Mira period-luminosity relations in Rubin z and y bands, and NGC 6822

85 Mira stars in the LMC give the first z and y band period-luminosity relations for Rubin, and one Mira puts NGC 6822 at 23.32 mag. We check the distance and the period-finding with fibub tools.

Mira stars are huge, cool red giants that swell and shrink over hundreds of days. The longer a Mira’s period, the brighter it is, so a measured period tells you how bright the star really is, and comparing that with how bright it looks gives its distance. A new paper sets up this rule for the near-infrared z and y filters of the Vera C. Rubin Observatory, which will watch the whole southern sky for ten years. It then uses one Mira to measure the distance to the nearby galaxy NGC 6822. This article shows how the period finder and the HR diagram let anyone check the distance and see why Rubin’s longer runs of data will matter.

The paper in brief

The team took repeated images of the Large Magellanic Cloud (LMC), a small galaxy next to ours, with the Dark Energy Camera on the 4-metre Blanco telescope in Chile. They tied the brightness scale directly to Rubin’s own early data release, Data Preview 2. From 85 oxygen-rich Miras they built the first period-luminosity relations in Rubin’s z and y bands: straight lines that link the logarithm of the period to the star’s brightness.

The lines are not straight all the way. They bend near a period of 300 days, so the calibrated relations only cover Miras with shorter periods. The team also looked for known Miras in Rubin’s preview data for three small galaxies in our Local Group. IC 1613 has too few points per star. NGC 6822 and the Sagittarius dwarf irregular galaxy have points over too short a time. Still, for one Mira in NGC 6822 they found a distance modulus of 23.32 ± 0.24 mag, in line with earlier measurements.

The paper is accepted by The Astronomical Journal. Why it matters: Rubin will find many thousands of Miras. A calibration made in Rubin’s own filters lets astronomers turn those light curves into distances straight away, as a check on the Cepheid and red giant distance ladders.

Checked with the tools

We used the paper’s NGC 6822 result, a published distance to the LMC, and fibub’s maths. We did not have the light curves, so Step 3 uses a simulated Mira and says so.

Step 1: from modulus to distance. The distance modulus μ is how much fainter a star looks than it would at 10 parsecs (a parsec is 3.26 light years), in magnitudes: μ = 5 log₁₀ d − 5, with d in parsecs. The HR diagram tool uses exactly this formula. Run backwards, μ = 23.32 gives 461 kiloparsecs, about 1.5 million light years (Figure 1).

Step 2: what ± 0.24 mag means. Magnitudes are a log scale, so the error is a percentage. The range 23.08 to 23.56 mag is 413 to 515 kiloparsecs, about ± 11%. That is good for a single star seen over a short time, and it shows why the paper asks for more data before claiming a precise distance.

Distance modulus and distanceDistance modulus and distance15202510 kpc100 kpc1 MpcDistance (parsecs, log scale)Distance modulus μ (mag)LMC 49.6 kpcNGC 6822 461 kpcLMC (literature)NGC 6822 (paper)fibub.com
Figure 1. The distance modulus curve drawn with the HR diagram tool's formula, with the LMC (literature value) and the paper's NGC 6822 Mira distance marked. The error bars are the quoted uncertainties in magnitudes.

Step 3: why the length of the data matters. We made a pretend Mira with a 250-day period (below the 300-day bend), a swing of about 1 magnitude and a little noise, and observed it two ways: 19 visits over 120 days, and 100 visits over three observing seasons. We ran the Period finder’s Lomb-Scargle maths on each, without telling it the answer (Figure 2). With 120 days, the data cover less than half a cycle. The periodogram is a broad hump and its top lands at 158 days, far from the truth. With three seasons, the peak is sharp and sits at 249.7 days, within half a percent of the 250 days we put in. A smaller peak near 750 days comes from the yearly gaps. This is the paper’s point about short baselines, made with a toy example.

Period finder on a simulated Mira: short run vs three seasonsPeriod finder on a simulated Mira: short run vs three seasons00.250.50.751501002005001000Trial period (days, log scale)Power300 d break120 days of dataThree seasonsfibub.com
Figure 2. Periodograms from the Period finder's maths for a simulated 250-day Mira: 120 days of data (broad hump) against three observing seasons (sharp peak). Simulated, not Rubin data.

Step 4: how much fainter. The LMC’s distance modulus is 18.477 mag (49.6 kiloparsecs). NGC 6822 is 4.84 mag further down the scale, so it is 9.3 times farther away, and a Mira of the same period there looks about 87 times fainter. That is why a big telescope like Rubin is needed to follow Miras in other galaxies.

