Pith. sign in

REVIEW 3 major objections 4 minor 52 references

Constraining young massive cluster properties with radio-continuum observations: The Arches cluster

T0 review · 3 major / 4 minor · reviewed 2026-08-15 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read Radio-continuum flux distributions of a young cluster's most massive stars can determine its age, mass, and IMF slope; applied to the Arches cluster they favour an age near 2.5 Myr and a mass around $2.7\times10^4\,M_\odot$ under an…

desk verdict A useful new forward-model application of radio continuum to YMC dating, but the recovery tests undermine the 'all models' age consistency claim and the mass/IMF constraints are partly calibrated to the target cluster. read the letter →

arxiv 2607.24580 v1 pith:BT26BTRT submitted 2026-07-27 astro-ph.GA astro-ph.SR

M. Cano-González , R. Schödel , A. Alberdi , F. Najarro , J. Moldón , M. Pérez-Torres , F. Nogueras-Lara This is my paper · ORCID
classification astro-ph.GAastro-ph.SR
keywords ArchesclusteryoungmassiveclustersradiocontinuumstellarwindsinitialmassfunctionGalacticCentrecolliding-windbinariesMCMCparameterinference
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper aims to establish that the radio-continuum flux density distribution of the most massive stars in a young massive cluster carries enough information, by itself, to constrain the cluster's age, mass, and initial mass function. For the Arches cluster, deep 10 GHz observations were compared against synthetic clusters built from three stellar evolutionary models at two metallicities; every combination prefers an age between 2 and 3 Myr, with a model-averaged IMF slope of $-1.85^{+0.28}_{-0.20}$. When the IMF slope from earlier infrared work is used as a prior, the recovered cluster mass peaks near $2.7\times10^4\,M_\odot$, with a lower limit around $2\times10^4\,M_\odot$. The payoff is practical: radio emission is unaffected by extinction, so the same approach could date and weigh newly discovered, heavily obscured clusters in the Galactic Centre without prior spectroscopy.

What carries the argument

The load-bearing object is the synthetic cluster's radio flux density distribution (FDD): the set of 10 GHz flux densities of its radio-bright stars, thresholded at the same $5\sigma$ detection limit as the observations. The FDD is assembled from two components: thermal emission from WN stars, computed from evolutionary-model mass-loss rates via a standard free-free wind prescription, and non-thermal emission from colliding-wind OB binaries, drawn from a Gamma distribution $f(S_X;\alpha_\Gamma,\theta_\Gamma)$ calibrated on the observed 10 GHz fluxes of 23 OB systems, with only a $7/32$ detection fraction applied. The observed FDD is compared with the simulated one through a likelihood that sums Gaussians centred on each simulated flux; star counts of WN, OB, and WC types add Poisson terms. An MCMC sampler explores age, cluster mass, and IMF slope, and the age posterior is dominated by the flux term, while the mass-IMF degeneracy lives in the counts terms.

What would settle it

Observe a second Galactic young massive cluster of independently known age with the same 10 GHz depth: if the radio-only posterior age disagrees by more than the quoted systematic uncertainties for all three evolutionary models, the radio-dating claim fails. Or enlarge the colliding-wind binary calibration sample: if the best-fit Gamma shape and scale parameters move by more than their current bootstrap uncertainties, the predicted IMF slope and mass range should shift correspondingly, and that shift can be checked against infrared star counts.

Watch

Extended reading notes

Core claim

The central claim is that the shape of a young massive cluster's stellar radio flux density distribution is a genuine diagnostic of cluster parameters, not just a by-product of stellar classification. The authors show this by modelling each synthetic cluster through two emission channels, thermal free-free radiation from hydrogen-rich Wolf-Rayet (WN) stars and non-thermal synchrotron radiation from colliding-wind OB binaries, and comparing the resulting flux distribution to the Arches data with a Bayesian likelihood. The recovered ages sit at $2\lesssim t_{\rm age}/{\rm Myr}\lesssim 3$ for all three evolutionary models, GENEC, PARSEC, and MIST, and both metallicities, solar and super-solar, with an IMF slope of $\alpha_{\rm IMF}=-1.85^{+0.28}_{-0.20}$ when the mass-IMF degeneracy is marginalized, and, after adopting the infrared IMF prior, a cluster mass of $\sim2.7\times10^4\,M_\odot$ bounded below by $\gtrsim2\times10^4\,M_\odot$. The age constraint is carried almost entirely by the flux term of the likelihood, which means the age survives even when the stellar types are not known spectroscopically.

