betaFeedback

WorksPhysics and Astronomy

A 105 M⊙10^5\,M_\odot Black Hole in the Bulgeless Spiral Galaxy IC 342: First Stellar-Dynamical MBHM_{\rm BH} Measurement with Keck/OSIRIS LGS-AO

Abstract

We present the first direct stellar-dynamical measurement of an intermediate-mass black hole (IMBH) in IC 342, a nearby nearly face-on (i≈31∘i\approx31^\circ) bulgeless spiral at D=3.45±0.13D=3.45\pm0.13 Mpc, using near-infrared integral-field spectroscopy with Keck/OSIRIS LGS-AO. The heavy foreground extinction (AF814W≈0.74A_{\mathrm{F814W}}\approx0.74 mag) renders the nucleus inaccessible to optical spectroscopy, making KK-band observations at ∼\sim2.3 μμm the only viable path to high-resolution nuclear kinematics. Two-dimensional stellar kinematics extracted from the CO bandheads via pPXF at 0.062′′0.062'' resolution reveal a central velocity dispersion rising from ∼\sim42 km s−1^{-1} at r≳0.2′′r\gtrsim0.2'' to a peak of ∼\sim49 km s−1^{-1}, providing kinematic evidence for a central dark mass within the nuclear star cluster (NSC; MNSC≈1.17×107 M⊙M_\mathrm{NSC}\approx1.17\times10^7~M_\odot). Modeling these kinematics with an ensemble of four JAM models, spanning cylindrical and spherical velocity-ellipsoid alignments and two PSF representations, combined with an MGE mass model constructed from a three-Sérsic plus exponential-disk decomposition of archival HST/WFC3 F814W imaging and incorporating a physically motivated stellar mass-to-light ratio gradient, driven by the young central NSC population, yields MBH=(1.95±0.61)×105 M⊙M_\mathrm{BH}=(1.95\pm0.61)\times10^5~M_\odot. Accounting for this stellar population gradient is essential for the detection: standard constant-M/LM/L models systematically overestimate the central stellar mass, hiding the BH signature and yielding only an upper limit. The stellar orbits are near-isotropic (β≈0β\approx0), typical of fast-rotating late-type spirals. This measurement places IC 342 among the rare bulgeless spirals with a direct dynamical MBHM_{\rm BH} at the IMBH mass scale, where both the NSC and BH are directly characterized, providing a new anchor for compact massive object co-evolution models.

Introduction

Introduction

The nuclei of late-type galaxies (LTGs) are among the most information-rich yet observationally challenging environments in the local Universe. Unlike the smooth cores of elliptical galaxies, LTG nuclei are shaped by a complex interplay of secular processes: bar-driven gas inflows, episodic starbursts, and the gradual accumulation of mass at the dynamical center. Two distinct central massive objects (CMOs) can inhabit these nuclei simultaneously: nuclear star clusters (NSCs; e.g., [8, 17, 128]) and massive black holes (BHs; MBH∼104M_{\mathrm{BH}} \sim10^{4}–106 M⊙10^{6}\ M_{\odot}; e.g., [100, 102]), yet the physical relationship between them ([53]) remains one of the most pressing open questions in galaxy evolution (e.g., [93, 105, 106]).

NSCs are compact (re∼2r_{\mathrm{e}} \sim2–55 pc), dense stellar systems common at the centers of both LTGs and dwarf early-type galaxies (ETGs: e.g., [121, 16, 61] (P. Côté et al. 2006)), with typical masses MNSC∼105M_{\mathrm{NSC}} \sim10^{5}–108 M⊙10^{8}\ M_{\odot} ([125]). Rather than being mutually exclusive, NSCs and BHs are now known to coexist across a broad range of galaxy types, from the Milky Way (e.g., [126, 50, 64, 124]) to nearby spirals (e.g., [51, 127, 99]) and dwarf ellipticals (e.g., [129, 101, 133, 59], and references therein), implying either independent formation channels or a physical coupling through shared gas supply and feedback.

The origin of this co-existence is deeply contested, with two competing pictures for how the two CMOs influence one another. In the first, the NSC comes first and feeds the BH: NSC formation, through the infall and coalescence of globular clusters or through secular gas inflows ([135, 4, 5]), assembles a dense central stellar system whose high stellar densities and runaway collisions provide both the seed and the fuel for subsequent BH growth ([40, 94, 32]). In the second, the BH ultimately limits the NSC: once sufficiently massive, the central BH suppresses further NSC growth through accretion feedback that heats and expels the nuclear gas reservoir ([2]). In merger-built galaxies, binary-BH scouring can additionally erode the cluster dynamically. Together, these processes produce an anti-correlation between MBHM_{\mathrm{BH}} and MNSCM_{\mathrm{NSC}} at intermediate masses ([3, 93]). The observed demographics reflect this competition: low-mass galaxies are NSC-dominated, high-mass galaxies are BH-dominated, and the handover occurs near a bulge mass of M⋆∼1010 M⊙M_{\star} \sim10^{10}\ M_{\odot} (or σe∼65 km s−1\sigma_{\mathrm{e}} \sim65\ \mathrm{km\ s^{-1}}), where the two CMOs overlap in mass and frequently co-exist ([94, 63]). The most powerful diagnostic of this competition is a direct, dynamical MBHM_{\mathrm{BH}} measurement in a galaxy whose NSC mass, age, and structural parameters are already well-characterized, precisely the transition regime that IC 342 occupies.

Additionally, calibrating the MBHM_{\mathrm{BH}}–host scaling relations ([86, 47, 48, 60, 69, 77, 54]) across the full range of galaxy morphologies is essential for understanding BH–galaxy co-evolution ([56]) and speciation ([58]). Yet of the ∼200\sim200 direct, dynamical MBHM_{\mathrm{BH}} measurements published to date, the majority reside in elliptical galaxies and lenticular galaxies with classical bulges ([117, 119 b]). Fewer than ∼30\sim30 measurements exist in late-type spiral galaxies (type ≥\geq Sbc), and the sample thins sharply toward bulgeless or pseudobulge-dominated Scd–Sd systems ([38]). This morphological census bias is not accidental: for LTGs with σe≲65 km s−1\sigma_{\mathrm{e}} \lesssim65\ \mathrm{km\ s^{-1}}, scaling relations predict MBH≲106 M⊙M_{\mathrm{BH}} \lesssim10^{6}\ M_{\odot}, placing the gravitational sphere of influence (SOI)1 at only ∼0.′′1\sim0.''1 even for the nearest targets (D∼3D \sim3 Mpc; [46, 98]). Resolving such small SOIs demands sub-0.′′10.''1 angular resolution, achievable in the optical with Hubble Space Telescope (HST) and the Multi Unit Spectroscopic Explorer (MUSE) on the Very Large Telescope (VLT), which is rendered ineffective by the heavy internal and/or foreground dust extinction prevalent in LTG nuclei.

The advent of near-infrared (NIR) adaptive optics (AO)-assisted integral-field spectroscopy (IFS) on 8–10 m telescopes has changed this picture. Observations targeting the CO band-head absorption features at ∼2.3 μm\sim2.3\ \mu\mathrm{m} achieve angular resolutions of ≲0.′′1\lesssim0.''1 while simultaneously penetrating dust extinction that would render optical counterparts unusable. When combined with Jeans Anisotropic Modeling (JAM)2 ([22, 24]) and Multi-Gaussian Expansion (MGE)3; Emsellem et al. [45]; M. Cappellari [21] decompositions of archival HST photometry, this approach has yielded robust SMBH mass determinations (MBH≈106M_{\mathrm{BH}} \approx10^{6}–1010 M⊙10^{10}\ M_{\odot}; D. D. Nguyen et al. [107, 109]); specially, in LTGs (D. D. Nguyen et al. [110]) and down to the intermediate-mass black hole (IMBH) mass (MBH≈102M_{\mathrm{BH}} \approx10^{2}–105 M⊙10^{5}\ M_{\odot}) frontier (H. N. Ngo et al. 2025a,b; D. D. Nguyen et al. 2025c; T. Q. T. Le et al. 2026).

IC 342 as the Ideal Laboratory

IC 342 (the “Hidden Galaxy”) is a low-mass (log⁡10(M⋆/M⊙)=10.37\log_{10}(M_{\star}/M_{\odot}) = 10.37; L. S. Pilyugin et al. [114]), nearby, face-on (i≈31∘i \approx31^{\circ}; L. P. Crosthwaite et al. 2000) LTG (SABcd) with a Tip-of-the-Red-Giant-Branch (TRGB) distance of D=3.45±0.13D = 3.45 \pm0.13 Mpc (P.-F. Wu et al. 2014), corresponding to a physical scale of ∼16\sim16 pc arcsec−1^{-1}, when adopting a flat Λ\LambdaCDM cosmology with H0=70H_{0} = 70 km s−1^{-1} Mpc−1^{-1}, ΩΛ=0.7\Omega_{\Lambda} = 0.7, and Ωm=0.3\Omega_{\mathrm{m}} = 0.3. It is the nearest massive bulgeless galaxy, and the dominant member of the IC 342/Maffei group (I. D. Karachentsev et al. [76]). Despite its intrinsic brightness, its proximity to the Galactic plane (b≈+10.6∘b \approx+10.6^{\circ}) subjects it to severe foreground extinction of AF814W≈0.74A_{\mathrm{F814W}} \approx0.74 mag (E. F. Schlafly & D. P. Finkbeiner [123]), rendering the nucleus effectively inaccessible to optical spectroscopy. NIR observations are therefore not merely advantageous, they are the only viable path to high-resolution nuclear kinematics in this galaxy.

The nuclear region of IC 342 has been extensively characterized across wavelengths, making it the best-studied bulgeless spiral galaxy nucleus in the local Universe. Photometric study shows that its central region is occupied by a prominent and young (∼60\sim60–630630 Myr) NSC with an effective radius of re≈r_{\mathrm{e}} \approx few pc and a stellar velocity dispersion of σNSC=33±3\sigma_{\mathrm{NSC}} = 33 \pm3 km s−1^{-1} (T. Böker et al. [18]). It is the most luminous among known NSCs in nearby LTGs and it displays a bar-like isophotal elongation on scales of ∼110\sim110 pc linked to a weak central stellar bar (T. Böker et al. [15]). The nucleus is embedded within a starburst ring of diameter ∼70\sim70 pc delineated by molecular gas, hosting two bright HII-region complexes of age ∼5\sim5 Myr (T. Böker et al. [15]; E. Schinnerer et al. [122]). High-resolution CO(2–1) interferometry (synthesized beam ≈1. ⁣′′2\approx1.\!^{\prime\prime}2) reveals ongoing non-circular inflow of molecular gas into the central few parsec at a rate of 0.0030.003–0.14 M⊙ yr−10.14\ M_{\odot}\ \mathrm{yr}^{-1} (E. Schinnerer et al. [122]), suggesting that conditions for both continued NSC growth and BH feeding are actively maintained. At larger scales, Chandra and XMM-Newton imaging reveals an extended nuclear X-ray source spatially coincident with the NSC, consistent with either starburst emission or an embedded low-luminosity active galactic nucleus (LLAGN; F. E. Bauer et al. [11]; D. S. S. Mak et al. [87]).

However, IC 342 lacks a direct, dynamical MBHM_{\mathrm{BH}} measurement. The only existing constraint comes from T. Böker et al. [18], who applied one-dimensional (1D) spherical Jeans models to aperture-integrated CO-bandhead spectroscopy of the NSC (σNSC=33±3\sigma_{\mathrm{NSC}} = 33 \pm3 km s−1^{-1}) and derived a dynamical NSC mass of ≈6×106 M⊙\approx6 \times10^{6}\ M_{\odot} together with an upper limit of MBH<5×105 M⊙M_{\mathrm{BH}} < 5 \times10^{5}\ M_{\odot} at their adopted distance of D=1.8D = 1.8 Mpc. Since both quantities scale linearly with distance, they become MNSC≈1.17×107 M⊙M_{\mathrm{NSC}} \approx1.17 \times10^{7}\ M_{\odot} and MBH<9.6×105 M⊙M_{\mathrm{BH}} < 9.6 \times10^{5}\ M_{\odot} at our adopted TRGB distance of D=3.45D = 3.45 Mpc (P.-F. Wu et al. 2014).

This Work

We present the first stellar-dynamical MBHM_{\mathrm{BH}} measurement in IC 342 using NIR IFS with Keck/OSIRIS observation assisted by the laser-guide-star AO (LGS-AO) system. With a point spread function (PSF) full width at half maximum (FWHM) of ≈0. ⁣′′053\approx0.\!^{\prime\prime}053–0. ⁣′′0620.\!^{\prime\prime}062 (§ 3.2), our data resolve the predicted SOI down to MBH∼105 M⊙M_{\mathrm{BH}} \sim10^{5}\ M_{\odot} at the adopted distance, enabling a direct, marginally resolved dynamical MBHM_{\mathrm{BH}} in this system. Combined with the well-characterized NSC properties (T. Böker et al. [15, 18]), this measurement directly probes the NSC–BH coexistence regime, tests whether the low-mass end of the MBHM_{\mathrm{BH}}–σe\sigma_{\mathrm{e}} relation extends to bulgeless Scd-type spirals, and provides a critical anchor for both the direct MBHM_{\mathrm{BH}} census in LTGs and theoretical models of CMO co-evolution in the secular-evolution regime.

This paper is structured as follows. § 3 details the Keck/OSIRIS observations, data reduction, and its PSF derivation. We extracted the nucleus stellar kinematics of IC 342 in § 4 and parametrize the galaxy mass model in § 5. § 6 presents the MBHM_{\mathrm{BH}} measurement via JAM dynamical modeling, and § 7 discusses its implications. We summarize our findings in § 8.

