component as an individual star, but if both components were considered then their cumulative brightness would be greater than Arcturus? The second thing is that the data table you used for Deneb must have had a typo because it’s showing up very small. My guess is that Deneb should be about 1.21 based on your scale?
The point that Anthony brings up about low-sample size statistics is very good I think, because in the case of stars, the brightest star in the night sky has changed over time and will continue to change. As you have pointed out in various Astronomical Calendars, Arcturus is moving through the galaxy in a near perpendicular orbit and as such appears very bright to us for a relatively short time in Earth’s history. Sirius won’t be the brightest star for very long either, and I think it will soon be eclipsed by either Aldebaran or Eltanin, both at northerly declination. And it is purely a random occurrence that Alpha Centauri, the nearest star system, appears in our southern sky.
You’re right. The combined magnitude of Alpha Centauri A (-0.01) and B (+1.33) is -0.29.
I had noticed this mistake happening in the case of Alpha Crucis, which appeared in the wrong order of brightness; I realized it was because of its being a double star, and “fudged” it in my program, but I forgot to do this for Alpha Centauri (or, for that matter, Sirius, though its faint secondary would make little difference). It would take half an hour to explain why this failure to add magnitudes for double stars happened in this small ad-hoc program (it has to do with just searching for stars by name, rather than running through a whole catalog of stars and choosing whether to plot them) and half a day to correct the program in this respect; which I shall not bother to do (now, anyway) because I don’t expect to make further use of the luminosity-rather-than-magnitude symbols.
Incidentally, if this adding of magnitudes was done, the program could not color the stars for their spectral types, since a double star would have two different types.
As for Deneb, what happened was that the program, searching for these stars by name, first found another of the several stars called Deneb: Epsilon Delphini, the “tail” of the little Dolphin constellation. I saw the too-small and wrongly colored symbol for great Deneb and corrected the mistake (by changing the name in my catalog to “Deneb Delphini”); but I now see that the correction got into one version of my picture and not the other.
These are some of the hazards of creating things to be put into a blog! By publishing something within a day or so of thinking of it, you’re exposing work almost in progress; not like making something that is to be printed months later.
Yes, over aeons the stars move about and different ones will be nearest and brightest for us, and there have been interesting calculations of that since the survey of stars by the Hipparcos satellite (1989-1993) was published.
A pattern of brightest stars that, like the Milky Way. will remain more stable is Gould’s Belt, the band around the sky formed by bright stars, inclined about 15 degrees to the plane of the Milky Way, and caused by the Sun’s being in a local spur of a spiral arm. Even that will presumably change over millions of years as the Sun rides up and down in relation to the galactic plane, like a horse in a carousel.
Geez, man! You confuse me too much. Why not just say how bright Venus is becoming? I was it at 4PM EST even without the moon as a guidepost,,,and although this month, the skies have not been that cooperative to show it,,, but I’ve got that trained eye… Maybe next month,,, but if skies are clear where you are this afternoon, Venus should be just about half-way twixt waxing moon and the sun… good luck,, and later…..
Thanks Guy, this is interesting. In any statistical analysis, random variations will have a much greater effect in a small sample than a large sample. Whether you’re talking about the locations of stars in the sky, the behavior of human beings, or the distribution of minerals on the surface of the Earth, if you only have a few data points you can’t really tell if you’re seeing a real phenomenon or just a fluke.
Regarding the absolute luminosity vs. apparent magnitude of stars, it seems good to be able to think in terms of both scales, and to understand the relationship between them. Logarithmic scales conform well to human perception.