Load-bearing premise

The load-bearing assumption is that the simulated non-thermal radio emission is faithful: the Gamma distribution fitted to only 23 observed colliding-wind OB binaries, plus the 7/32 detection fraction, governs how many and how bright the synthetic OB radio stars are, so if that calibration is unrepresentative the inferred mass and IMF slope lose their independent constraining power, though the age result rests mainly on the thermal Wolf-Rayet term and is less affected.

Editorial extensions

If this is right

  • A cluster's age can be measured from radio continuum alone, with no need for prior spectroscopic classification of the stellar content.
  • The Arches cluster is young: its radio flux distribution is incompatible with the roughly 3.5 Myr age inferred in some infrared work and favours about 2.5 Myr.
  • With the infrared IMF prior, the total initial mass of the Arches cluster is bracketed between roughly $2\times10^4$ and $3.7\times10^4\,M_\odot$.
  • The estimated IMF slope of about $-1.85$ favours a top-heavy IMF in the Galactic Centre, though the mass-IMF degeneracy means it is not yet a decisive statistical exclusion of the standard steep IMF.
  • The same forward-modelling machinery can be applied to other Galactic young massive clusters, provided individual radio stars can be resolved and their emission types identified.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • Because the age posterior decouples better at fainter flux densities, deeper radio surveys should be able to distinguish the 2-3 Myr range from older alternatives more sharply; this is a testable prediction for next-generation radio arrays.
  • A larger sample of colliding-wind OB binaries would shrink the Gamma-distribution calibration errors and, in turn, tighten the IMF slope and mass, so the method's biggest lever is observational, not theoretical.
  • If the framework is applied to a cluster with multiple star-formation episodes, the posterior should develop two age peaks; the absence of such structure in the Arches data is consistent with a single coeval burst.
  • The same radio flux distribution could be used to build a luminosity function for unresolved young clusters, potentially dating extragalactic super star clusters once sensitivity allows resolving their radio light.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 4 minor

Summary. The manuscript presents a Bayesian forward-modeling analysis of the Arches cluster using VLA 10 GHz continuum flux densities of 22 radio stars (excluding F6). Synthetic clusters are generated with SPISEA using GENEC, MIST, and PARSEC evolutionary tracks at solar and super-solar metallicity; a mixture-like likelihood compares simulated and observed radio flux densities plus Poisson count terms for WN, OB, and WC populations. The authors report preferred ages in the 2-3 Myr range for all models and metallicities, a model-averaged IMF slope of -1.85 (+0.28/-0.20), and, when the Hosek et al. (2019) IMF prior is adopted, a cluster mass near 2.7e4 Msun with a lower limit of roughly 2e4 Msun. They argue that radio continuum data alone can usefully constrain YMC parameters, particularly age, and that their results favor a top-heavy IMF in the Galactic Centre.

Significance. If the result holds, this is a valuable new avenue: radio observations bypass extinction and can provide age constraints for heavily obscured Galactic YMCs without prior spectroscopy. The paper's strengths include the use of three independent evolutionary models, explicit treatment of thermal (WN) versus non-thermal (OB colliding-wind binary) emission, mock recovery tests with quoted systematic uncertainties, and an unusually detailed discussion of model-specific mass-loss and clumping prescriptions. The falsifiable prediction that radio flux-density distributions of the most massive stars can date a YMC at ~8 kpc is clear and testable with future SKA-Mid observations. The main caveats concern calibration of the non-thermal component and the recovery tests' limited ability to validate the age scale at the older end.