OSIRIS Observations and Data Reduction

IFS Observations and Data Reduction

We observed the nucleus of IC 342 with the OH Suppressing InfraRed Imaging Spectrograph (OSIRIS; J. Larkin et al. [80]) with the Keck I LGS-AO system (M. A. van Dam et al. (2006); P. L. Wizinowich et al. [143]). OSIRIS is a lenslet array IFU spectrograph with a 2048 ×\times 2048 Hawaii-2 detector and spectral resolution in the KK band of R∼3600R \sim3600. The observations were conducted on 15 Jan 2025, and consisted of two 900-second object frames, with an intervening 900-second sky frame (i.e., 30 min on object), in the standard object-sky-object sequence, at an airmass of 1.54. The plate scale was 0. ⁣′′020.\!^{\prime\prime}02, which at the distance of IC 342 (3.45 Mpc) equates to 0.32 pc pixel−1^{-1}. We observed with the KbbKbb filter (1965-2381 nm), which is ideally suited for CO bandhead observations.

Each object frame was reduced using the OSIRIS data reduction pipeline (DRP) version 6.0, using the sky frames for background subtraction. The DRP uses “Rectification Matrices” which incorporate all the flat fielding, bias and dark frames in one, along with mapping each pixel to a lenslet/spectral element. These matrices are retrieved from the Keck repository; for our data, these were dated Feb 2025. The resulting data cubes have fields of view of 0\farcs38 \times1\farcs28 (19 × 64 pixels).

The data cubes were centered and combined, then trimmed for display purposes, resulting in images with about 1040 spectra. The resulting data cube values were cleaned of spikes and geometric artifacts; these are mainly a product of the DRP. Telluric correction was performed with observation of the A0 star, HIP 12906 (magK_K = 7.77). The stellar spectrum is interpolated over the hydrogen absorption features and a representative blackbody is assumed. This is divided through the data cube spectra to remove the telluric absorption features. Flux calibration is performed by integrating the stellar spectrum over a clean region (away from telluric and stellar features), then matched to a reference Vega spectrum.

OSIRIS Point Spread Functions

Synthetic diffraction-limited PSF

The Keck/OSIRIS PSF at 2.3 μ\mum was modeled using the diffraction-limited routine osirispsf.pro from the OSIRIS Data Reduction Pipeline (DRP; K. E. Lockhart et al. [82]), which computes the PSF as the squared modulus of the Fourier transform of the complex telescope pupil wavefront. The pupil is constructed via the companion routine osirispupil.pro, encoding the Keck aperture geometry (primary mirror, secondary obscuration, and spider supports) at a physical plate scale du=λ/(Npix δθ)du = \lambda/(N_{\rm pix}\,\delta\theta), where λ\lambda is the effective wavelength, NpixN_{\rm pix} the padded pupil array size, and δθ\delta\theta the detector plate scale in radians per pixel. Zero-padding enforces two conditions: du<0.10du < 0.10 m pixel−1^{-1} (adequate pupil sampling) and Npix du>12N_{\rm pix}\,du > 12 m (full coverage of the 10 m aperture). A phase ramp is then applied for sub-pixel PSF centroid positioning before the wavefront is Fourier-transformed and rebinned to the native detector scale. For IC 342, we adopted δθ=20\delta\theta= 20 mas pixel−1^{-1} and λ=2.29\lambda= 2.29 μ\mum (blue edge of the CO bandhead fitting window). The resulting PSF image was normalized to unit flux and parameterized in MGE form using the mge_fit_sectors routine from the MgeFit package (M. Cappellari [21]); the best-fit parameters are listed in Table 1.

jj(Light fraction)j_jσj\sigma_j (arcsec)FWHMPSFtot\mathrm{FWHM}_{\mathrm{PSF}}^{\mathrm{tot}} (arcsec)
(1)(2)(3)(4)
Empirical PSF (standard star HIP 12906)
10.1480.0180.062
20.8520.036
Synthetic diffraction-limited PSF
10.9850.023
20.0070.1280.053
30.0081.405

Table 1. Keck/OSIRIS KbbK_{bb} PSF MGE models (pixel size = 0. ⁣′′020.\!^{\prime\prime}02).

Although this theoretical PSF omits residual wavefront errors and the extended AO halo, the Keck LGS-AO system delivers typical KK-band Strehl ratios of 30–40% under normal seeing conditions (M. A. van Dam et al. 2006), ensuring a core-dominated PSF that is adequate for dynamical modeling purposes.

Empirical PSF Constrained from the Standard Star HIP 12906

The empirical OSIRIS PSF was constructed from a dedicated standard star observation obtained immediately after the IC 342 science exposures, with the AO loops closed and identical instrument settings (20 mas pixel−1^{-1}, KbbKbb filter). We observed HIP 12906 with a short exposure time to avoid saturation; the PSF image was formed by collapsing the normalized, median-combined data cube over 2.050–2.360 μ\mum, spanning the spectral range used for stellar kinematic extraction and fully contained within the KbbKbb bandpass.

The PSF image was parameterized using the mge_fit_sectors routine of MGEFIT4 (M. Cappellari [21]); the resulting MGE model is reported in Table 1 and convolved with all JAM model predictions prior to comparison with the data (§ 6). The circularly-averaged radial profile of this empirical OSIRIS PSF and its best-fit MGE model are shown in Figure 1.

Figure 1. Circularly-averaged radial profile of the empirical Keck/OSIRIS Kbb PSF (open circles), constructed from the standard star HIP 12906 over 2.050–2.360 μ\mum, overlaid with its best-fit MGE model (red; Table 1). The flux is normalized and shown on a logarithmic scale.

Stellar Kinematics Measurements

Defining the Galaxy Center

Defining the photometric center of IC 342 requires caution because of the multi-layered dust environment along the line-of-sight (LOS). Optical HST images are unreliable for centroiding by the heavy foreground extinction (AF814W≈0.74A_{\mathrm{F814W}} \approx0.74 mag; E. F. Schlafly & D. P. Finkbeiner [123]) combined with patchy internal obscuration from the nuclear starburst ring. Archival HST/WFC3 IR imaging exists in several narrow- and medium-band filters, but each is unsuitable:

**Note—**Columns list: (1) Gaussian component index; (2) fractional contribution to the total light; (3) Gaussian dispersion along the major axis; (4) intensity-weighted mean FWHM of the PSF.

F128N, F164N, and F187N capture Paβ\beta, [Fe II] λ1.644 μm\lambda1.644\,\mu\mathrm{m}, and Paα\alpha emission, respectively, whose peaks are dominated by the two circumnuclear H II complexes located ∼4′′\sim4^{\prime\prime} east and west of the NSC ([15]), systematically offset from the stellar mass center. F127M and F153M are designed for planetary atmosphere studies, do not reliably trace galactic stellar continuum.

We therefore adopt the surface-brightness peak of the OSIRIS KbbK_{bb} white-light image as the coordinate origin, assuming the central BH coincides with the NIR continuum peak. At 2.3 μm2.3\,\mu\mathrm{m} (A2.3 μm≈0.15A_{2.3\,\mu\mathrm{m}} \approx0.15 mag; [123]), this image is effectively immune to both foreground and internal dust, achieves the highest angular resolution available, and is spatially registered to the science data by construction.

The nuclear centroid was determined as the intensity-weighted barycenter of the 120 brightest pixels within a 0. ⁣′′2×0. ⁣′′40.\!^{\prime\prime}2 \times0.\!^{\prime\prime}4 aperture of the OSIRIS KbbK_{bb} white-light image. Reassuringly, this photometric peak coincides to within a single spaxel with both the central peak of the velocity-dispersion map and the zero-crossing of the stellar rotation field, so that the photometric, dispersion, and kinematic centers are mutually consistent. We adopt this common location as the dynamical center and the coordinate origin for all subsequent LOSVD maps, mass models, and figures. We anchor on the OSIRIS continuum rather than the HST imaging because at 2.3 μm2.3\,\mu\mathrm{m} the NIR peak is largely immune to the dust-induced centroid shifts affecting the obscured optical and near-IR HST images.

PPXF Setup & The Optimal Stellar Template

Stellar kinematics of the IC 342 nucleus were extracted from the Keck/OSIRIS KbbK_{bb}-band data cube via the Penalized PiXel-Fitting method (PPXF5; M. Cappellari & E. Emsellem [28]; M. Cappellari [23, 25]), fitting the full KbbK_{bb} spectral range of 2.050–2.360 μm\mu\mathrm{m} and thereby including the CO bandhead absorption features at the red end of the bandpass. Following the results of D. D. Nguyen et al. [107, 110], who found that LOSVDs derived from the empirical X-shooter Spectral Library (XSL6; K. V. Verro et al. [140]) agree within 3% with those from the higher-resolution PHOENIX synthetic library (T.-O. Husser et al. [74]), we adopt XSL as our primary template source. The XSL provides broad spectral coverage (3000–25,000 Å) at R∼10,000R \sim10{,}000, encompassing a diverse range of O–M and evolved stars.

To account for the mixed stellar populations in the IC 342 nucleus and mitigate template-mismatch bias ([88, 130]), we derived a fixed optimal XSL template from the full DR3 library of 830 spectra from 683 stars and from a large spatial region’s integrated spectrum, namely the global spectrum. The global spectrum was formed by co-adding all spaxels within the red aperture in the left panel of Figure 2, achieving S/N ∼240\sim240 per spectral pixel. It was then logarithmically rebinned to a constant velocity scale of 32 km s−1^{-1} pixel−1^{-1} following Eq. 8 of M. Cappellari [23], and the XSL templates were broadened to the instrumental resolution by convolution with a Gaussian kernel whose dispersion was computed via Eq. 5 of M. Cappellari [23]. The fixed optimal XSL template was subsequently derived by fitting these broadened templates to the global spectrum with PPXF under a Gaussian LOSVD (moments = 2) to obtain the rotational velocity VV (w.r.t. the systemic velocity VsysV_{\mathrm{sys}} of the nucleus) and velocity dispersion σ\sigma as our primary kinematic quantities.

Figure 2. Left: Keck/OSIRIS KbbK_{bb}-band white-light image of the IC 342 nucleus, constructed by collapsing the data cube along the spectral axis (logarithmic stretch); white contours trace the stellar continuum in steps of 1 mag arcsec−2^{-2}. Right: Global PPXF fit to the nuclear spectrum obtained by co-adding all spaxels within the red aperture (left panel). The observed spectrum (black) and best-fit XSL template convolved with the optimal LOSVD (red) are shown over the spectral range of 2.050–2.360 μ\mum; fit residuals are plotted as green dots.

Since additive Legendre polynomials can alter the depth of the CO bandhead absorption features and bias the recovered LOSVDs in the KbbK_{bb}-band OSIRIS data, we disabled them in all PPXF fits (degree = −1-1). Multiplicative polynomials (mdegree) of orders 0–4 were tested to correct for residual continuum shape, sky subtraction, and calibration systematics; the recovered kinematics stabilize and become insensitive to polynomial order at mdegree ≥2\ge2 across the full fitting window (2.050–2.360 μm\mu\mathrm{m}), so we adopt mdegree = 2 as our fiducial choice to ensure stable LOSVD recovery while minimizing overfitting.

This fitting process produced a fixed optimal stellar template, defined as a linear combination of 13 giant-star spectra that faithfully reproduce the full wavelength range of the OSIRIS spectra (Figure 2, right panel). The fixed optimal template is dominated by four giant stars: BMB 106 (contributing 50% of the flux), HD 111631 (11%), [W65]c2 (10%), and IRAS 09484-6242 (5%). These four stars alone provide a relatively good fit to the global spectrum; however, the remaining nine stars contribute an additional 34% of the flux. The properties and fractional flux contributions of all 13 XSL template stars are listed in Table 2. The global best-fit kinematics are V=34±2V = 34 \pm2 km s−1^{-1} (heliocentric) and σ=41±2\sigma= 41 \pm2 km s−1^{-1}; uncertainties are the 1σ1\sigma scatter from 200 Monte Carlo realizations (M. Cappellari & E. Emsellem [28]). While this OSIRIS heliocentric velocity is consistent

No.Star NameXSL IDSp. TypeTeffT_{\mathrm{eff}} (K)log⁡g\log gOSIRIS
(1)(2)(3)(4)(5)(6)(7)
1BMB 106X0169M632700.400.50
2HD 111631X0903M0V38884.560.11
3[W65] c2X0591C––0.10
4IRAS 09484−6242X0860R––0.05
5[W65] c2X0609C––0.05
6HIP 75423X0684M2V34904.780.04
7SHV 0518161−683543X0530C40780.200.04
8SHV 0606101−724012X0121M733370.120.04
9HD 57060X0116O7 Iafpvar318403.410.02
10HD 46375BMD014M2.0V––0.02
11BMB 289X0155M9––0.01
12ISO-MCMS J005332.4−730501X0031–43911.050.01
13SHV 0542111−683837X0810C40800.470.01

Table 2. Optimal XSL Template Stars for IC 342. Note—(1) Order number; (2) Star name; (3) XSL DR3 identifier (K. Verro et al. [140]); (4) Spectral type; (5) Effective temperature; (6) Surface gravity; (7) Fractional contribution to the total light from the optimal template fit to the global OSIRIS spectrum.

with the NED7 value, the central velocity dispersion measured here exceeds that of T. Böker et al. [18] by a factor of ∼1.2\sim1.2. This offset is expected: T. Böker et al. [18] reported an aperture-integrated σ\sigma over the NSC using seeing-limited long-slit spectroscopy on the NASA Infrared Telescope Facility (IRTF), whereas our OSIRIS measurements resolve the central few parsec individually, where the gravitational influence of any central dark mass elevates the local velocity dispersion above the NSC-integrated value. The elevated central σ\sigma is therefore consistent with (and indeed motivates) our dynamical search for a central dark mass in IC 342.

