During the winter months, the constellation Orion, seen from the northern hemisphere of the planet, offers a magnificent spectacle, able to fascinate even those who raise their eyes towards the starry sky only distractedly, maybe once in a while. One of the most iconic elements of the constellation is the Belt, known since ancient times, formed by three very bright stars, well recognizable in a row next to each other. Their description has been handed down to us from the most diverse and distant cultural traditions.
In the West, the three stars of Orion’s Belt, as well as most of the brightest stars in the sky, have names of Arab origin. The easternmost of the three is Alnitak, a blue supergiant that is part of a triple system. The westernmost is Mintaka, which we see with the naked eye as a single star but is actually a multiple system consisting of at least six stars. The central star of the Belt, also a blue supergiant, is Alnilam, whose name derives from an Arabic word that means “row of pearls” and evidently refers to the asterism rather than to the single star.
The pearls of the Belt
These three bright gems of the night sky are most often mentioned in the scientific literature with the names attributed to them in 1603 by Johann Bayer’s Star Atlas, the Uranometria — Alnitak is Zeta Orionis (ζ Ori), Mintaka is Delta Orionis (δ Ori), Alnilam is Epsilon Orionis (ε Ori). Their apparent visual magnitudes are, respectively, 1.79, 2.41 and 1.69, which makes Alnilam the brightest star of Orion’s Belt  and, in general, the 29th brightest star in the sky.
The three stars of the Belt, together with many others found in their surroundings, are part of a rich star-forming region, in which there are open clusters made up of massive and bright young stars, molecular clouds that host legions of newborn stars, emission nebulae, reflection nebulae, and a large population of small- and medium-mass stars, many of which have only recently formed.
This vast set of stars belongs to the so-called Orion Molecular Cloud Complex and consists of a stellar population of relatively similar age and origin, known as Orion OB1 Association. The abbreviation OB refers to the fact that many of the constellation’s brightest, hottest, and most massive blue stars belonging to the Association are classified with spectral types O or B.
In 1964, the Dutch astronomer Adriaan Blaauw divided the Association OB1 into four subgroups, distinguished by the age of its stars. The idea was that the Orion complex had produced in the last 8–12 million years a series of subsequent stellar formation episodes triggered by feedback mechanisms. In such a scenario, the Association OB1a is the oldest element of an interconnected system, in which the Association OB1d — inside which is M42, the famous Orion Nebula — is instead the youngest element, full of still forming stars. The three stars of the Belt, together with the multiple system of Sigma Orionis, are part of the Association OB1b, which contains stars with an estimated age of 4–6 million years.
These subdivisions are obviously loose and very general. At an average distance of over 1,000 light-years from Earth, it is challenging to assess the three-dimensional structure of these stellar associations and to determine precisely which objects belong to them and which do not. It is also difficult to assess the age of individual light sources accurately. According to a 2017 study, it seems, for example, that the easternmost part of Orion OB1b, that is, the one that contains Alnitak, the Flame Nebula, and the Horsehead Nebula, has a stellar population that is on average younger than the rest of the association.
Whatever the exact age of this subgroup, there is no doubt that its brightest stars are all a few million years old at most. It makes them very young compared to the Sun, which landed on the main sequence over 4.5 billion years ago, but not very young compared to their true life expectancy. The brightest stars, in fact, are generally also the most massive, and this means that they burn their nuclear fuel reserve much faster than the Sun.
Stellar parameters of Alnilam
Alnilam is no exception to this universal rule. According to a paper published in 1974 by the Dutch astrophysicist Henny Lamers, the central star of Orion’s Belt is 3 or 4 million years old. Still, it is to be considered already a very evolved star, which has ended its reserve of nuclear hydrogen and is now burning the hydrogen contained in a shell outside the stellar core.
From the point of view of stellar classification, Alnilam belongs to the class B0 Ia. The assignment of the blue supergiant in Orion to this class, which includes the brightest B-type stars, dates back to 1953, when a revised version of the stellar spectra atlas of the Morgan-Keenan classification system was published, the first version of which was ten years old. The spectrum of Epsilon Orionis, that is, Alnilam, was adopted in the 1953 atlas as the reference standard for class B0 Ia stars . Apart from minor additions, the standard defined then is still in force, so that the central star of the Belt can rightly be considered the archetype of the brightest B-type supergiants.