major comments (3)
  1. [§3.4, §4.2, §4.4, Fig. 9, Table 3] The abstract's central claim that all models and metallicities return preferred ages in the 2-3 Myr range is not supported by the paper's own recovery tests. In §4.2 the MIST models show a significant, increasing underestimation of recovered age with true age; the authors do not apply the ~0.6 Myr bias correction because the fit is deemed statistically unreliable, and instead add only ≲0.3 Myr of systematic error. In §4.4 the GENEC solar-metallicity posterior is explicitly described as an upper limit at ≲2.5 Myr with no discernible peak, and Table 3 lists it as '≲2.5'. Only PARSEC shows insignificant age bias. The model-averaged ages (2.54 and 2.43 Myr) are therefore pulled down by an uninformative GENEC upper limit and a potentially uncorrected MIST bias; if the MIST bias were real, the corrected age would be ≳3.3 Myr. Furthermore, the recovery tests use true ages only within ±0.5 Myr of the observed posterior peak (§3.4), so they cannot discriminate against the ~3.5 Myr literature ages that the paper argues against. The abstract and Section 5 need to be reworded to distinguish PARSEC from the other models and to present the GENEC result as an upper limit.
  2. [§3.2.2, Eq. (1), §4.6] The non-thermal component is calibrated to the very data used for inference. The Gamma distribution in Eq. (1) is fitted to the observed flux densities of 23 OB colliding-wind binaries drawn from De Becker & Raucq (2013) combined with the Arches and Quintuplet radio stars, i.e., a sample that includes the target cluster, and the detection probability for synthetic OB systems is set to 7/32, the Arches observed fraction. The minimum orbital period of ~100 days is likewise chosen to reproduce the observed Arches O supergiants. Consequently, the OB count likelihood L_OB and the non-thermal part of the flux likelihood are not independent predictions but re-statements of the data, weakening the claimed constraining power on IMF slope and cluster mass. The age result is less affected because §4.6 shows the age posterior is dominated by the flux term, but the published alpha_IMF and M_cl constraints require this dependence to be stated explicitly and tested with an external calibration sample. In addition, the fit of Eq. (1) does not appear to account for the 5σ detection limit of 0.012 mJy, which should bias the Gamma shape and scale estimates for a truncated sample.
  3. [§3.2.1, §3.3.2, Table 3] Multiple ingredients of the forward model are tuned to the Arches data without an explicit sensitivity analysis: the GENEC clumping scaling by sqrt(1.75/10) applied in the 0.3<X_H<0.6 range, the WNh selection thresholds (0.2<X_H<0.6 and M_ini>80 Msun), the sigma_floor term introduced to keep acceptance fractions acceptable, and the orbital-period cutoff just mentioned. The systematic uncertainties reported in Table 3 are derived only from mock-recovery scatter and do not propagate variations in these choices. Because the paper draws quantitative conclusions from the resulting posteriors, such as M_cl in the 2.0-3.7e4 Msun range and alpha_IMF of -1.85 with asymmetric uncertainties, the authors should either marginalize over these settings or demonstrate that the results are stable when they are varied over sensible ranges.
minor comments (4)
  1. [Abstract] The sentence 'All models and metallicities return preferred ages in the 2≲t_age/Myr≲3 range' should be qualified to reflect that the GENEC solar value is an upper limit and that MIST carries a large, uncorrected systematic bias.
  2. [Table 3] The entry '≲2.5' for GENEC at Z=0.014 does not contain the same statistical uncertainty information as other entries; consider reporting a quantile-based upper limit or a caveat in the table notes.
  3. [§3.2] The footnote stating that the author is willing to share modified scripts upon reasonable request should be replaced by a persistent repository URL to meet standard reproducibility expectations.
  4. [§3.3.4] The DBSCAN parameters are quoted as ranges (min_size 200-600, eps 0.4-0.65); the paper should state the exact values used for each posterior or justify why a range is acceptable.

Circularity Check

2 steps flagged · score 6.0 of 10

Non-thermal OB flux distribution and OB detectability are calibrated to Arches data and then used as constraints on Arches; partial circularity.

  1. fitted input called prediction [Sect. 3.2.2 (Eq. 1); likelihood Eqs. (3)-(6) in Sect. 3.3.2]
    "We combined the OB-OB binary supergiant catalogue from De Becker & Raucq (2013) with the OB radio-stars identified in the Arches and Quintuplet clusters (Cano-González et al. 2024, 2025), that we also assume to be colliding wind binaries. ... We then modelled the distribution with a Gamma function ... The best fit parameters to the observed distribution are αΓ=0.19+0.21−0.18 (shape parameter) and θ=0.29±0.09 mJy (scale parameter), where the uncertainties were obtained via bootstrapping."