Two-Dimensional Stellar Kinematic Maps

Two-dimensional stellar LOSVDs across the IC 342 nucleus were extracted by first adaptively binning spaxels with PowerBin8 ([26]) to a target S/N of 150 per spectral pixel. Spaxels within r_{\rm unbin}=0\farcs05 (close to one PSF FWHM of OSIRIS element) were left unbinned to preserve the spatial resolution critical for kinematic measurements in the innermost nuclear region, where individual spaxel S/N exceeds 90 per spectral pixel; spaxels at r>runbinr>r_{\rm unbin} were grouped adaptively, yielding N=70N=70 bins in total. Each resulting spectrum was logarithmically rebinned in wavelength and fitted with PPXF (M. Cappellari [23]), using the fixed optimal XSL template broadened to the instrumental resolution with second-degree multiplicative Legendre polynomials to correct for residual continuum shape differences, sky subtraction, and calibration systematics (following § 4.2; disable additive Legendre polynomials and set moments = 2), recovering the stellar LOS velocity VV, dispersion σ\sigma, and their associated uncertainties.

The resulting kinematic maps are presented in Figure 3: the logarithmic integrated intensity (panel A), a representative PPXF fit to a central spaxel (panel B), and 2D maps of VV (panel C), σ\sigma (panel D), and root-mean-squared velocity V2+σ2\sqrt{V^2+\sigma^2} (panel E). The systemic heliocentric velocity, derived from the OSIRIS velocity map, is vsys=35±2v_{\mathrm{sys}} = 35 \pm2 km s−1^{-1}, which is 4 km s−1^{-1} larger than the NED value. After subtracting vsysv_{\mathrm{sys}}, the nucleus exhibits a rotation reaching ∣V∣≈10±2|V| \approx10 \pm2 km s−1^{-1} at the edge of the OSIRIS FOV.

Figure 3. Keck/OSIRIS KbbK_{\rm bb}-band view of the IC 342 nucleus from stellar kinematic extraction. All panels are restricted to the spatial region enclosed by the red aperture in Figure 2, where spectra are free from artificial detector artifacts. Panel A: Nuclear white-light image obtained by spectrally collapsing the OSIRIS data cube (logarithmic intensity scale, same as left panel of Figure 2). Panel B: Representative PPXF spectral fit at the spaxel marked in Panel A. The OSIRIS spectrum (black) is shown across the spectral range of 2.050–2.360 μ\mum, with the best-fit fixed optimal XSL template overplotted in red; green points show fit residuals, and vertical gray lines indicate spectral regions excluded from the fit. Panels C–E: 2D stellar kinematic maps of the LOS velocity VV, velocity dispersion σ\sigma, and Vrms≡V2+σ2V_{\rm rms} \equiv\sqrt{V^2+\sigma^2}, with white contours overlaid at 1 mag arcsec−2^{-2} intervals tracing the stellar continuum morphology.

We measure a global kinematic position angle of PAkin=38.4∘±1∘\mathrm{PA}_{\mathrm{kin}} = 38.4^\circ\pm1^\circ, consistent with the kinematic position angle of PA=37∘\mathrm{PA} = 37^\circ derived from the H I velocity field by L. P. Crosthwaite et al. (2000). Both vsysv_{\mathrm{sys}} and PAkin\mathrm{PA}_{\mathrm{kin}} were derived from the velocity map using the PAFIT9 package (D. Krajnović et al. [78]). The σ\sigma profile, which dominates over VV, rises from ∼42±1.3\sim42 \pm1.3 km s−1^{-1} at r≳0.2′′r \gtrsim0.2^{\prime\prime}, to a central peak of ∼49.0±1.3\sim49.0 \pm1.3 km s−1^{-1}. This central increase in both σ\sigma and VrmsV_{\mathrm{rms}} provides strong evidence for the presence of a central compact dark mass. We present these stellar-kinematic measurements in Table 3.

Δ\DeltaR.A.Δ\DeltaDecl.VVΔV\Delta Vσ\sigmaΔσ\Delta\sigma
(arcsec)(arcsec)(km s−1^{-1})(km s−1^{-1})(km s−1^{-1})(km s−1^{-1})
−0.018-0.0180.0120.0121.111.111.211.2149.4949.491.261.26
0.0020.002−0.008-0.0080.510.511.111.1149.0449.041.151.15
0.0220.0220.0320.0322.312.311.111.1148.8848.881.151.15
0.0220.022−0.008-0.0080.140.141.081.0848.4748.471.121.12
0.0220.0220.0120.0121.651.651.161.1648.4148.411.201.20
0.0020.0020.0120.0121.311.311.071.0748.2948.291.111.11
0.0020.0020.0320.0322.772.771.081.0848.1748.171.121.12
..................

Table 3. Keck/OSIRIS Kinematic Data of the IC 342 Nucleus. (This table is available in its entirety in machine-readable form)

No Evidence of AGN in the Keck/OSIRIS Kbb IFU

Following recent VLT/SINFONI (M. Cappellari et al. [29]), VLT/MUSE (D. A. Simon et al. [131]), and JWST/NIRSpec (D. D. Nguyen et al. [110]) studies, we checked for AGN continuum dilution of the CO bandhead features, which can bias the recovered σ\sigma and mimic spurious stellar population or M/LM/L variations in the nucleus. We quantified this using the stellar line-strength parameter γ≡stars/(stars+AGN)\gamma\equiv\mathrm{stars}/(\mathrm{stars}+\mathrm{AGN}) ([29] (R. P. van der Marel & M. Franx 1993)), which approaches unity where AGN contamination is negligible and drops below the unity toward the center if an AGN continuum is present. We find γ≈1\gamma\approx1 across the entire OSIRIS FOV, indicating no detectable AGN contribution in the nucleus of IC 342 in the NIR wavelength; consequently, unlike D. D. Nguyen et al. [110]; D. A. Simon et al. [131], no AGN subtraction is applied to the galaxy photometric model.

This conclusion is further supported by multiwavelength observations of the IC 342 nucleus. The central X-ray source detected by Chandra and XMM-Newton has a luminosity of LX≲1039L_X \lesssim10^{39} erg s−1^{-1} (D. S. S. Mak et al. [87]; C. Zhang et al. [146]), well within the regime expected for normal star-forming galaxies and consistent with high-mass X-ray binary (HMXB) populations associated with the nuclear starburst. The source is also spatially extended rather than point-like, arguing against a well-defined AGN (D. S. S. Mak et al. [87]). Similarly, the extended radio emission from the IC 342 nucleus is fully consistent with starburst-driven synchrotron emission rather than AGN-powered jets (D. S. Balser et al. [10]). We therefore conclude that any AGN contribution to the nuclear emission of IC 342 is negligible, and that the γ≈1\gamma\approx1 recovered from our OSIRIS data reflects genuine stellar-continuum dominance rather than AGN dilution.

Galaxy Photometric Model

The HST/WFC3/UVIS F814W surface brightness profile of IC 342 was modeled using archival imaging (Proposal ID 12163; PI: Barth) with a pixel scale of 0.040.04 pixel−1^{-1}. This image was adopted for the structural analysis because it provides the highest available spatial resolution (PSF FWHM =0.069′′= 0.069^{\prime\prime}) while being less affected by the galaxy’s patchy internal dust than shorter-wavelength optical images. Figure 4 shows the galaxy light at 0.814 and 1.53 μ\mum. Dust lanes and heavily obscured regions were masked using SAOIMAGE DS9 prior to the isophote analysis, which used the mskregions task in IRAF.

Nucleus of IC 342 in WFC3/IR F153M and HST/WFC3/UVIS F814W

Figure 4. Nucleus of IC 342. The central dense nuclear star cluster is surrounded by a somewhat ill-formed and knotty ∼3\sim3–4′′4^{\prime\prime} (∼48\sim48–6464 pc) nuclear ‘disk’. (Images from HST Proposal ID 12163. PI: Barth.)

We use a 1D profile rather than a full 2D MGE fit to average the light distribution, thereby smoothing out the uneven, non-axisymmetric features from the patchy dust. The isophotes were fit using the IRAF tasks Isofit and Cmodel (B. C. Ciambur [33]), while allowing the ellipticity and position angle to vary with radius but keeping the photometric centre fixed at the kinematic centre (§ 4.1). This constraint was imposed because patchy dust extinction produces apparent shifts in the isophotal centroids with radius that would otherwise bias a fully free-geometry fit. The resulting F814W surface brightness profile, shown in Figure 5, was calibrated to the Vega magnitude system using a WFC3/UVIS F814W photometric zero-point of 24.747 mag.

Four-component surface-brightness decomposition of the IC 342 nucleus

Figure 5. Four-component (three Sérsic + exponential disk) surface brightness decomposition of the IC 342 nucleus from HST/WFC3 UVIS F814W imaging. Upper panel: The observed major-axis profile (filled circles) is overlaid with the best-fit total model (solid line), decomposed into an inner Sérsic component tracing the NSC (dashed line), an outer Sérsic component describing the central stellar distribution (dash-dotted line), and an exponential disk (dotted line). Lower panel: Fit residuals (data−model) in mag arcsec−2^{-2}.

We performed a decomposition of the geometric-mean and major-axis surface brightness profile using a Python 3 version of the Profiler code (B. C. Ciambur [34])10. The light was modeled with four Sérsic components. During the fitting process, each intrinsic Sérsic profile was first convolved with the image PSF, and the convolved profiles were then summed to reproduce the observed (PSF-blurred) light profile.

The best-fitting major-axis model comprises: an inner Sérsic component tracing the NSC with μe,1=12.31\mu_{e,1} = 12.31 mag arcsec−2^{-2}, r_{e,1} = 0\farcs14 (≈2.34\approx2.34 pc), and n1=2.37n_1 = 2.37; a second Sérsic component with μe,2=15.44\mu_{e,2} = 15.44 mag arcsec−2^{-2}, r_{e,2} = 0\farcs63, and n2=0.12n_2 = 0.12 tracing the puffy stars and star cluster around the NSC; an immediate Sérsic component with μe,3=16.87\mu_{e,3} = 16.87 mag arcsec−2^{-2}, r_{e,3} = 2\farcs05, and n3=0.10n_3 = 0.10 that captures what may likely be considered a nuclear disk in less well-resolved galaxies located at greater distances (M. Balcells et al. [9]). The large-scale disk is described by the outer Sérsic component, which is consistent with an exponential component with μ0=15.90\mu_0 = 15.90 mag arcsec−2^{-2} and h = 2\farcs93 (≈49\approx49 pc). Fitting the geometric-mean axis yields magnitudes of 13.92 and 14.66 mag for the inner and middle components, respectively, and thus the inner component makes up 66.4 percent of the combined F814W flux of these two components.

Our four-component decomposition yields NSC parameters that differ substantially from the two-component GALFIT fit (i.e., Sérsic + exponential) of D. J. Carson et al. [31], who report only for the central excess (NSC), n=3.79−0.40+0.48n = 3.79^{+0.48}_{-0.40} and re=1.38−0.13+0.19r_e = 1.38^{+0.19}_{-0.13} pc from the same F814W image but fitting just the inner 1\farcs3. While the effective radii agree by ∼\sim70% (2.34 vs. 1.38 pc), the Sérsic indices differ by ∼\sim60% (2.37 vs. 3.79). Discrepancy is expected given D. J. Carson et al., their Figure 6 adopted a single Sérsic plus exponential disk model, so their model’s inner component absorbs flux from the extended inner stellar envelope (r \sim0\farcs4–0\farcs8), that our fit separates as an independent component from the dense NSC, inflating both nn and the apparent magnitude relative to the innermost component of our model. Indeed, the summed magnitude (mF814W≈13.47m_{\mathrm{F814W}} \approx13.47 mag) of our inner two Sérsic components, coming from the fit to the geometric-mean radial profile, and thus taking the ellipticity profile into account, agrees to within 0.04 mag with the D. J. Carson et al. value (mF814W=13.43m_{\mathrm{F814W}} = 13.43 mag). However, not separating these two features, evident in the image (Figure 4), forces the model with a single inner Sérsic function to have a high-nn tail in an effort to capture the excess flux at intermediate radii (r \sim0\farcs4–0\farcs8), in turn demanding a high central concentration/spike at radii well below the image resolution. Not only does it yield a stronger amplitude in the snake-like residual light profile, but it impacts the optimal extraction of the BH mass.

Figure 6. 1D stellar kinematics extracted from the OSIRIS KbbK_{\mathrm{bb}} data cubes of IC 342, shown along the major axis (red, offset by +5+5 km s−1^{-1} for clarity) and the minor axis (blue), overlaid with the best-fit JAMcyl_{\mathrm{cyl}} (top) and JAMsph_{\mathrm{sph}} (bottom) models, for the two PSF models (FWHM = 0. ⁣′′0530.\!^{\prime\prime}053, left; 0. ⁣′′0620.\!^{\prime\prime}062, right), and with the MBH=0M_{\mathrm{BH}}=0 model shown as the grey dashed curve.

We caution that while our second and third Sérsic components broadly capture the complex morphology seen at optical wavelengths, their parameters should not be over-interpreted. Rather, their inclusion enabled us to better isolate and quantify the innermost component, required for the JAM dynamical modeling. Beyond ∼\sim3, the decline in the light profile shown in Figure 5 is somewhat due to dust (see Figure 4).