The main physical parameters of Alnilam reported in the 1974 study by Lamers are as follows:
- effective temperature: 28,800 K ± 2,000 K;
- radius: 31.3 solar radii with an uncertainty of +8.2 and −2.3 solar radii;
- mass : 35 solar masses with an uncertainty of +25 and −8 solar masses;
- surface gravity: 10 m/s² ;
- bolometric magnitude : −9.6;
- total brightness: 600,000 times the solar luminosity .
The last value is enough to give us an immediate awareness of what incredible power emanates from the main star of Orion’s Belt. With a brightness 600,000 times higher than that of the Sun, a few seconds of exposure of the Earth to Alnilam’s radiation would be enough to reduce our planet to a glowing desert. If the blue supergiant were, in fact, in place of the Sun, it would hit our atmosphere with an energy equal to 817 megawatts per square meter; a hell of radiation against which there would be no defense.
If Alnilam were in the place of the Sun, it would hit our atmosphere with an energy of 817 megawatts per square meter.
But where did the measurements reported by Lamers come from? The effective temperature was obtained from the analysis of the spectrum of the star in visible light, corrected based on the distribution of the energy emitted in the ultraviolet. Gravity was also derived from the visible spectrum. The stellar radius measure was instead an extrapolation based on two data, only one of which from direct observation: the angular diameter of Alnilam, resulting in 0.69 thousandths of an arc second at the wavelength of 4.430 Å .
But how far is it really?
The other measure necessary for calculating Alnilam’s physical radius was obviously its distance from Earth, but in this case, there were no direct measurements available. Before Lamers, several authors had calculated Orion OB1 Association’s mean distance from Earth, using photometric studies that took into account, among other things, the excess of color due to the extinction along our line of sight towards Orion. These studies compared the apparent magnitude of a series of stars with their calculated absolute magnitude, obtaining the so-called distance modulus, a simple mathematical formula that allows deriving the desired distance, but with a certain amount of approximation depending on the combination of various factors.
Lamers, based on the values reported by previous authors, adopted a distance modulus d = 8.1 magnitudes, with an uncertainty of +0.5 and −0.1 magnitudes. It resulted in a distance of 420 parsecs (1,370 light-years) for Alnilam with an error of 110 parsecs more and 20 parsecs less. The considerable uncertainty in calculating the distance spread to all the measures deriving from it: radius, stellar mass, and brightness.
To better tighten Alnilam’s stellar parameters, a more precise distance measurement than that available at the time of Lamers’ study was needed. Unfortunately, it was not easy to obtain. The best way to estimate the distance of stars, at least the closest, is to calculate their parallax angle. But, in the case of a bright and distant supergiant like Alnilam, this method was not able to provide reliable results, at least not using telescopes located on Earth.
The parallax of Alnilam was finally measured from space by the Hipparcos astrometric satellite, during a mission that lasted from 1989 to 1993. But to get the official data, we had to wait until June 1997, when ESA finally published the catalog of parallax angles of 120,000 stars measured by Hipparcos. Alnilam’s parallax was then 2.43 thousandths of an arc second, a value that resulted in a distance of just under 412 parsecs, surprisingly close to the 420 parsecs obtained from the distance modulus adopted by Lamers twenty-three years earlier. But once again, the margin of error was very high — 0.91 thousandths of an arc second. It meant that Alnilam, based on the data provided by Hipparcos, could be at any distance between 300 and 658 parsecs, which made the calculation of the radius, mass, and brightness still somewhat uncertain.
Ten years later, in 2007, an astronomer from the University of Cambridge, Floor van Leeuwen, published a new version of the Hipparcos catalog, which, according to the author, ensured four times higher precision in calculating the parallax angles of the more brilliant stars. The parallax of Alnilam in the new version of the catalog appeared to be 1.65 ± 0.45 thousandths of an arc second. It corresponded to a distance of 606 parsecs (1,977 light-years) with an uncertainty of 227 more parsecs and 130 fewer parsecs.
Therefore, with the new version of the Hipparcos catalog, Alnilam was at least 600 light-years farther than the distance obtained from the parallax angle reported in the previous version of the catalog.