Check Paper Our tool result
Distance to NGC 6822 μ = 23.32 ± 0.24 mag 461 kpc (413 to 515 kpc)
Uncertainty as a percentage ± 0.24 mag about ± 11% in distance
Period from a short run baselines too short for a full check 158 days found for a 250-day star (simulated, 120 days)
Period from three seasons expected from the first Rubin data release 249.7 days found (simulated)
NGC 6822 against the LMC (LMC used for the calibration) 9.3 times farther, 87 times fainter

How we made the graphs

The charts and every number above come from scripts/blog-figures/mira-period-luminosity-rubin-zy-ngc-6822-distance.mjs in fibub’s public code. It calls the tools’ own functions:

  • modulus from src/lib/sci/hrd.ts, the maths behind the HR diagram. The script turns a modulus back into a distance by searching for the distance whose modulus matches, so it uses the same formula as the tool.
  • lombScargle and peaks from src/lib/sci/lombscargle.ts, the engine of the period finder.

Where the inputs come from: the 23.32 ± 0.24 mag modulus and the 300-day bend are from the paper; the LMC’s 18.477 ± 0.026 mag is from Pietrzyński and colleagues (2019). The simulated Mira’s period (250 days), shape, swing (0.5 mag each way), noise (0.03 mag), visit spacing (about every 6 nights) and season length (200 nights) are our own choices, labelled in the script.

How we test the maths

On every change to fibub, tests/sci.mjs and tests/sci-tools/hr-diagram.mjs check these functions against known answers:

  • The Period finder must recover a 2.345-day sine from 120 unevenly spaced points to within 0.002 days, and its fast method must match a direct sum to nine decimal places.
  • In pure noise, about 10% of 200 trials must cross the 10% false-alarm line, as they should if the false-alarm numbers are right.
  • The HR diagram tool must put the Sun at 10 parsecs with a modulus of exactly 0 and an absolute magnitude of 4.83, and its absolute magnitudes for sample stars must match NumPy and astropy to a billionth of a magnitude.

Check it yourself

The distance, on the HR diagram:

  1. In One star per line, with a header row, type two lines: name,B-V,V and then Mira-like star,1.5,18.0 (a made-up star of magnitude 18).
  2. Set Distances to One distance for every star and type 461000 in Distance (pc).
  3. The summary reads median distance 461000 pc (modulus 23.32), and Absolute magnitude M(V) reads −5.32: the star’s apparent magnitude minus 23.32. Try 413000 and 515000 to see the modulus move to 23.08 and 23.56.

The period, on the period finder:

  1. Paste a Mira light curve as time (days) and magnitude, one point per line. Public light curves of Miras are available from survey archives such as OGLE or from variable star observers.
  2. Leave Values are magnitudes (brighter is up) ticked. Type 100 in Shortest period and 1000 in Longest period, the Mira range.
  3. Read Best period and tap the top row of Strongest peaks to fold the data. If the data span less than two cycles, expect a broad peak and a doubtful period, as in Figure 2.

What researchers can use it for

  • Turn a distance modulus into a distance and an honest percentage error, with the HR diagram.
  • Find a long period in uneven data and see whether the run is long enough to trust it, with the period finder.
  • Plan observations: simulate your own cadence and season length to see how many seasons a 300-day Mira needs.
  • Compare galaxies: place stars from two galaxies at their own distances and read off absolute magnitudes.
  • Teach how pulsating stars give distances, one of the first rungs of the cosmic distance ladder.

The fine print

  • The paper’s relations themselves (their slopes and zero points) are not used here; we only check the distance it reports and the period-finding problem it describes.
  • The light curves in Figure 2 are simulated. They show what a short run does to a Lomb-Scargle search, not what Rubin’s preview data contain.
  • We ignored dust. Any dimming by dust in our galaxy towards NGC 6822 must be removed before a modulus becomes a distance; in the near-infrared z and y bands this correction is smaller than in visible light.
  • The NGC 6822 distance rests on one star, as the paper says; more Miras and longer data from Rubin will tighten it.

Sources

  • Ngeow, C.-C. and Bhardwaj, A. (2026), Calibrated Period-Luminosity Relations for Mira variables in the Large Magellanic Cloud in the Vera C. Rubin Observatory zy-band Photometric Systems, accepted by The Astronomical Journal; arXiv:2611.17628 (posted 13 November 2026). Licence CC BY 4.0, checked on the arXiv page, 2026-11-16. Values used here, from the abstract: 85 oxygen-rich Miras in the LMC, DECam light curves calibrated to Rubin Data Preview 2, a break in the relations near 300 days, and a distance modulus of 23.32 ± 0.24 mag for one Mira in NGC 6822.
  • LMC distance modulus 18.477 ± 0.026 mag (literature value used for comparison): Pietrzyński, G. et al. (2019), A distance to the Large Magellanic Cloud that is precise to one per cent, Nature 567, 200-203.
  • Mira period range (about 100 to 1,000 days) and their nature as pulsating red giants: standard textbook facts.

If you are an author or publisher and want a credit or a passage changed or removed, tell us and we will act quickly.

Tools in this article

Read next