Our best-fit Sérsic functions were converted into an MGE model using the mge_fit_1d routine (M. Cappellari [21]), applied to the circularized 1D major-axis profile sampled on a logarithmic radial grid. The observed axial ratio qj′q'_j of each Gaussian component was assigned from the ellipticity measured at the corresponding photometric scale, as required by the JAM dynamical models (§ 6). The resulting MGE model is listed in Table 4.

jjΣ⋆,j\Sigma_{\star,j} (L⊙_\odot pc−2^{-2})σj\sigma_j (arcsec)qj′=bj/ajq'_j=b_j/a_j
(1)(2)(3)(4)
129861430.0250.94
214086770.0570.94
34691080.1260.94
4802650.4280.94
5170841.7650.94
638254.8790.94

Table 4. HST/WFC3 UVIS F814W MGE model. Notes: Columns list: (1) Gaussian component index; (2) Luminosity surface density; (3) Gaussian dispersion along the major axis; (4) Observed axial ratio

Dynamical Modelling

Jeans Anisotropic Models

The observed root-mean-squared velocity (V2+σ2)(\sqrt{V^2+\sigma^2}) maps of IC 342 were modeled with both the cylindrically aligned (JAMcyl_{\mathrm{cyl}}; M. Cappellari [22]) and spherically aligned (JAMsph_{\mathrm{sph}}; M. Cappellari [24]) implementations of JAM, each under the assumption of an axisymmetric gravitational potential. The two implementations bracket the range of physically plausible velocity-ellipsoid orientations and thereby quantify how this modeling choice affects the inferred MBHM_{\mathrm{BH}}.

In JAMcyl_{\mathrm{cyl}}, the velocity ellipsoid is tied to cylindrical coordinates (R,ϕ,z)(R,\phi,z), with anisotropy βz≡1−(σz/σR)2\beta_z \equiv1-(\sigma_z/\sigma_R)^2 (σz≠σR=σϕ\sigma_z \ne\sigma_R=\sigma_\phi); in JAMsph_{\mathrm{sph}}, it is tied to spherical coordinates (r,θ,ϕ)(r,\theta,\phi), with anisotropy βr≡1−(σθ/σr)2\beta_r \equiv1-(\sigma_\theta/\sigma_r)^2 (σr≠σθ=σϕ\sigma_r \ne\sigma_\theta=\sigma_\phi). Both are computed through the jam_axi_proj routine of JamPy (M. Cappellari [24]) via align=‘cyl’ or align=‘sph’, producing the projected second-moment map ⟨Vlos2⟩1/2≡V2+σ2\langle V_{\mathrm{los}}^2\rangle^{1/2}\equiv\sqrt{V^2+\sigma^2}. For JAMcyl_{\mathrm{cyl}}, the intrinsic second moments ⟨σz2⟩\langle\sigma_z^2\rangle and ⟨σϕ2⟩\langle\sigma_\phi^2\rangle are approximated numerically to accelerate model evaluation; JAMsph_{\mathrm{sph}} instead computes the BH gravitational potential exactly without approximation.

The four free parameters common to both models are: MBHM_{\mathrm{BH}}, sampled logarithmically; and three linear parameters: M/LF814WM/L_{\mathrm{F814W}}, the inclination ii (converted internally to the minimum intrinsic axial ratio qmin⁡q_{\min}), and the orbital anisotropy (reported as dispersion ratios σz/σR\sigma_z/\sigma_R and σθ/σr\sigma_\theta/\sigma_r for JAMcyl_{\mathrm{cyl}} and JAMsph_{\mathrm{sph}}, respectively). In both JAMcyl_{\mathrm{cyl}} and JAMsph_{\mathrm{sph}}, M/LF814WM/L_{\mathrm{F814W}} is parameterized to vary spatially as a linear function of radius (Section 5.2.1) rather than being held constant. For the orbital anisotropy, JAMcyl_{\mathrm{cyl}} instead assumes a spatially uniform value, whereas JAMsph_{\mathrm{sph}} adopts a logistic function of radius (§ 6.2.2). Search ranges and initial guesses for all free parameters are given in Table 5.

Parameter (Unit)Search rangeJAMcyl_{\mathrm{cyl}} Best-fitting value (3)JAMcyl_{\mathrm{cyl}} 1σ\sigma Error (16–84)% (4)JAMcyl_{\mathrm{cyl}} 3σ\sigma Error (0.14–99.86)% (5)JAMsph_{\mathrm{sph}} Best-fitting value (3)JAMsph_{\mathrm{sph}} 1σ\sigma Error (16–84)% (4)JAMsph_{\mathrm{sph}} 3σ\sigma Error (0.14–99.86)% (5)
Empirical PSFFWHMPSFtot=0. ⁣′′062_{\mathrm{PSF}}^{\mathrm{tot}} = 0.\!^{\prime\prime}062JAMcyl_{\mathrm{cyl}}JAMsph_{\mathrm{sph}}
log⁡10(MBH/M⊙)\log_{10}(M_{\mathrm{BH}}/M_{\odot})(0→7)(0 \to7)5.32−0.08,+0.07-0.08,+0.07−0.42,+0.18-0.42,+0.185.38−0.15,+0.13-0.15,+0.13−0.40,+0.37-0.40,+0.37
M/LF814W0M/L_{F814W}^{0} (M⊙/L⊙M_{\odot}/L_{\odot})(0.1→3)(0.1 \to3)0.09−0.02,+0.02-0.02,+0.02−0.05,+0.06-0.05,+0.060.07−0.02,+0.02-0.02,+0.02−0.05,+0.06-0.05,+0.06
M/LF814W0. ⁣′′15M/L_{F814W}^{0.\!^{\prime\prime}15} (M⊙/L⊙M_{\odot}/L_{\odot})(0.1→3)(0.1 \to3)0.38−0.01,+0.01-0.01,+0.01−0.03,+0.03-0.03,+0.030.36−0.01,+0.02-0.01,+0.02−0.03,+0.04-0.03,+0.04
(σz/σR)cyl(\sigma_z/\sigma_R)_{\mathrm{cyl}}(0.1→2)(0.1 \to2)0.98−0.03,+0.03-0.03,+0.03−0.08,+0.10-0.08,+0.10–––
(σθ/σr)0(\sigma_{\theta}/\sigma_r)_0(0.5→1)(0.5 \to1)–––0.92−0.10,+0.06-0.10,+0.06−0.38,+0.08-0.38,+0.08
(σθ/σr)∞(\sigma_{\theta}/\sigma_r)_{\infty}(1→1.3)(1 \to1.3)–––1.16−0.10,+0.10-0.10,+0.10−0.16,+0.14-0.16,+0.14
Diffraction-limited PSFFWHMPSFtot=0. ⁣′′053_{\mathrm{PSF}}^{\mathrm{tot}} = 0.\!^{\prime\prime}053JAMcyl_{\mathrm{cyl}}JAMsph_{\mathrm{sph}}
log⁡10(MBH/M⊙)\log_{10}(M_{\mathrm{BH}}/M_{\odot})(0→7)(0 \to7)5.23−0.09,+0.07-0.09,+0.07−0.27,+0.20-0.27,+0.205.26−0.16,+0.15-0.16,+0.15−0.39,+0.37-0.39,+0.37
M/LF814W0M/L_{F814W}^{0} (M⊙/L⊙M_{\odot}/L_{\odot})(0.1→3)(0.1 \to3)0.10−0.01,+0.02-0.01,+0.02−0.04,+0.04-0.04,+0.040.09−0.02,+0.02-0.02,+0.02−0.06,+0.04-0.06,+0.04
M/LF814W0. ⁣′′15M/L_{F814W}^{0.\!^{\prime\prime}15} (M⊙/L⊙M_{\odot}/L_{\odot})(0.1→3)(0.1 \to3)0.38−0.01,+0.01-0.01,+0.01−0.02,+0.02-0.02,+0.020.35−0.01,+0.01-0.01,+0.01−0.03,+0.04-0.03,+0.04
(σz/σR)cyl(\sigma_z/\sigma_R)_{\mathrm{cyl}}(0.1→2)(0.1 \to2)0.96−0.02,+0.03-0.02,+0.03−0.07,+0.08-0.07,+0.08–––
(σr/σθ)0(\sigma_r/\sigma_{\theta})_0(0.5→1)(0.5 \to1)–––0.93−0.10,+0.05-0.10,+0.05−0.28,+0.07-0.28,+0.07
(σr/σθ)∞(\sigma_r/\sigma_{\theta})_{\infty}(1→1.3)(1 \to1.3)–––1.16−0.09,+0.09-0.09,+0.09−0.16,+0.13-0.16,+0.13

Table 5. JAM models best-fitting parameters and formal uncertainties.

In total, we perform four JAM model runs to bracket the main sources of systematic uncertainty in our dynamical analysis. These span two choices of velocity-ellipsoid alignment (e.g., JAMcyl_{\mathrm{cyl}} and JAMsph_{\mathrm{sph}}) and two representations of the OSIRIS PSFs in KbbK_{\mathrm{bb}} band (e.g., empirical or diffraction-limited) as characterized in § 3.2 and listed in Table 1.

Inclination Prior and MGE Deprojection Degeneracy

The photometric deprojection from a 2D surface brightness distribution into a 3D intrinsic mass density is formally non-unique for any galaxy not viewed precisely edge-on (G. B. Rybicki 1987; {citet}J. Magorrian 1999). In the MGE formalism, one particular deprojection is obtained analytically for each assumed inclination $i$ ({citet}M. Cappellari 2002), but the resulting intrinsic density can vary substantially across the family of solutions consistent with the observed surface brightness when $i$ is small ({citet}P.-Y. Lablanche et al. 2012). This degeneracy is particularly severe for nearly face-on systems: {citet}M. D. Smith et al. [132] and {citet}P. Dominiak et al. [43] demonstrated for NGC 524 ($i\lesssim40^\circ$) and NGC 1387 ($i\lesssim13.6^\circ$), respectively, that low inclinations introduce strongly asymmetric posteriors and inflate the uncertainties on both $M_{\mathrm{BH}}$ and $M/L$, while {citet}P.-Y. Lablanche et al. [79]` showed explicitly that JAM models of a simulated galaxy at i=25∘i=25^\circ, can produce intrinsic density distributions that are visually indistinguishable in projection yet physically very different in 3D.

IC 342 is nearly face-on at i≈31∘i\approx31^\circ (L. P. Crosthwaite et al. 2000), placing it squarely in this problematic regime. Treating ii as a free parameter in the JAM fit would therefore not constrain the inclination from the stellar kinematics alone, but would instead allow the MCMC sampler to explore a wide range of degenerate 3D mass distributions, inflating the uncertainties on MBHM_{\mathrm{BH}} without providing additional physical insight. We therefore fixed i=31∘i=31^\circ throughout all JAM model variants, adopting the kinematic inclination measured from the HI velocity field by L. P. Crosthwaite et al. (2000), which is well constrained independently of the stellar dynamics and represents the most reliable available inclination estimate for this galaxy.

Linearly Varying M/LF814WM/L_{\mathrm{F814W}} Profile

While the heavy foreground extinction toward IC 342 (AF814W≈0.74A_{\mathrm{F814W}}\approx0.74 mag) precludes verifying any radial stellar population variation with optical data, our JAM models consistently disfavor a spatially constant M/LF814WM/L_{\mathrm{F814W}} when fitting the OSIRIS stellar kinematics. We therefore adopt a radially varying M/LF814WM/L_{\mathrm{F814W}} profile that increases linearly from a central value M/LF814W0M/L_{\mathrm{F814W}}^0 to an outer value M/L_{\mathrm{F814W}}^{0\farcs15} at r=0\farcs15, beyond which M/LF814WM/L_{\mathrm{F814W}} is held fixed at M/L_{\mathrm{F814W}}^{0\farcs15}. The transition radius of 0\farcs15 corresponds to the shorter dimension of the OSIRIS FOV, within which the 2D stellar kinematics provide the tightest spatial constraints on the M/LF814WM/L_{\mathrm{F814W}} gradient. See § 6.5.2 for more details.

In practice, this profile is implemented in JAM by assigning a distinct (M/LF814W)j(M/L_{\mathrm{F814W}})_j to each Gaussian component of dispersion σj\sigma_j in the MGE model (Table 4):

Υj={γ0′′+γ0′′15−γ0′′0′′15σj,σj<0′′15γ0′′15,σj≥0′′15(1)\Upsilon_j = \begin{cases} \gamma^{0\prime\prime} + \dfrac{\gamma^{0\prime\prime15} - \gamma^{0\prime\prime}}{0\prime\prime15}\sigma_j, & \sigma_j < 0\prime\prime15 \\ \gamma^{0\prime\prime15}, & \sigma_j \ge0\prime\prime15 \end{cases} \tag*{(1)}

with γ=M/LF814W\gamma= M/L_{\mathrm{F814W}}.

Logistic Anisotropy for JAMsph Models

The radial anisotropy profile we adopt for the JAMsph_{\mathrm{sph}} models is motivated by a now well-established empirical pattern: stellar orbits across galaxies of very different masses and morphologies tend to settle into a tangentially biased configuration within the BH’s SOI, while remaining closer to isotropic or mildly radial further out (M. Cappellari [27], their Figure 10). The leading physical explanation for this pattern in massive ellipticals invokes binary SMBH scouring during dry mergers, a process that preferentially depletes radial orbits near the center (e.g., [91, 115]); the fact that a similar tangential bias appears even in lower-mass, fast-rotating systems with no comparable merger history suggests the behavior may be a more general feature of galactic nuclei rather than a merger relic specifically. Rather than treating this as a qualitative expectation only, we encode it directly as a prior on the model, following the implementation of D. A. Simon et al. [131], in which the anisotropy βr(r)≡1−σθ2/σr2\beta_r(r) \equiv1 - \sigma_\theta^2/\sigma_r^2 varies with log⁡r\log r according to a logistic function:

βr(r)=βr,0+βr,∞−βr,01+(ra/r)α(2)\beta_r(r) = \beta_{r,0} + \frac{\beta_{r,\infty} - \beta_{r,0}}{1 + (r_a/r)^\alpha} \tag*{(2)}

with βr,0\beta_{r,0} and βr,∞\beta_{r,\infty} setting the inner and outer anisotropy plateaus and rar_a marking the radius of transition between them; for ease of interpretation, we report this throughout as the equivalent dispersion ratio σr/σθ=(1−βr)−1/2\sigma_r/\sigma_\theta= (1-\beta_r)^{-1/2}.