If the new distance was really more reliable than the previous one, it meant that Alnilam is even larger, more massive and brighter than what appeared from the already impressive data provided by Lamers in his study of 1974.
An update of Alnilam’s stellar parameters based on the new distance derived from Hipparcos’ data appears in a study by Raul E. Puebla and other authors, published at the end of 2015 in the Monthly Notices of the Royal Astronomical Society. In this study, there is a table, reproduced below, which reports all the parameters twice. In the first row, the values are calculated based on the distance of 412 parsecs obtained from the first Hipparcos catalog, while in the second row, they are calculated based on the distance of 606 parsecs, derived from the updated version of the catalog. As it is easy to notice, the only values that change are those related to the radius, mass, and brightness.
When taking into account the greater distance, Alnilam’s parameters becomes truly remarkable. The radius rises to 42 solar radii, that is, 29.22 million km — the equivalent of 4,581 Earth radii. The mass increases to 64.5 solar masses , a value that corresponds to 21.5 million times the mass of the Earth!
Finally, with the new distance, the brightness stands at the incredible value of 832,000 solar luminosities — the equivalent of an energy output of 3.2 × 10³² watts. But, if we introduce in the calculation also the margin of uncertainty present in the new measure of the distance derived by Hipparcos’ data, then the maximum value we obtain is more than double, i.e., 1,738,000 solar luminosities! This would make Alnilam one of the intrinsically brightest stars in the whole Milky Way.
However, as we have seen so far, the data in our possession are rather uncertain, so we cannot know how far away this extraordinary blue supergiant really is. Considering also the distance modulus proposed in the past by various authors, the actual distance of Alnilam is likely to fall between 410 and 600 parsecs. It results in an absolute brightness between 600,000 and 800,000 solar luminosities, more than enough to make it one of the brightest stars of the Milky Way.
But, if this is the case, it is natural to ask why Alnilam occupies only the 29th place in the list of the brightest stars observable from Earth. Why, for example, does Deneb, that is at the same time more distant and intrinsically less luminous than Alnilam, appear to us more brilliant, albeit slightly, seen with the naked eye?
To understand why, one must take into account the fact that the brightness perceived by the human eye depends on various factors. First, the star’s distance and intrinsic brightness are important. But other elements also matter, such as the extinction caused by interstellar dust interposed along the line of sight towards the terrestrial observer, or the spectral distribution of the energy emitted by the star. In the case of Alnilam, the last factor is particularly important. With a photospheric temperature of around 28,000 K, Orion’s Belt’s central star is about three times hotter than Deneb. It means that a large part of its radiation is emitted in the ultraviolet region, which does not generate a luminosity perceptible to the human eye. On the contrary, Deneb has an ultraviolet emission lower than that of Alnilam, but slightly higher in the narrow band of the electromagnetic spectrum that corresponds to visible light ; this is why the Swan’s supergiant appears brighter to us than Alnilam.
Mass loss, rotation, variability, disintegration
To complete the description of the blue supergiant, a few mentions of its mass loss and rotation speed cannot be missing.
It has long been known that hot and massive stars such as Alnilam face significant mass losses through the emission of a dense and powerful stellar wind. The radiation pressure of the star powers the mechanism that produces this dispersion of matter. Highly energetic photons that emerge at the stellar surface are absorbed by ions of various metallic species present in its atmosphere, transferring moment to them. This results in an accelerating force directed outwards.
Many observations have been made over the years to determine how fast Alnilam is losing mass through its stellar wind. According to a 1986 study, the estimated rate is 3.1 × 10⁻⁶ M⊙/a, i.e., just over three millionths of a solar mass per year. It may seem little, but it is a very high rate when compared to that of the Sun, which is eight orders of magnitude lower . In a study by the aforementioned Henny Lamers and another author, published in 1993, the mass loss of Alnilam was estimated through emissions in radio waves , and the value found, 4 × 10⁻⁶ M⊙/a, was in good agreement with the estimate provided by the 1986 study. A similar research carried out in the infrared region by the Indian astronomer B.S. Shylaja led in 1994 to an estimate of Alnilam’s mass loss once again in line with the previous ones: 3.1 × 10⁻⁶ M⊙/a.