    Equation (1) is the generative model for every non-thermal OB flux that enters the simulated FDD, and L_S (Eqs. 3-6) compares that FDD to the observed Arches X-band fluxes. The 23 OB systems used to fit Eq. (1) explicitly include the Arches and Quintuplet OB radio-stars (Cano-González et al. 2024, 2025) - i.e., the target data. Thus the simulated non-thermal flux distribution is the empirical distribution of the very data it is used to constrain; the OB part of the flux-likelihood is a smoothed comparison of the data to a resampling of itself, not an independent prediction. The paper states (Sect. 4.6) that the flux term L_S dominates the age posterior, so this calibration feeds the central age/mass/IMF inference.

  2. self definitional [Sect. 3.2.2; Poisson count likelihood Eq. (9) in Sect. 3.3.2]
    "In particular, only 7/32∼22% of the Arches O supergiants (types O Ia (+)) are clear radio detections. Therefore, even if a synthetic multiple O-type system met the previous criteria, only 22% of them were considered to have properties that make them observable. ... The Arches cluster contains N_obs_WN=15 WN(h) radio-stars, N_obs_OB=7 OB radio-stellar systems, and no detected WC stars (N_obs_WC=0)."

    The forward model sets the probability that a synthetic OB system enters the FDD to the observed Arches fraction 7/32, and additionally tunes the minimum orbital period (≳100 days) to 'better reproduce' the observed Arches radio O supergiants. The same observed count, N_obs_OB=7, is then used as the data in the Poisson likelihood L_OB=P(7|λ_OB). Hence the expected OB count is calibrated to the very count it is asked to predict; the OB count term cannot independently constrain M_cl or α_IMF, but only select parameters consistent with its own calibration. This matters because Sect. 4.6 says the WN/OB/WC count terms encode the M_cl-α_IMF degeneracy.

full rationale

Two components of the forward model are calibrated to the very Arches data that the likelihood then uses as constraints. First, the non-thermal OB flux distribution is a best-fit Gamma function to 23 observed OB colliding-wind binaries, a sample that explicitly includes the Arches (and Quintuplet) radio-stars; synthetic non-thermal fluxes are drawn from this fit, so the flux-likelihood term for the OB population compares the data to a resampling of the calibration sample rather than to an independent prediction. Second, the probability that a synthetic OB system is observable is set equal to the observed Arches detection fraction (7/32), and the same observed count (N_obs_OB=7) is then the datum of the OB Poisson likelihood; the orbital-period cutoff is also chosen to reproduce the observed Arches radio O supergiants. These steps make the OB non-thermal and counts channels partially self-definitional. The thermal WN channel is more independent: it is based on evolutionary-model mass-loss rates, Wright-Barlow free-free emission, and published clumping factors, though the GENEC clumping scaling is checked against the same Arches WNh mass-loss range. The age result therefore retains independent content in the WN thermal fluxes, but the abstract's 'radio data alone' claim is weakened by the non-thermal calibration. The recovery-test issues (GENEC degeneracy above ~2.5 Myr, MIST bias) are correctness/validation concerns, not circularity, and do not change this score.

Assumptions & free parameters 7 free parameters · 6 assumptions · 0 invented entities

The central inference depends on physically motivated but partly target-calibrated modeling assumptions. The nonthermal OB flux distribution, the OB detection fraction, and the GENEC clumping scaling are all fitted or tuned using the Arches and Quintuplet observations themselves, which reduces the independence of the mass and IMF results. The age result is largely set by thermal WN fluxes and is more robust, and mock recovery tests add supporting evidence.