The value of imposing this particular functional form rests on its track record: D. A. Simon et al. [131], D. D. Nguyen et al. [110], and D. D. Nguyen et al. [111] applied the identical prescription to M87, NGC 4258, and NGC 3031, respectively, and recovered MBHM_{\mathrm{BH}} values in each case consistent with independent literature measurements within their reported uncertainties. These three systems span a substantial range in morphology, from the giant elliptical M87 to the spirals NGC 4258 and NGC 3031, providing a reasonably broad test of the prescription’s validity prior to its application here. We adopt the same prior ranges here on that basis, fixing α=2\alpha= 2 to keep the parameter space manageable and bounding the inner and outer dispersion ratios to 0.5<(σr/σθ)0<10.5 < (\sigma_r/\sigma_\theta)_0 < 1 and 1<(σr/σθ)∞<1.31 < (\sigma_r/\sigma_\theta)_\infty< 1.3, consistent with the tangentially-biased-to-isotropic transition described above. The transition radius itself is held fixed at ra=0′′09r_a = 0\prime\prime09, approximately matching the wavelength-stable effective radius of the nuclear cluster reported by D. J. Carson et al. [31]. This value sits outside the SOI of the IC 342 BH (RSOI≈0′′05R_{\mathrm{SOI}} \approx0\prime\prime05; § 6.4) and the OSIRIS PSF FWHM so that the logistic curve has room to complete its transition from the inner tangential regime to the outer, more isotropic one within the radial range actually constrained by the data.

MCMC Fitting and Uncertainty Estimation

Posterior distributions for the JAM parameters were generated by Markov chain Monte Carlo (MCMC) sampling, specifically the adaptive Metropolis algorithm of H. Haario et al. [66] as implemented in ADAMet.11 (M. Cappellari et al. [30]), running each chain for 5×1045 \times10^4 steps and discarding the first 20% as burn-in. We take the maximum-likelihood solution as the best-fit value for each parameter, with 1σ1\sigma and 3σ3\sigma confidence levels (CLs) read off the marginalized posteriors at the 16th–84th and 0.14th–99.86th percentiles, respectively.

A subtlety of fitting such a large number of high-precision OSIRIS kinematic bins is that the resulting parameter uncertainties can end up unrealistically small if any systematics remain unaccounted for in the model, and that the χ2\chi^2 can become dominated by the sheer number of bins at large radii rather than by the comparatively few bins inside the SOI where the BH actually dominates the gravitational force on the kinematics. In the conventional approach, both issues can be mitigated using the Δχ2\Delta\chi^2 inflation scheme introduced by R. C. E. van den Bosch & G. van de Ven (2009), which in a Bayesian framework corresponds to rescaling the kinematic uncertainties by a factor of (2N)1/4(2N)^{1/4} (e.g., [92, 103, 104, 97]). To keep the central kinematics fully informative on MBHM_{\mathrm{BH}}, the inflation is applied only to bins outside the BH SOI, while bins within the SOI retain their formal uncertainties. This selective-inflation strategy has become common practice in both stellar- and gas-dynamical MBHM_{\mathrm{BH}} measurements (e.g., D. A. Drehmer et al. 2015; S. Thater et al. 2022; D. D. Nguyen et al. 2026b,c; D. A. Simon et al. 2024), and has been adopted in recent analyses (T H. T. Ho et al. submitted; H. N. Ngo et al. submitted).

For IC 342, however, we deliberately choose not to apply the selective-inflation scheme, for two reasons. First, the total number of kinematic bins used in the JAM fit is small (N=70N = 70), so the nominal inflation factor (2N)1/4≈1.9(2N)^{1/4} \approx1.9 would uniformly enlarge the outer-bin errors without a compelling physical justification. Second, the BH SOI inferred here (rSOI≈0′′05r_{\mathrm{SOI}} \approx0\prime\prime05, § 6.4) is only marginally resolved, being comparable to, and in fact slightly smaller than, the effective spatial resolution of the Keck/OSIRIS data. Under these conditions, the SOI/non-SOI partition on which selective inflation relies is not cleanly defined, and applying it would risk introducing rather than mitigating a systematic bias on MBHM_{\mathrm{BH}}. We therefore retain the formal kinematic uncertainties for all bins and rely on the model-averaging procedure described above to capture the associated systematic budget.

The IMBH Dynamical Mass

Table 5 summarizes the best-fit parameters and 1σ1\sigma CLs from all four JAM configurations. The inferred MBHM_{\mathrm{BH}} spans log⁡10(MBH/M⊙)≈5.23\log_{10}(M_{\mathrm{BH}}/M_{\odot}) \approx5.23–5.385.38 across the ensemble, a factor of ∼1.4\sim1.4 range that reflects the combined sensitivity to velocity-ellipsoid geometry and PSF treatment. Both the inner and outer stellar mass-to-light ratios, M/LF814W′′0≈0.07M/L^{\prime\prime0}_{\mathrm{F814W}} \approx0.07–0.10 M⊙/L⊙0.10\,M_{\odot}/L_{\odot} and M/LF814W′′0.15≈0.35M/L^{\prime\prime0.15}_{\mathrm{F814W}} \approx0.35–0.38 M⊙/L⊙0.38\,M_{\odot}/L_{\odot}, are stable across all configurations, confirming that the photometric normalization at both small and large radii is insensitive to the assumed orbital geometry. The JAMsph_{\mathrm{sph}} models systematically return higher MBHM_{\mathrm{BH}} than their JAMcyl_{\mathrm{cyl}} counterparts by ∼0.03\sim0.03–0.060.06 dex, with the JAMsph_{\mathrm{sph}} ensemble spanning log⁡10(MBH/M⊙)≈5.26\log_{10}(M_{\mathrm{BH}}/M_{\odot}) \approx5.26–5.385.38 against the JAMcyl_{\mathrm{cyl}} range of log⁡10(MBH/M⊙)≈5.23\log_{10}(M_{\mathrm{BH}}/M_{\odot}) \approx5.23–5.325.32.

Figure 6 compares each best-fit JAM model to the observed VrmsV_{\mathrm{rms}} profile along the kinematic major and minor axes. The OSIRIS data span ∣Δr∣≲0. ⁣′′4|\Delta r| \lesssim0.\!^{\prime\prime}4 and ≲0. ⁣′′2\lesssim0.\!^{\prime\prime}2 along those axes, respectively, sampling kinematics from within the IMBH SOI out to the surrounding stellar component. Both PSF variants reproduce the global shape of the two profiles well, including the central VrmsV_{\mathrm{rms}} peak at ∣Δr∣≲0. ⁣′′1|\Delta r| \lesssim0.\!^{\prime\prime}1 that constitutes the primary kinematic signature of the central dark mass. Residuals between data and model are typically ≲2%\lesssim2\% (∼1\sim1 km s−1^{-1}

across the full kinematic field. Crucially, the MBH=0M_{\mathrm{BH}}=0 model (grey dashed curve) produces a central VrmsV_{\mathrm{rms}} dip in place of the observed peak and fails to reproduce the data in all four configurations. We nonetheless regard this as a marginal detection: because RSOI≈0.04′′R_{\mathrm{SOI}}\approx0.04''–0.06′′0.06'' is comparable to, or slightly below, the OSIRIS PSF FWHM (0.053′′0.053''–0.062′′0.062''), the SOI is only marginally resolved, so the formal uncertainties likely understate the true systematic error on MBHM_{\mathrm{BH}}.

Figure 7 shows the 2D posterior distributions from the AdaMet MCMC optimization for the two JAMcyl_{\mathrm{cyl}} configurations (empirical and diffraction-limited OSIRIS PSF). All four free parameters: log⁡10(MBH/M⊙)\log_{10}(M_{\mathrm{BH}}/M_{\odot}), M/LF814W0′′M/L^{0''}_{\mathrm{F814W}}, M/LF814W0.15′′M/L^{0.15''}_{\mathrm{F814W}}, and σz/σR\sigma_z/\sigma_R, are recovered with well-defined, unimodal posteriors in both cases. Two distinct covariance structures are visible in the joint distributions. First, MBHM_{\mathrm{BH}} exhibits a negative covariance with M/LF814W0′′M/L^{0''}_{\mathrm{F814W}}: a higher central stellar M/LM/L assigns more mass to the stellar component at the scale of the BH SOI, reducing the BH contribution required to reproduce the observed central VrmsV_{\mathrm{rms}} peak, and vice versa. This is the classical BH–M/LM/L degeneracy operating at the resolution scale of the kinematic data ([22, 134]). Second, MBHM_{\mathrm{BH}} shows a positive covariance with M/LF814W0.15′′M/L^{0.15''}_{\mathrm{F814W}}, the outer M/LM/L value: because M/LF814W0.15′′M/L^{0.15''}_{\mathrm{F814W}} governs the stellar mass budget at radii beyond the SOI, increasing it raises the overall stellar mass normalization of the model, which in turn requires a more massive BH to maintain the observed VrmsV_{\mathrm{rms}} profile at the center.

Figure 7. Corner plot summarizing the AdaMet MCMC posteriors for the JAMcyl_{\mathrm{cyl}} fits to the OSIRIS nuclear kinematics of IC 342, with all chains shown post-burn-in. Each triangle plot shows the 2D projected probability distributions of the model parameters, with 1D marginalized histograms along the diagonal; solid and dashed vertical lines indicate the best-fit values and 1σ1\sigma uncertainties, respectively. Insets: Direct comparison of the observed (top) and model (bottom) VrmsV_{\mathrm{rms}} maps on a common velocity stretch, illustrating how closely each model reproduces the kinematics across the OSIRIS FOV. These maps are symmetrized assuming axisymmetry and rotated so that the major and minor axes align with the horizontal and vertical directions using the symmetrize_velfield routine from the PLOTBIN package12, and are displayed with plot_velfield.

This positive covariance is partially degenerate with the limited radial extent of the OSIRIS FOV relative to the spatial scale over which the M/LM/L gradient operates, meaning the two M/LM/L values cannot be fully decoupled from MBHM_{\mathrm{BH}} by the data alone. Both covariances are nonetheless contained within the 3σ3\sigma credible region, and the 1D marginalized histograms of MBHM_{\mathrm{BH}} remains well-constrained in both PSF variants.

The joint distribution between M/LF814W0′′M/L^{0''}_{\mathrm{F814W}} and M/LF814W0.15′′M/L^{0.15''}_{\mathrm{F814W}} reveals a negative covariance: when the central M/LM/L is high, the stellar mass near the BH SOI is already elevated, and the model compensates by preferring a lower outer M/LM/L to maintain a consistent fit to the global VrmsV_{\mathrm{rms}} profile shape. This anti-correlation between the two endpoints of the linear M/LM/L gradient reflects the limited radial leverage of the OSIRIS FOV, which cannot fully decouple the inner and outer mass normalization without wider kinematic coverage.

The MBHM_{\mathrm{BH}}–σz/σR\sigma_z/\sigma_R joint distribution shows a positive covariance, a direct manifestation of the classical mass–anisotropy degeneracy ([14]): more radially biased orbits (σz/σR>1\sigma_z/\sigma_R>1) inflate the projected LOS VrmsV_{\mathrm{rms}}, which the model accommodates by simultaneously increasing MBHM_{\mathrm{BH}}. However, the tight constraint on σz/σR\sigma_z/\sigma_R near isotropy (≈0.96\approx0.96–0.980.98) significantly limits the range over which this degeneracy can operate, leaving the MBHM_{\mathrm{BH}} posterior well-constrained despite the covariance.

Both JAMcyl_{\mathrm{cyl}} configurations recover a well-constrained orbital anisotropy of σz/σR=0.96±0.03\sigma_z/\sigma_R = 0.96 \pm0.03 and 0.98±0.030.98 \pm0.03 for the diffraction-limited and empirical PSF models, respectively, with narrow, symmetric posteriors in both cases. These values are consistent with isotropic stellar orbits (βz≡1−(σz/σR)2≈0\beta_z \equiv1 - (\sigma_z/\sigma_R)^2 \approx0) within 1σ1\sigma, indicating that the nuclear stellar kinematics of IC 342 require no significant vertical anisotropy to reproduce the observed VrmsV_{\mathrm{rms}} field. The near-isotropic solution is stable across both PSF treatments, confirming that this result is not sensitive to the adopted PSF model. Near-isotropic orbits are a characteristic feature of fast-rotating, disk-dominated systems (e.g., [22]) and have been consistently recovered in stellar dynamical studies of similar late-type spirals (e.g., [110 c]), lending further confidence that the JAMcyl_{\mathrm{cyl}} orbital structure recovered here is physically realistic.

Similarly, the 2D JAMsph_{\mathrm{sph}} posteriors share the same MBHM_{\mathrm{BH}}–M/LM/L covariance structure as the JAMcyl_{\mathrm{cyl}} models: MBHM_{\mathrm{BH}} exhibits a negative covariance with M/LF814W0′′.15M/L^{0^{\prime\prime}.15}_{\mathrm{F814W}} and a positive covariance with M/LF814W0′′.15M/L^{0^{\prime\prime}.15}_{\mathrm{F814W}}, while the two M/LM/L endpoints are anti-correlated with each other. However, the JAMsph_{\mathrm{sph}} posteriors reveal additional covariance structures arising from the logistic anisotropy parameterization that are absent in the JAMcyl_{\mathrm{cyl}} case.