From 2000 onwards, numerous other authors have calculated the mass loss of the star, finding slightly lower values, ranging from 1.7 and 2.5 millionths of a solar mass per year. The only value that deviates significantly from previous estimates is that provided by F. Martins and others in a 2015 study, in which the mass loss of Alnilam is calculated at 5.6 × 10⁻⁷ M⊙/a. In any case, the annual dispersion of matter by the blue supergiant is tens of millions of times greater than that of the Sun.
A lot of research has also been done to estimate the rotation speed of Alnilam. Unfortunately, the inclination of the star’s rotation axis with respect to the Earth observer is not known. Consequently, the estimates of its rotation speed provide only the minimum value, which could correspond to the actual rotation speed only if the axis on which Alnilam rotates is exactly perpendicular to our line of sight. In the studies carried out from the 70s of the last century onwards, the values found by researchers range from a minimum of 42 to a maximum of 85 km/s. These estimates correspond to a minimum rotation period of approximately 20 to 50 Earth days if one considers a stellar radius of 22.9 million km. This data is sufficient to state that Alnilam is not a fast rotator.
Another feature highlighted by spectroscopic studies is the variability of the star. In fact, it has been found that Alnilam is an Alpha Cygni variable. From the complex variations of the spectral lines, a period of 1.9 days emerges, which is probably associated with a cyclic series of pulsations, which could have to do with the advanced evolutionary stage reached by Alnilam.
It remains to be asked what the future fate of this magnificent star will be. From the ratio of atmospheric carbon to nitrogen, researchers estimate that Alnilam is now going for the first time through the evolutionary phase that will lead it to become a red supergiant. A series of expansions and contractions, temperature changes, and further mass losses will follow. All this will eventually lead to an inevitable conclusion, the disintegration of Alnilam during a powerful supernova explosion, which — due to the star’s high initial mass — will most likely leave behind a stellar-mass black hole.
When Alnilam explodes, those who witness it will attend an extraordinary show. The relative proximity of the event will allow astronomers of that future era to study the explosion and formation of the supernova remnant with an unprecedented wealth of detail. What we learned from SN 1987 A, the supernova that exploded in the Large Magellanic Cloud thirty-three years ago, which remains the closest supernova to Earth and best studied from the invention of the telescope onwards, will pale in comparison with what we can learn from the explosion of Alnilam.
 For those who feel confused, stellar magnitudes work in reverse: smaller values indicate higher brightness.
 The division of the brightness class I into Ia and Ib was introduced precisely on that occasion.
 Despite searches, a binary companion of Alnilam was never found. If it had been found, the mass of both stars could have been derived from their orbital parameters, using Newtonian gravity formulas. Failing this, supergiant mass estimates can only be indirect, that is, based on stellar evolution models.
 The surface gravity is reported in the scientific literature with a value that represents the base-10 logarithm of the acceleration of gravity, expressed in the CGS system of units (centimeter/gram/second). In the study examined here, for example, Alnilam’s gravity is reported with the following formula: log g = 3.0 ± 0.1. Carrying out the calculation, we obtain 10³ = 1,000, i.e., 1,000 cm/s² or 10 m/s² — a gravity acceleration which, coincidentally, is just a little higher than the Earth’s gravity at sea level (9.81 m/s²).
 That is, the magnitude that Alnilam would have considered the radiation emitted in each region of the electromagnetic spectrum if we observed the star from a distance of 10 parsecs.
 The exact value reported in the study by Lamers is log(L/L⊙)=5.78 +0.22 −0.14. L⊙ represents the solar luminosity.
 A measurement obtained in 1966 with interferometric observations by R. Hanbury Brown, J. Davis, and L. R. Allen.
 64.5 solar masses are 128 thousand billion billion billion kilograms or, more briefly, 1.283 × 10³² kg.
 Deneb’s apparent visual magnitude is, in fact, equal to 1.25, while that of Alnilam is equal to 1.69.
 According to data reported in a 2008 study, the Sun loses mass at a rate equal to 9.13×10⁻¹⁴ M⊙/a, that is, just over nine hundred thousandths of a billionth of a solar mass per year.
 Alnilam emits radiation in almost all bands of the electromagnetic spectrum, from X-rays to radio waves.