free parameters (7)
  • Gamma shape parameter alpha_Gamma = 0.19 (+0.21/-0.18)
    Fitted to the observed X-band fluxes of 23 OB colliding-wind binaries from De Becker & Raucq (2013) and Cano-González et al. (2024, 2025), including Arches sources; used in Eq. (1) to generate synthetic nonthermal fluxes.
  • Gamma scale parameter theta_Gamma = 0.29 +/- 0.09 mJy
    Same sample as alpha_Gamma; defines the characteristic flux of the simulated nonthermal OB population.
  • OB radio detection fraction = 7/32 ~ 0.22
    Fraction of Arches O supergiants detected as radio sources; applied to eligible synthetic OB systems, then used in the L_OB count likelihood.
  • Minimum orbital period for synthetic OB CWBs = larger than 100 days
    Tuned so the number of synthetic nonthermal systems reproduces the 7 observed Arches O supergiant radio stars; a target-specific calibration.
  • GENEC clumping mass-loss scaling = sqrt(1.75/10) ~ 0.418
    Applied to Gräfener & Hamann (2008) mass-loss rates for hydrogen-rich WN stars with X_H > 0.3 so that GENEC thermal fluxes match the observed Arches WNh mass-loss range.
  • WNh selection thresholds = 0.2 < X_H < 0.6; M_ini > 80 M_sun
    Chosen so synthetic WN stars resemble the helium abundances and spectral types of Arches WNh stars; the lower limit is tied to Arches helium values.
  • sigma_floor = <= 0.1 fractional variance
    Added to observed flux uncertainties in Eq. (5) to smooth the posterior and keep acceptance fraction above 0.1; a regularization hyperparameter rather than a physical parameter.
assumptions (6)
  • domain assumption WNh stars are the only thermal emitters and O-type non-WR stars emit only non-thermally
    Section 3.1: motivated by the observed spectral-index bimodality; this separates the synthetic flux into thermal and nonthermal components and underpins the whole FDD construction.
  • domain assumption Thermal radio flux follows the Wright and Barlow (1975) free-free prescription with a fully ionized wind, Z_wind = gamma = 1, T_e = 10^4 K
    Section 3.2.1: converts synthetic mass-loss rates into thermal radio flux; a standard prescription in hot-star radio astronomy.
  • domain assumption The nonthermal flux of OB supergiants is distributed as the Gamma function fitted to the 23-source CWB sample
    Section 3.2.2, Eq. (1): needed to assign random flux values to synthetic OB systems; the fitted parameters include Arches and Quintuplet sources.
  • domain assumption The cluster is coeval and dominated by single-star evolution, so binary interactions do not significantly alter the radio population
    Section 4.8: follows Clark et al. (2023); justifies using SPISEA single-star isochrones for all cluster members.
  • domain assumption The cluster is at 8 kpc and the distance uncertainty is not propagated
    Section 3.2, footnote 1: authors choose not to vary the distance because it affects all members equally and is not precisely defined.
  • domain assumption The Hosek et al. (2019) Gaussian prior on IMF slope is used for the mass estimates
    Section 3.3.3: an external infrared constraint; adopted in the runs that produce the 2.7 x 10^4 M_sun mass and the 20,000 M_sun lower limit.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Constraining young massive cluster properties with radio-continuum observations: The Arches cluster." pith.science (2026). https://pith.science/paper/BT26BTRT

@misc{pith2026260724580,
  author       = {Pith},
  title        = {Pith review of: Constraining young massive cluster properties with radio-continuum observations: The Arches cluster},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/BT26BTRT}},
  note         = {Machine review of arXiv:2607.24580}
}
abstract

The Arches cluster, located in the Galactic Centre (GC) is one of the best astrophysical laboratories to study the properties of massive stars and young massive clusters (YMCs). However, several fundamental parameters of the Arches cluster remain uncertain. Our goal is to constrain key cluster parameters (cluster age, mass, and initial mass function, IMF) by comparing the observed stellar radio flux density distribution of the Arches cluster to those derived from a set of synthetic clusters. We use the deep X-band (10 GHz) Very Large Array data from our previous radio continuum study of the Arches cluster. We model each simulated cluster with three parameters: age, mass, and IMF slope. We use three different stellar evolutionary models: GENEC, PARSEC, and MIST at two different metallicities, solar ($Z=0.014$) and super-solar ($Z=0.020$). We run Markov-chain Monte-Carlo simulations for each model/metallicity combination in order to explore parameter space. All models and metallicities return preferred ages in the $2\lesssim t_{\rm age}/{\rm Myr}\lesssim 3$ range. We obtain an IMF slope of $\alpha_{\rm IMF}=-1.85^{+0.28}_{-0.20}$, averaged over all models, where uncertainties are dominated by the degeneracy between cluster mass and IMF slope. If we use the IMF slope from previous infrared studies as prior, the cluster mass distributions peak at $\sim2.7\times10^4\, M_\odot$ and we can establish a lower limit at $\gtrsim2\times10^4 M_\odot$ for the Arches cluster mass. Radio continuum observations of their most massive stars can be used to constrain YMC parameters. In the case of the Arches cluster, age can be determined regardless of prior spectroscopic information, which can be useful to characterise newly discovered YMCs in the GC. Our results support the idea that a top-heavy IMF may be preferred in the GC or in YMCs in general.

Figures

Figures reproduced from arXiv: 2607.24580 by the authors.