Most notably, MBHM_{\mathrm{BH}} shows a strong negative covariance with the inner dispersion ratio (σr/σθ)0(\sigma_r/\sigma_\theta)_0: a more tangentially biased central orbit distribution (lower (σr/σθ)0(\sigma_r/\sigma_\theta)_0) suppresses the projected VrmsV_{\mathrm{rms}} in the innermost bins, requiring a more massive BH to compensate and reproduce the observed central peak. This is a manifestation of the mass–anisotropy degeneracy operating specifically at the scale of the BH SOI, as similar as the JAMcyl_{\mathrm{cyl}}, and represents the dominant source of the broader MBHM_{\mathrm{BH}} uncertainties in JAMsph_{\mathrm{sph}}. Conversely, MBHM_{\mathrm{BH}} exhibits a mild positive covariance with the outer dispersion ratio (σr/σθ)∞(\sigma_r/\sigma_\theta)_\infty: more radially biased orbits at large radii inflate the outer VrmsV_{\mathrm{rms}}, which is partially compensated by increasing MBHM_{\mathrm{BH}} to steepen the central profile, though this covariance is weaker given that the outer anisotropy primarily affects bins well outside the SOI.

The two anisotropy parameters themselves are anti-correlated: a more tangentially biased center (low (σr/σθ)0(\sigma_r/\sigma_\theta)_0) is preferentially paired with a more radially biased outer region (high (σr/σθ)∞(\sigma_r/\sigma_\theta)_\infty), reflecting the physical prior that the logistic function must produce a monotonically increasing anisotropy profile from center to large radii.

The logistic anisotropy profile is well-behaved across both JAMsph_{\mathrm{sph}} configurations. The inner dispersion ratio (σr/σθ)0(\sigma_r/\sigma_\theta)_0 is constrained to the tangentially biased regime at ≈0.92\approx0.92–0.930.93, consistent with the physically motivated prior (σr/σθ)0<1(\sigma_r/\sigma_\theta)_0 < 1 and confirming that the OSIRIS kinematics genuinely prefer tangential orbits within the BH SOI rather than being driven to this solution solely by the prior boundary. The outer dispersion ratio (σr/σθ)∞≈1.16(\sigma_r/\sigma_\theta)_\infty\approx1.16 is recovered in the mildly radial regime, also consistent with its prior range, with a narrow posterior indicating that the kinematic data at larger radii provide meaningful constraints on the outer orbital structure. Together, these two parameters define a logistic transition from tangential to radial anisotropy with increasing radius, qualitatively consistent with the anisotropy profiles reported in stellar dynamical studies of other galactic nuclei hosting central dark masses (e.g., [131, 110 c]).

Figure 10 presents all four MBHM_{\mathrm{BH}} measurements with their 1σ1\sigma uncertainties, spanning the two velocity-ellipsoid alignments and two PSF treatments. To extract a single fiducial estimate from this ensemble, we take the median across the four models and propagate the spread via 1σ1\sigma bootstrap resampling, obtaining MBH=(1.95−0.16+0.12)×105 M⊙M_{\mathrm{BH}} = (1.95^{+0.12}_{-0.16}) \times10^5\,M_\odot.

Four independent JAM measurements of $M_{\mathrm{BH}}$ for IC 342

Figure 10. Four independent JAM measurements of MBHM_{\mathrm{BH}} for IC 342, obtained from the Keck/OSIRIS stellar kinematics by combining two velocity-ellipsoid alignments (JAMcyl_{\mathrm{cyl}}, JAMsph_{\mathrm{sph}}) with two treatments of the OSIRIS PSF (empirical, diffraction-limited). Points with error bars give each measurement and its 1σ1\sigma uncertainty; the dashed vertical line marks our adopted MBHM_{\mathrm{BH}}, with the shaded band showing the total uncertainty after combining the statistical and systematic components in quadrature.

The total uncertainty on this fiducial mass combines two components added in quadrature. The first is the statistical component, taken as the MCMC CL of each individual model. The second is a systematic component quantifying the model-to-model scatter across the four JAM runs, which captures the sensitivity of MBHM_{\mathrm{BH}} to the choice of velocity-ellipsoid geometry and PSF model. Distance uncertainty is handled separately (§ 6.5.1) rather than folded into this budget, since MBHM_{\mathrm{BH}} scales linearly with DD and its contribution is most transparently reported in isolation. Combining all terms yields our adopted black hole mass of MBH=(1.95±0.61)×105 M⊙M_{\mathrm{BH}} = (1.95 \pm0.61) \times10^5\,M_\odot for IC 342.

As mentioned in § 2.1, IC 342 has no prior direct, dynamical MBHM_{\mathrm{BH}} measurement. The only existing constraint is an upper limit of MBH<9.6×105 M⊙M_{\mathrm{BH}} < 9.6 \times10^5\,M_\odot, scaled to our adopted TRGB distance of D=3.45D = 3.45 Mpc (P.-F. Wu et al. 2014), was inferred by T. Böker et al. [18] from 1D spherical Jeans models of the NSC using seeing-limited ground-based spectroscopy at ∼5′′\sim5^{\prime\prime} resolution, a scale more than two orders of magnitude larger than the BH SOI. At such resolution, the BH gravitational influence is entirely unresolved and the upper limit carries little constraining power on the actual MBHM_{\mathrm{BH}}. Our Keck/OSIRIS LGS-AO data, with a PSF FWHM of 0. ⁣′′0620.\!^{\prime\prime}062, marginally resolve the BH SOI (RSOI≈0. ⁣′′05R_{\mathrm{SOI}} \approx0.\!^{\prime\prime}05 when

of 0. ⁣′′0620.\!^{\prime\prime}062, marginally resolve the BH SOI (RSOI≈0. ⁣′′05R_{\mathrm{SOI}} \approx0.\!^{\prime\prime}05 when adopted the NSC velocity dispersion of 33 km s−1^{-1} ([18]) and represent the first dataset capable of delivering a direct, stellar-dynamical MBHM_{\mathrm{BH}} measurement in IC 342. The result presented here is therefore the first direct dynamical MBHM_{\mathrm{BH}} determination for this galaxy, filling a long-standing gap in the black hole census of nearby late-type spirals.

With MBH=(1.95±0.61)×105 M⊙M_{\mathrm{BH}} = (1.95 \pm0.61) \times10^{5}\,M_{\odot}, IC 342 joins the small sample of IMBHs with direct dynamical mass measurements, becoming the eighth such measurement to date: NGC 205 ([102]), NGC 5102 and NGC 5206 ([101, 102]), NGC 404 ([39]), NGC 3593 ([105]), NGC 4395 ([41]), and IC 750 ([145]). IC 342 is also now the second bulgeless spiral galaxy (i.e., it joins NGC 4395) with a direct dynamical MBHM_{\mathrm{BH}} measurement, extending this rare dynamical census to a galaxy without a classical bulge.

Additional Uncertainty Budgets

Distances

We adopt the TRGB distance D=3.45±0.13D = 3.45 \pm0.13 Mpc (P.-F. Wu et al. 2014) as our fiducial value throughout. The NASA/IPAC Extragalactic Database lists 17 independent distance determinations for IC 342, spanning 1.0–6.0 Mpc and yielding a formal mean of ⟨D⟩=3.51±1.42\langle D\rangle= 3.51 \pm1.42 Mpc; the large scatter is driven by older, less reliable Tully–Fisher estimates rather than by genuine uncertainty in the true distance. Restricting to the six most reliable measurements: Cepheids ([118]) and TRGB (P.-F. Wu et al. 2014), yields a weighted mean of 3.39±0.073.39 \pm0.07 Mpc, in excellent agreement with our adopted value and confirming that D≈3.45D \approx3.45 Mpc is well-supported by the best available distance indicators. Since MBH∝DM_{\mathrm{BH}} \propto D, the ±0.13\pm0.13 Mpc uncertainty on our adopted distance propagates directly into a systematic uncertainty of ±3.8%\pm3.8\% on MBHM_{\mathrm{BH}}.

Constant versus Varying M/LF814WM/L_{\mathrm{F814W}} Parameterization

As a test of whether the radial variation in M/LF814WM/L_{\mathrm{F814W}} is required by the stellar kinematics, we repeated all four JAM configurations using a spatially constant M/LF814WM/L_{\mathrm{F814W}}. The constant-M/LM/L models give systematically worse fits than the corresponding varying-M/LF814WM/L_{\mathrm{F814W}} models, with

χconstant M/L2−χvarying M/L2=206–241\chi^2_{\mathrm{constant}\ M/L} - \chi^2_{\mathrm{varying}\ M/L} = 206\text{--}241

across the four JAM configurations. Since the varying model introduces only one additional free parameter, this difference quantifies the preference of the stellar kinematics for a radial M/LF814WM/L_{\mathrm{F814W}} gradient.

In all four JAM configurations, the constant-M/LF814WM/L_{\mathrm{F814W}} models allow MBH=0M_{\mathrm{BH}} = 0 within 1σ1\sigma, providing only an upper limit rather than a black-hole detection (Figure 11). This is in contrast to the models with a varying M/LF814WM/L_{\mathrm{F814W}}, which consistently give a non-zero black-hole mass.

Figure 11. Same as Figure 6 but for the constant-M/LF814WM/L_{\mathrm{F814W}} JAM configurations. Top: JAMcyl_{\mathrm{cyl}}; Bottom: JAMsph_{\mathrm{sph}}; diffraction-limited (left) and empirical (right) OSIRIS PSF in each case. All four models fail to reproduce simultaneously the central VrmsV_{\mathrm{rms}} peak and the outer kinematic bins, leaving larger systematic residuals (∼\sim4 km s−1^{-1}) at both radial extremes compare to the the detection case as shown in Figure 6. The posteriors consequently include MBH=0 M⊙M_{\mathrm{BH}} = 0\,M_{\odot} within 1σ1\sigma (see Figure 12), motivating the adoption of a radially varying M/LF814WM/L_{\mathrm{F814W}} profile (§ 6.5.2).

The difference between the two parameterizations is also apparent in the fitted VrmsV_{\mathrm{rms}} profiles. The constant-M/LF814WM/L_{\mathrm{F814W}} models produce systematically larger residuals, exceeding ∼\sim4 km s−1^{-1} in the innermost bins and near the outer edge of the OSIRIS field, whereas the varying-M/LF814WM/L_{\mathrm{F814W}} models generally reproduce the observed kinematics to within ≲2%\lesssim2\% (∼1\sim1 km s−1^{-1}; § 6.4). Figure 11 and Figure 12 show the corresponding 2D posterior distributions and 1D marginalized distributions. None of the four constant-M/LF814WM/L_{\mathrm{F814W}} configurations can simultaneously reproduce the observed VrmsV_{\mathrm{rms}} field and constrain MBHM_{\mathrm{BH}} above zero.

Figure 12. Same as Figure 7 and Figure 8–Figure 9, but for JAM models assuming a spatially constant M/LF814WM/L_{\mathrm{F814W}} across the OSIRIS FOV. Top panels: JAMcyl_{\mathrm{cyl}} models with the diffraction-limited (left) and empirical (right) PSF. Bottom panels: JAMsph_{\mathrm{sph}} models with the same PSF ordering. In all four configurations, the posteriors include MBH=0M_{\mathrm{BH}} = 0 within 1σ1\sigma, yielding only an upper limit on MBHM_{\mathrm{BH}} rather than a detection. The OSIRIS stellar kinematics are inconsistent with a constant M/LF814WM/L_{\mathrm{F814W}} parameterization and instead require a radially varying M/LF814WM/L_{\mathrm{F814W}} profile (§ 6.5.2).

Figure 8. Same as Figure 7 but for the JAMsph_{\mathrm{sph}} fit with the empirical OSIRIS PSF (Table 1).

Figure 9. Same as Figure 7 but for the JAMsph_{\mathrm{sph}} fit with the diffraction-limited OSIRIS PSF (Table 1).

The preference for a varying M/LF814WM/L_{\mathrm{F814W}} is also physically plausible. IC 342 hosts a prominent NSC with an estimated age of ∼\sim60–630 Myr (T. Böker et al. [18]), younger than the surrounding stellar population. The dynamical models give a central M/LF814W≈(0.07M/L_{\mathrm{F814W}} \approx(0.07–0.10) M⊙/L⊙0.10)\ M_\odot/L_\odot, increasing to M/LF814W≈(0.35M/L_{\mathrm{F814W}} \approx(0.35–0.38) M⊙/L⊙0.38)\ M_\odot/L_\odot at larger radii. This lower central M/LF814WM/L_{\mathrm{F814W}} is consistent with the presence of a younger stellar population in the nucleus. The central dynamical M/LF814WM/L_{\mathrm{F814W}} represents a luminosity-weighted average of the nuclear cluster and surrounding stars within the innermost resolution elements, and therefore is expected to be higher than the M/LF814WM/L_{\mathrm{F814W}} of the nuclear cluster alone. For comparison, stellar-population models for a young nuclear cluster with a Kroupa IMF and solar metallicity give M/LI∼0.015M/L_I \sim0.015–0.16 M⊙/L⊙0.16\ M_\odot/L_\odot (G. Bruzual & S. Charlot [20]), broadly consistent with the lower central dynamical values once the contribution from the surrounding older population is included.

The strong foreground extinction toward IC 342 (AF814W≈0.74A_{\mathrm{F814W}} \approx0.74 mag; E. F. Schlafly & D. P. Finkbeiner [123]), however, prevents an independent optical stellar-population measurement of the central M/LF814WM/L_{\mathrm{F814W}} gradient. We therefore rely on the stellar kinematics to determine whether a radial M/LF814WM/L_{\mathrm{F814W}} variation is required, while the young nuclear population provides independent physical motivation for the inferred gradient. Taken together, the model comparison and the physical properties of the nuclear cluster support the use of a radially varying M/LF814WM/L_{\mathrm{F814W}} in our dynamical modeling.

Discussion

Isotropic Nuclear Orbits as a Signature of Merger-Free Secular Growth

Beyond providing a robust MBHM_{\mathrm{BH}}, the orbital structure recovered by our JAM models carries direct information about the assembly history of the IC 342 nucleus. Both JAMcyl_{\mathrm{cyl}} configurations return tightly constrained, near-isotropic orbits (e.g., σz/σR≈0.96\sigma_z/\sigma_R \approx0.96–0.980.98; βz≈0\beta_z \approx0), and the JAMsph_{\mathrm{sph}} models find only a mild tangential bias confined to the innermost region (e.g., (σr/σθ)∞≈0.92(\sigma_r/\sigma_\theta)_\infty\approx0.92–0.930.93; βr<0\beta_r < 0). This stands in contrast to the anisotropy signatures characteristic of merger-built nuclei: violent relaxation during dissipationless mergers imprints radially biased orbits in classical bulges (e.g., L. Hernquist [71]), while binary SMBH scouring in gas-poor major mergers preferentially ejects stars on radial orbits, leaving strongly tangentially biased cores ([91, 115]). The absence of either signature in IC 342 indicates a phase space that has never been dynamically disturbed by a significant merger: a kinematic counterpart to the photometric evidence for a bulgeless disk (§ 5). The nuclear structure of IC 342, comprising a young NSC and an IMBH embedded in molecular gas with ongoing inflow (E. Schinnerer et al. [122]), is therefore most naturally explained by purely secular, gas-driven growth, which is consistent with the bar-driven inflow channel discussed in § 2. In this sense, the near-isotropic orbits constitute a kinematic “smoking gun” for merger-free CMO assembly, and establish IC 342 as a clean laboratory for the secular-growth mode explored in the remainder of this section.

IC 342 on the BH-Galaxy Scaling Relations

The secular, merger-free assembly history inferred above provides the natural context for interpreting the position of IC 342 on the BH–galaxy scaling relations. Figure 13 shows the location of IC 342 on the MBHM_{\mathrm{BH}}–σ\sigma and MBHM_{\mathrm{BH}}–M⋆M_{\star} diagrams alongside galaxies with direct dynamical MBHM_{\mathrm{BH}} measurements, velocity dispersions, and stellar masses ([120, 58], references therein). In both panels, IC 342 lies below the spiral galaxy relations (cyan lines), indicating that its IMBH is somewhat undermassive for its host galaxy properties, even relative to other late-type galaxies. Although these offsets are modest (∼\sim0.45 dex in the MBHM_{\mathrm{BH}}–σ\sigma plane and ∼\sim0.3 dex in the MBHM_{\mathrm{BH}}–M⋆M_{\star} plane), IC 342 remains well within the intrinsic 1σ\sigma scatter of the spiral galaxy relations.

Figure 13. Location of IC 342 (blue star) on the MBHM_{\mathrm{BH}}–σ0\sigma_{0} (left) and MBHM_{\mathrm{BH}}–M⋆,galM_{\star,\mathrm{gal}} (right) scaling diagrams. Symbol types and scaling relation lines follow A. W. Graham [57]; A. W. Graham & N. Sahu [62]; A. W. Graham [58] as indicated in the legend.

Nonetheless, its systematic location below the relations may be physically meaningful. IC 342 is a confirmed bulgeless SABcd galaxy whose nucleus is dominated by an NSC and has evolved primarily through secular processes rather than through the violent relaxation associated with the formation of a classical bulge. Bulgeless and pseudobulge galaxies have long been recognised as preferentially occupying the lower-right side of the classical-bulge sequences in both the MBHM_{\mathrm{BH}}–σ\sigma and MBHM_{\mathrm{BH}}–M⋆M_{\star} diagrams ([52 J. Hu] 2008; J. E. Greene et al. [65]; J. Kormendy & L. C. Ho [77], although spiral galaxies as a population are likewise systematically offset (Figure 13).

These offsets likely reflect differences in the co-evolution of black holes and their host galaxies. At fixed σ\sigma, spiral galaxies appear to host systematically lower-mass black holes than early-type galaxies. Moreover, their larger disk-to-bulge stellar mass ratios naturally displace them further from the MbhM_{\mathrm{bh}}–M⋆,bulgeM_{\star,\mathrm{bulge}} relation than is the case for early-type galaxies in the MbhM_{\mathrm{bh}}–M⋆M_{\star} plane. The position of IC 342 below the spiral-galaxy scaling relations is therefore consistent with a growth history in which the galaxy assembled largely through secular evolution, efficiently building its nuclear star cluster while, thus far, experiencing comparatively limited black-hole growth.

The scientific significance of this measurement extends beyond its position on the scaling diagrams. At the low-mass end (MBH≲106 M⊙M_{\mathrm{BH}} \lesssim10^{6}\ M_{\odot}), direct dynamical MBHM_{\mathrm{BH}} measurements remain extremely scarce, and the shape, normalization, and scatter of the MBHM_{\mathrm{BH}}–σ\sigma relation are poorly constrained (J. E. Greene et al. [65]; D. D. Nguyen et al. [101, 102]). This is precisely the regime where theoretical models of BH seeding diverge most strongly: light-seed models from the remnants of Population III stars (P. Madau & M. J. Rees [84]), predict a wide scatter and high BH occupation fraction in low-mass galaxies, while heavy-seed models via the direct collapse of gas clouds (G. Lodato & P. Natarajan [83]) predict a lower occupation fraction but a higher minimum BH mass (M. Volonteri et al. [142]). Populating the IMBH end of the scaling relations with direct, well-resolved dynamical measurements in well-characterized hosts is therefore essential for discriminating between competing seeding scenarios. As a nearby, bulgeless spiral with a well-studied NSC and now a directly measured IMBH, IC 342 provides one of the most robust anchors currently available for this low-mass regime, and demonstrates that the MBHM_{\mathrm{BH}}–σ\sigma relation for spiral galaxies extends continuously down to at least MBH∼105 M⊙M_{\mathrm{BH}} \sim10^{5}\ M_{\odot}.

IC 342 on the BH–NSC Scaling Relation

Narrowing from galaxy-wide properties to the CMO themselves, Figure 14 shows the location of IC 342 on the MBHM_{\mathrm{BH}}–MNSCM_{\mathrm{NSC}} diagram of A. W. Graham [55], updated with recent measurements from A. W. Graham et al. [59]. IC 342 (blue star) is plotted at its dynamically measured MBH≈(1.95±0.61)×105 M⊙M_{\mathrm{BH}} \approx(1.95 \pm0.61) \times10^{5}\ M_{\odot} and NSC mass MNSC≈1.17×107 M⊙M_{\mathrm{NSC}} \approx1.17 \times10^{7}\ M_{\odot} (T. Böker et al. [18]), scaled to our adopted distance. IC 342 falls below the best-fit relation (solid line) but within the 1σ1\sigma scatter region (dotted lines), with a mass ratio MBH/MNSC≈0.017M_{\mathrm{BH}}/M_{\mathrm{NSC}} \approx0.017, placing it near the 10−210^{-2} iso-ratio line. This is consistent with the general picture in which lower-mass systems tend to be NSC-dominated, with the BH contributing only a small fraction of the total CMO mass

MBH–MNSC scaling diagram

Figure 14. MBHM_{\mathrm{BH}}–MNSCM_{\mathrm{NSC}} scaling diagram adapted from A. W. Graham [55] with updates from A. W. Graham et al. [59], showing IC 342 (blue star) at its dynamically measured MBHM_{\mathrm{BH}} and NSC mass MNSC≈1.17×107 M⊙M_{\mathrm{NSC}} \approx1.17 \times10^{7}\ M_{\odot} (T. Böker et al. [18]). Red stars and orange squares denote systems hosting NSCs and UCDs, respectively; gray diagonal lines indicate constant MBH/MNSCM_{\mathrm{BH}}/M_{\mathrm{NSC}} ratios as labeled. The solid line shows the best-fit relation with 1σ1\sigma intrinsic scatter indicated by the dotted lines. Systems with MBH>108 M⊙M_{\mathrm{BH}} > 10^{8}\ M_{\odot} (gray shaded region) are omitted as their NSCs are expected to have been depleted by binary SMBH scouring during dry mergers (K. Bekki & A. W. Graham [13]); flattened nuclear disks, which can extend to tens of parsecs in size (e.g., M. Balcells et al. [9]), are likewise excluded as they do not follow the NSC–BH relation. IC 342 falls within the 1σ1\sigma scatter of the relation, confirming its membership in the NSC–BH coexistence regime at the IMBH mass scale.

(A. W. Graham & L. R. Spitler [64]; A. W. Graham [55]; N. Neumayer et al. [93]).

The gray shaded region at the top of Figure 14 marks the regime MBH≳108 M⊙M_{\mathrm{BH}} \gtrsim10^{8}\ M_{\odot} where massive merger-built ETGs dominate and NSCs start to be eroded by binary SMBH scouring (K. Bekki & A. W. Graham [13]); IC 342 is far from this regime, firmly in the NSC–BH coexistence zone where both CMOs grow simultaneously through secular processes. The position of IC 342 on this diagram is particularly significant because it represents one of the few systems where MBHM_{\mathrm{BH}} at the IMBH mass scale (MBH∼105 M⊙M_{\mathrm{BH}} \sim10^{5}\ M_{\odot}) is directly and dynamically measured, rather than inferred from indirect indicators, making it a robust anchor for the low-mass end of the MBHM_{\mathrm{BH}}–MNSCM_{\mathrm{NSC}} relation.

Beyond galaxy evolution, the calibration of the MBHM_{\mathrm{BH}}–MNSCM_{\mathrm{NSC}} relation at the IMBH end also bears directly on predictions for gravitational wave science, a point we return to in § 7.6.

An Evolutionary Sequence of In-Situ CMO Growth: from M33 to IC 342

The nuclear region of IC 342 may be showing us NSC formation in action. As illustrated in Figure 4, the circumnuclear disk of gas and dust hosts several off-center knots of star formation. Such clumps are the natural outcome of a cold, gas-rich disk fragmenting and collapsing once local overdensities exceed the Jeans criterion; once formed, dynamical friction against the background disk drains their orbital energy, driving them to spiral inward and merge at the bottom of the potential well (e.g., S. D. Tremaine et al. [135]; K. Bekki & A. W. Graham [13]; M. Arca-Sedda & R. Capuzzo-Dolcetta [5]; N. Neumayer et al. [93]). The observed configuration: a young, massive NSC surrounded by off-center star-forming clumps (T. Böker et al. [15]; D. J. Carson et al. [31]) and ongoing molecular gas inflow (E. Schinnerer et al. [122]), suggests that both the NSC and the IMBH of IC 342 are still actively growing through this secular, in-situ channel.

This picture invites comparison with M33, a nearby bulgeless spiral that appears to sit earlier along the same evolutionary pathway. M33 hosts a well-studied NSC (D. J. Carson et al. [31]), yet carries one of the most stringent stellar-dynamical BH limits known, MBH≲(1.5−3)×103 M⊙M_{\mathrm{BH}} \lesssim(1.5 - 3) \times10^{3}\ M_{\odot} (K. Gebhardt et al. [49]; D. Merritt et al. [90]): its nucleus has evidently not yet grown, or fed, a substantial central BH. If the in-situ channel operating in IC 342 is generic, the sequence M33 → IC 342 traces the gradual, NSC-mediated assembly of central BHs in secular disk galaxies. This constitutes an argument against the universality of massive (≳105 M⊙\gtrsim10^{5}\ M_{\odot}) BH seeds, since it implies that at least some disk galaxies begin with little or no central BH and grow one only gradually within their nucleus (e.g., N. Neumayer & A. C. Seth [94]; M. Volonteri et al. [142]; A. W. Graham & R. Soria [63]). We note that the Large Magellanic Cloud, which lacks a prominent NSC, may nonetheless host a ≃6×105 M⊙\simeq6 \times10^{5}\ M_{\odot} IMBH inferred from hypervelocity-star trajectories (J. J. Han et al. [68]); if confirmed, its inverted configuration, a BH without an NSC, would point to a distinct growth channel outside the in-situ sequence traced here. Direct dynamical measurements at the IMBH scale in further bulgeless disks will be required to populate, or refute, this proposed sequence.

IC 342 as a Live Analog of Stripped Nuclei: Centauri and B023-G078

The NSC–IMBH configuration of IC 342 finds a striking counterpart in two nearby stellar systems believed to be the surviving nuclei of tidally disrupted galaxies. ω\omega Centauri, the most massive globular cluster of the Milky Way, is now widely regarded as the stripped NSC of an accreted dwarf galaxy (H. Zinnecker et al. [147]; K. Bekki & K. C. Freeman [12]; D. Massari et al. [89]; R. A. Ibata et al. [75]), most plausibly the Gaia-Enceladus/Sequoia progenitor accreted ∼10\sim10 Gyr ago. M. Häberle et al. [67] recently identified seven fast-moving stars within its central 0.08 pc, establishing a firm lower limit of MBH≳8,200 M⊙M_{\mathrm{BH}} \gtrsim8{,}200\ M_{\odot} (and plausibly up to ∼5×104 M⊙\sim5 \times10^{4}\ M_{\odot}) for a central IMBH. An even closer analog in mass is B023-G078, the most massive globular cluster of M31 and likewise a stripped-nucleus candidate, for which R. Pecetti et al. (2022) measured a dynamical BH mass of ∼105 M⊙\sim10^{5}\ M_{\odot}, which is nearly identical to our measurement in IC 342.

These systems plausibly represent two evolutionary stages of the same phenomenon. IC 342 shows the in-situ configuration: an IMBH of ∼105 M⊙\sim10^{5}\ M_{\odot} embedded in a massive, actively growing NSC (MNSC≈1.17×107 M⊙M_{\mathrm{NSC}} \approx1.17 \times10^{7}\ M_{\odot}) at the center of an intact, gas-rich, bulgeless disk galaxy, still fed by molecular inflows (E. Schinnerer et al. [122]). Should a galaxy like IC 342 be accreted by a more massive host, tidal stripping would remove its disk while the dense nucleus, and its IMBH, survives, leaving behind precisely the kind of massive, chemically complex star cluster observed as ω\omega Cen and B023-G078 (J. Pfeffer et al. [113]; N. Neumayer et al. [93]). IC 342 thus offers a rare view of the progenitor state of stripped nuclei, with both CMO masses dynamically measured, anchoring the initial conditions that stripped-nucleus and globular-cluster IMBH studies must reproduce.