Figure 1
Figure 1. Average WN and WC counts for a 20 000 M⊙ cluster with αIMF = −1.8 (Hosek et al. 2019) and two different metallicities: solar (Z = 0.014) and super-solar (Z = 0.020). Each panel shows the results obtained for a different evolutionary model. We computed mean counts per cluster, age, and metallicity from 50 realisations. As reference, the Arches cluster hosts 15 known WN(h) stars and no WC star has been observed. of al… view at source ↗
Figure 2
Figure 2. Blue: Counts drawn from a truncated Γ distribution (1 500 iter￾ations with lower limit of 0.012 mJy, the 5σ rms of the observations used here; Cano-González et al. 2024). Orange: Observed flux densi￾ties from the 23 OB supergiants from De Becker & Raucq (2013) and Cano-González et al. (2024, 2025). corresponding central frequency of 10 GHz of the VLA X-band at the assumed distance of 8 kpc to the GC. The orange his￾… view at source ↗
Figure 3
Figure 3. Average counts of expected CWBs per 50 simulated clusters for the three evolutionary models considered. All clusters host a total of 20 000 M⊙ in stellar mass, and were created with a top-heavy IMF slope of αIMF = −1.8. As reference, in the Arches cluster, we detected a total of 7 OIa supergiants with the VLA. 3.3. MCMC simulations We explored the preferred regions of the parameter space with Markov-chain Monte-Carl… view at source ↗
Figures from the paper (5 more)
Figure 5
Figure 5. Figure 5: Posterior distribution of the three evolutionary models combined, at solar metallicity. All priors were flat (ln p(θ) = 0), which results in a clear degeneracy between Mcl and αIMF. 16000 24000 32000 40000 48000 Mcl (M ¯ ) 1.6 2.0 2.4 2.8 3.2 Age (yr) 1e6 2.2 2.0 1.8 1…
Figure 6
Figure 6. Figure 6: Same as [PITH_FULL_IMAGE:figures/full_fig_p007_6.png]
Figure 9
Figure 9. Figure 9: Recovered age versus true input age for all three models at solar (filled symbols) and super-solar (empty symbols) metallicity. The x = y line is shown as measure of goodness of recovery. Note the nearly horizontal tendency of the GENEC models, as well as the large sca…
Figure 10
Figure 10. Figure 10: Comparison of estimated ages of the Arches cluster in previous literature (black) and our results (blue). From left to right, the values are obtained from Figer et al. (2002); Schneider et al. (2014); Lohr et al. (2018); Hosek et al. (2019). ondary DBSCAN clusters are…
Figure 12
Figure 12. Figure 12: MIST radio luminosity function of a 20 000 M⊙ cluster for dif￾ferent ages. Vertical lines represent the 5σ sensitivity limit achieved with the VLA in (dashed line, Cano-González et al. 2024) and the SKAO-Mid sensitivity assuming 10 h of observations at band 5b (dotted…

Review history (2 revisions) →

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

52 extracted references · 32 canonical work pages

  1. [1]

    2007, A&A, 466, 137

    Ascenso, J., Alves, J., Beletsky, Y ., & Lago, M. 2007, A&A, 466, 137

    work page 2007

  2. [2]

    R., & Meyer, M

    Bastian, N., Covey, K. R., & Meyer, M. R. 2010, ARA&A, 48, 339

    2010

  3. [3]

    2012, MNRAS, 427, 127 Cano-González, M., Schödel, R., Alberdi, A., et al

    Bressan, A., Marigo, P., Girardi, L., et al. 2012, MNRAS, 427, 127 Cano-González, M., Schödel, R., Alberdi, A., et al. 2024, A&A, 692, A23 Cano-González, M., Schödel, R., Alberdi, A., et al. 2025, A&A, 700, A246

    work page 2012

  4. [4]

    2015, MNRAS, 452, 1068

    Chen, Y ., Bressan, A., Girardi, L., et al. 2015, MNRAS, 452, 1068

    2015

  5. [5]

    2016, ApJ, 823, 102

    Choi, J., Dotter, A., Conroy, C., et al. 2016, ApJ, 823, 102

    2016

  6. [6]

    Clark, J., Negueruela, I., Crowther, P., & Goodwin, S. P. 2005, A&A, 434, 949

    work page 2005

  7. [7]

    S., Lohr, M

    Clark, J. S., Lohr, M. E., Najarro, F., Patrick, L. R., & Ritchie, B. W. 2023, MNRAS, 521, 4473

    2023

  8. [8]

    S., Negueruela, I., Crowther, P

    Clark, J. S., Negueruela, I., Crowther, P. A., & Goodwin, S. P. 2005, A&A, 434, 949

    2005

Show all 52 references
  1. [9]

    2012, ApJ, 751, 132

    Clarkson, W., Ghez, A., Morris, M., et al. 2012, ApJ, 751, 132

    2012

  2. [10]

    A., Caballero-Nieves, S., Bostroem, K., et al

    Crowther, P. A., Caballero-Nieves, S., Bostroem, K., et al. 2016, MNRAS, 458, 624 De Becker, M. 2007, A&A Rev., 14, 171 De Becker, M. & Raucq, F. 2013, A&A, 558, A28

    2016

  3. [11]

    D., Cotera, A., et al

    Dong, H., Wang, Q. D., Cotera, A., et al. 2011, MNRAS, 417, 114

    2011

  4. [12]

    M., Beasley, A

    Dougherty, S. M., Beasley, A. J., Claussen, M. J., Zauderer, B. A., & Boling- broke, N. J. 2005, ApJ, 623, 447 Duchêne, G. & Kraus, A. 2013, ARA&A, 51, 269 Ekström, S., Georgy, C., Eggenberger, P., et al. 2012, A&A, 537, A146

    2005

  5. [13]

    J., & Melnick, J

    Espinoza, P., Selman, F. J., & Melnick, J. 2009, A&A, 501, 563

    2009

  6. [14]

    Figer, D. F. 2004, arXiv preprint astro-ph/0403088

    2004 arXiv

  7. [15]

    Figer, D. F. 2005, Nature, 434, 192

    2005

  8. [16]

    F., Najarro, F., Gilmore, D., et al

    Figer, D. F., Najarro, F., Gilmore, D., et al. 2002, ApJ, 581, 258

    2002

  9. [17]

    W., Lang, D., & Goodman, J

    Foreman-Mackey, D., Hogg, D. W., Lang, D., & Goodman, J. 2013, PASP, 125, 306

    2013

  10. [18]

    T., Schödel, R., Alberdi, A., et al

    Gallego-Calvente, A. T., Schödel, R., Alberdi, A., et al. 2021, A&A, 647, A110

    2021

  11. [19]

    & Geman, D

    Geman, S. & Geman, D. 1984, IEEE Transactions on pattern analysis and ma- chine intelligence, 721

    1984

  12. [20]

    2012, A&A, 542, A29 Gräfener, G

    Georgy, C., Ekström, S., Meynet, G., et al. 2012, A&A, 542, A29 Gräfener, G. & Hamann, W.-R. 2008, A&A, 482, 945

    2012

  13. [21]

    H., Meynet, G., Ekström, S., & Georgy, C

    Groh, J. H., Meynet, G., Ekström, S., & Georgy, C. 2014, A&A, 564, A30

    2014

  14. [22]

    G., Flaccomio, E., Albacete-Colombo, J

    Guarcello, M. G., Flaccomio, E., Albacete-Colombo, J. F., et al. 2024, A&A, 682, A49

    2024

  15. [23]

    2013, A&A, 556, A26

    Habibi, M., Stolte, A., Brandner, W., Hußmann, B., & Motohara, K. 2013, A&A, 556, A26

    2013

  16. [24]

    2019, A&A, 625, A57

    Hamann, W.-R., Gräfener, G., Liermann, A., et al. 2019, A&A, 625, A57

    2019

  17. [25]

    2010, MNRAS, 409, 628

    Harfst, S., Portegies Zwart, S., & Stolte, A. 2010, MNRAS, 409, 628

    2010

  18. [26]

    R., Anderson, J., et al

    Hosek, Matthew W., J., Lu, J. R., Anderson, J., et al. 2019, ApJ, 870, 44

    2019

  19. [27]

    W., Do, T., Lu, J

    Hosek, M. W., Do, T., Lu, J. R., et al. 2022, ApJ, 939, 68

    2022

  20. [28]

    W., Lu, J

    Hosek, Jr., M. W., Lu, J. R., Lam, C. Y ., et al. 2020, AJ, 160, 143

    2020

  21. [29]

    S., Figer, D

    Kim, S. S., Figer, D. F., Kudritzki, R. P., & Najarro, F. 2006, ApJ, 653, L113

    2006

  22. [30]

    2005, AJ, 130

    Lang, C., Johnson, K., Goss, W., & Rodriguez, L. 2005, AJ, 130

    2005

  23. [31]

    2012, ARA&A, 50, 107

    Langer, N. 2012, ARA&A, 50, 107

    2012

  24. [32]

    2018, A&A, 617, A66

    Lohr, M., Clark, J., Najarro, F., et al. 2018, A&A, 617, A66

    2018

  25. [33]

    & Bodensteiner, J

    Marchant, P. & Bodensteiner, J. 2024, ARA&A, 62, 21 Martínez-Arranz, Á., Schödel, R., Nogueras-Lara, F., Hosek, M., & Najarro, F. 2024, A&A, 683, A3

    2024

  26. [34]

    J., Paumard, T., et al

    Martins, F., Hillier, D. J., Paumard, T., et al. 2008, A&A, 478, 219

    2008

  27. [35]

    & Palacios, A

    Martins, F. & Palacios, A. 2013, A&A, 560, A16

    2013

  28. [36]

    2025, A&A, 701, A258

    Nguyen, C., Costa, G., Bressan, A., et al. 2025, A&A, 701, A258

    2025

  29. [37]

    2020, A&A, 641, A141

    Nogueras-Lara, F., Schödel, R., Neumayer, N., et al. 2020, A&A, 641, A141

    2020

  30. [38]

    1998, A&A, 333, 956

    Nugis, T., Crowther, P., & Willis, A. 1998, A&A, 333, 956

    1998

  31. [39]

    & Lamers, H

    Nugis, T. & Lamers, H. 2000, A&A, 360, 227

    2000

  32. [40]

    2011, Journal of Machine Learning Research, 12, 2825

    Pedregosa, F., Varoquaux, G., Gramfort, A., et al. 2011, Journal of Machine Learning Research, 12, 2825

    2011

  33. [41]

    Runacres, M. C. & Owocki, S. P. 2002, A&A, 381, 1015

    2002

  34. [42]

    E., de Koter, A., et al

    Sana, H., de Mink, S. E., de Koter, A., et al. 2012, Science, 337, 444

    2012

  35. [43]

    G., de Mink, S

    Schneider, F., Izzard, R. G., de Mink, S. E., et al. 2014, ApJ, 780, 117

    2014

  36. [44]

    2024, arXiv preprint arXiv:2406.04022

    Schoedel, R., Alberdi, A., Jimenez-Serra, I., et al. 2024, arXiv preprint arXiv:2406.04022

    2024 arXiv

  37. [45]

    2008, Nature, 455, 201

    Smith, N. 2008, Nature, 455, 201

    2008

  38. [46]

    K., Brandner, W., & Figer, D

    Stolte, A., Grebel, E. K., Brandner, W., & Figer, D. F. 2002, A&A, 394, 459 Szécsi, D., Agrawal, P., Wünsch, R., & Langer, N. 2022, A&A, 658, A125

    2002

  39. [47]

    Vink, J. S., ed. 2015, Astrophysics and Space Science Library, V ol. 412, Very Massive Stars in the Local Universe

    2015

  40. [48]

    S., Muijres, L

    Vink, J. S., Muijres, L. E., Anthonisse, B., et al. 2011, A&A, 531, A132

    2011

  41. [49]

    D., Dong, H., & Lang, C

    Wang, Q. D., Dong, H., & Lang, C. 2006, MNRAS, 371, 38

    2006

  42. [50]

    E., Heger, A., & Weaver, T

    Woosley, S. E., Heger, A., & Weaver, T. A. 2002, Reviews of modern physics, 74, 1015

    2002

  43. [51]

    Wright, A. E. & Barlow, M. J. 1975, MNRAS, 170, 41

    1975

  44. [52]

    2022, MNRAS, 511, 2814 Article number, page 12 of 14 M

    Yusof, N., Hirschi, R., Eggenberger, P., et al. 2022, MNRAS, 511, 2814 Article number, page 12 of 14 M. Cano-González et al.: Constraining young massive cluster properties with radio-continuum observations: The Arches cluster 16000 24000 32000 40000 48000 Mcl (M ⊙ ) 2.50 2.75 ...

    2022

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.