IC 342 as a Local EMRI Laboratory and a Caution on NSC Structural Fits

The evolutionary framework above also positions IC 342 as a valuable target for gravitational wave science. Dense NSCs surrounding IMBHs are expected to be the dominant sites of extreme-mass-ratio inspirals (EMRIs), in which stellar-mass compact remnants spiral into the central BH through gravitational wave emission (P. Amaro-Seoane et al. [1]). At D=3.45D = 3.45 Mpc, IC 342 is among the nearest systems in which both the NSC mass (≈1.17×107 M⊙\approx1.17 \times10^{7}\ M_{\odot}) and the IMBH mass (≈105 M⊙\approx10^{5}\ M_{\odot}), squarely within the optimal LISA sensitivity band) are directly and dynamically characterized. The directly measured NSC–IMBH system of IC 342 therefore provides an ideal local test bed and target template for modeling EMRI capture rates and waveform templates for the upcoming LISA mission.

We close with an important caveat for such modeling efforts. EMRI rate calculations require the stellar density profile of the NSC, typically encoded in Sérsic, King, or Plummer fits to the observed light distribution. However, as the archival HST imaging makes clear, the nuclear region of IC 342 is a lumpy conglomerate of star-forming clumps and clusters rather than a single, smoothly declining stellar system akin to a globular cluster (T. Böker et al. [15]; D. J. Carson et al. [31]). Single-component structural fits to distant, poorly resolved NSCs implicitly assume such smooth profiles, and the resulting sizes and masses, which propagate directly into published NSC scaling relations and EMRI rate predictions, may consequently carry substantial, unquantified systematic errors. The reported MNSCM_{\mathrm{NSC}} of IC 342 itself, derived from aperture-integrated kinematics (T. Böker et al. [18]), should be understood as the mass of the nuclear conglomerate rather than of a single relaxed cluster. High-resolution imaging and spectroscopy of nearby, resolved systems such as IC 342 are therefore essential for calibrating the structural assumptions underpinning both NSC demographics and LISA-era EMRI forecasts.

Conclusions

We have presented the first direct stellar-dynamical measurement of an IMBH in IC 342 using NIR LGS-AO IFS with Keck/OSIRIS. The unprecedented combination of high angular resolution (PSF FWHM ≈0.053′′−0.062′′\approx0.053^{\prime\prime}-0.062^{\prime\prime}, corresponding to ≈0.8\approx0.8–1 pc at D=3.45D = 3.45 Mpc) and KK-band observations that penetrate the heavy foreground extinction (AF814W≈0.74A_{\mathrm{F814W}} \approx0.74 mag) delivers the first spatially resolved view of the nuclear stellar kinematics in this system. Our main results are as follows:

  1. High-resolution nuclear stellar kinematics. The 2D LOSVDs extracted from the CO bandheads (∼2.3 μ\sim2.3\ \mum) via PPXF reveal a velocity dispersion rising from ∼42 km s−1\sim42\ \mathrm{km\ s^{-1}} at r≳0.2′′r \gtrsim0.2^{\prime\prime} to a central peak of ∼49 km s−1\sim49\ \mathrm{km\ s^{-1}} within the innermost resolution element. This central increase in both σ\sigma and V2+σ2\sqrt{V^2+\sigma^2} provides unambiguous kinematic evidence for a central dark mass embedded within the IC 342 NSC.

  1. First direct IMBH mass measurement. Modeling the observed V2+σ2\sqrt{V^2+\sigma^2} maps with an ensemble of four JAM models, spanning cylindrical (JAMcyl_{\mathrm{cyl}}) and spherical (JAMsph_{\mathrm{sph}}) velocity-ellipsoid alignments and two OSIRIS PSF treatments (empirical and diffraction-limited), combined with a three-component MGE decomposition of archival HST/WFC3 F814W imaging, yields an IMBH mass of MBH=(1.95±0.61)×105 M⊙M_{\mathrm{BH}} = (1.95 \pm0.61) \times10^5\ M_{\odot}. All four model variants are consistent within 1σ1\sigma, and the SOI (RSOI≈0.05′′R_{\mathrm{SOI}} \approx0.05^{\prime\prime}) is marginally resolved by the OSIRIS PSF (≈0.053′′\approx0.053^{\prime\prime}–0.062′′0.062^{\prime\prime}), making this the first direct dynamical MBHM_{\mathrm{BH}} determination for IC 342.

  1. Stellar orbital structure. Both JAMcyl_{\mathrm{cyl}} and JAMsph_{\mathrm{sph}} consistently recover near-isotropic orbits in the nuclear region (σz/σR≈0.96\sigma_z/\sigma_R \approx0.96–0.980.98 for JAMcyl_{\mathrm{cyl}}; βz≈0\beta_z \approx0), consistent with the orbital structure typical of fast-rotating, disk-dominated systems [22, 110 c]. The JAMsph_{\mathrm{sph}} logistic anisotropy profile confirms the physically expected transition from tangentially biased orbits in the center ((σr/σθ)0≈0.92(\sigma_r/\sigma_\theta)_0 \approx0.92–0.930.93), to mildly radial orbits at large radii ((σr/σθ)∞≈1.16(\sigma_r/\sigma_\theta)_\infty\approx1.16), consistent with independent validation on M87 [131], NGC 4258 [110], NGC 3031 [111], NGC 5055 (T. H. T. Ho et al. submitted), and NGC 4736 (H. N. Ngo et al. submitted).

  1. Radially varying stellar M/LF814WM/L_{\mathrm{F814W}}. The OSIRIS kinematics consistently disfavor a spatially constant M/LF814WM/L_{\mathrm{F814W}}: constant-M/LM/L models include MBH=0M_{\mathrm{BH}} = 0 within 1σ1\sigma and produce systematically worse residuals exceeding ∼4 km s−1\sim4\ \mathrm{km\ s^{-1}} in both the innermost and the edge bins. The fiducial radially varying profile, with inner value M/LF814W0′′≈0.07M/L_{\mathrm{F814W}}^{0^{\prime\prime}} \approx0.07–0.10 M⊙/L⊙0.10\ M_{\odot}/L_{\odot} and outer value M/LF814W15′′≈0.35M/L_{\mathrm{F814W}}^{15^{\prime\prime}} \approx0.35–0.38 M⊙/L⊙0.38\ M_{\odot}/L_{\odot}, is qualitatively consistent with the young NSC stellar population (∼60\sim60–630 Myr; [18]) dominating the innermost parsecs atop an older surrounding component. This result serves as a critical caveat for dynamical searches of IMBHs in LTGs: assuming a constant M/LM/L in systems with young, luminous NSCs systematically overestimates the central stellar mass. This “phantom” stellar mass artificially accounts for the central rise in velocity dispersion, thereby absorbing the gravitational signature of any central compact object and yielding false-negative null results (i.e., mere upper limits). Correctly modeling stellar population and M/LM/L gradients is therefore a vital prerequisite for conducting a robust, unbiased IMBH census at the low-mass end of the scaling relations.

  1. Photometric and systematic robustness. Replacing our four-component photometric decomposition with the single-Sérsic GALFIT model of D. J. Carson et al. [31] yields MBHM_{\mathrm{BH}} values consistent within 1σ1\sigma across all four JAM configurations, confirming insensitivity to the photometric model choice. Conversely, omitting dust masking from the decomposition shifts MBHM_{\mathrm{BH}} by more than 3σ3\sigma, validating dust masking as an essential step for a reliable dynamical measurement in this heavily obscured system.

  1. NSC–BH coexistence. With MBH≈1.95×105 M⊙M_{\mathrm{BH}} \approx1.95 \times10^5\ M_{\odot} and MNSC≈1.17×107 M⊙M_{\mathrm{NSC}} \approx1.17 \times10^7\ M_{\odot}, IC 342 falls in the transition regime where NSCs and BHs are expected to coexist, with a mass ratio MBH/MNSC≈0.017M_{\mathrm{BH}}/M_{\mathrm{NSC}} \approx0.017. This measurement directly probes the NSC–BH coexistence regime and provides a new, well-constrained anchor for theoretical models of CMO co-evolution in the secular-evolution regime. This direct dynamical measurement at the IMBH mass scale anchors the low-mass end of the MBHM_{\mathrm{BH}}–MNSCM_{\mathrm{NSC}} relation, which is a key input for forecasting EMRI event rates detectable by future gravitational wave observatories (e.g., [1]).

  1. An evolutionary context for in-situ CMO growth. The nuclear configuration of IC 342: a young NSC surrounded by off-center star-forming clumps and fed by ongoing molecular gas inflow, captures the in-situ, secular growth of both CMOs in action. This places

IC 342 within a broader evolutionary sequence of NSC-mediated BH growth in disk galaxies. At the early end sits M33, whose well-studied NSC hosts at most a ≲(1.5–3)×103 M⊙\lesssim(1.5\text{--}3) \times10^{3}\ M_{\odot} BH [49, 90]. IC 342 represents the active phase, with both CMOs growing concurrently. The end state is exemplified by ω\omega Centauri [67] and B023-G078 [112]: massive clusters now classified as globular clusters, but widely regarded as the stripped nuclei of disrupted galaxies, each retaining an IMBH of mass comparable to that of IC 342. This sequence argues against the universality of massive (≳105 M⊙\gtrsim10^{5}\ M_{\odot}) BH seeds, suggesting instead that central BHs in at least some disk galaxies assemble gradually within their NSCs.

8. Scaling relations. Our MBHM_{\rm BH} measurement places IC 342 below but within the scatter of the MBHM_{\rm BH}–σ\sigma and MBHM_{\rm BH}–M⋆,galM_{\star,\rm gal} relations for spiral galaxies, consistent with the known systematic offset of bulgeless hosts from relations calibrated on classical-bulge and merger-built systems [77, 62]. Each direct dynamical MBHM_{\rm BH} at the IMBH mass scale carries disproportionate weight for constraining the low-mass end of these relations, where the scatter is largest and the shape remains poorly determined. This regime is of particular theoretical importance: light-seed models predict a wide mass distribution and high BH occupation fraction in low-mass galaxies, while heavy-seed models predict a higher minimum MBHM_{\rm BH} and lower occupation fraction [142, 84, 83]. As the sample of direct dynamical measurements in bulgeless spirals grows, currently populated by only a handful of systems including NGC 4395 [42], NGC 2748 [7], possibly NGC 5055 (T. H. T. Ho et al. submitted), and now IC 342, it will become possible to statistically discriminate between competing BH seeding scenarios directly from the observed distribution of MBHM_{\rm BH} at fixed host-galaxy properties.

9. Bulgeless spiral with a direct dynamical MBHM_{\rm BH}. The surface brightness decomposition of IC 342 shows no evidence for a classical bulge component, with all Sérsic indices n<1n < 1, confirming its bulgeless nature. IC 342 thus joins the rare class of bulgeless spirals with a direct dynamical MBHM_{\rm BH} determination: a regime currently populated by only a handful of galaxies including NGC 4395 (M. den Brok et al. 2015b) and NGC 2748 [7], and extends this census down to the IMBH mass scale (MBH∼105 M⊙M_{\rm BH} \sim10^{5}\ M_{\odot}). This is particularly significant because bulgeless systems are thought to follow a more stochastic evolutionary path decoupled from the tight co-evolutionary channel governing classical spheroids [77], yet direct MBHM_{\rm BH} measurements to test this prediction at the low-mass end remain extremely scarce.

Acknowledgements

DDN acknowledges support from the ELT postdoctoral research fellowship at the Department of Astronomy, University of Michigan. TNW would like to thank the ICISE staff for their enthusiastic support. This work is partially supported by a grant from the Simons Foundation to IFIRSE, ICISE (916424, N.H.). RS acknowledges support from the INAF grant number 1.05.23.04.04. IC’s research is supported by the SAO Telescope Data Center. He also acknowledges support from the NASA ADAP-22-0102 and HST-GO-16739 grants.

Some of the data presented herein were obtained at Keck Observatory, which is a private 501(c)3 non-profit organization operated as a scientific partnership among the California Institute of Technology, the University of California, and the National Aeronautics and Space Administration. The Observatory was made possible by the generous financial support of the W. M. Keck Foundation. These observations are associated with the Keck Program ID: 2024B_W294.

The authors wish to recognize and acknowledge the very significant cultural role and reverence that the summit of Maunakea has always had within the Native Hawaiian community. We are most fortunate to have the opportunity to conduct observations from this mountain.

Facility: Keck/OSIRIS and HST/WFC3.

Software: Python 3.12: [139], Matplotlib 3.6: [73], NumPy 1.22: [70], SciPy 1.3: [141], photutils 0.7: [19], AstroPy 5.1 [6], AdaMet 2.0 [30], JamPy 7.2.4 [24], MgeFit 6.0.4 [21], Profiler [33], and Profiler [34].

References

  1. [1]Amaro-Seoane, P., Audley, H., Babak, S., et al. 2017, arXiv e-prints, arXiv:1702.00786, doi: 10.48550/arXiv.1702.00786
  2. [2]Antonini, F., Barausse, E., & Silk, J. 2015a, ApJ, 812, 72, doi: 10.1088/0004-637X/812/1/72DOI
  3. [3]Antonini, F., Barausse, E., & Silk, J. 2015b, ApJL, 806, L8, doi: 10.1088/2041-8205/806/1/L8DOI
  4. [4]Antonini, F., Capuzzo-Dolcetta, R., Mastrobuono-Battisti, A., & Merritt, D. 2012, ApJ, 750, 111, doi: 10.1088/0004-637X/750/2/111DOI
  5. [5]Arca-Sedda, M., & Capuzzo-Dolcetta, R. 2014, MNRAS, 444, 3738, doi: 10.1093/mnras/stu1683DOI
  6. [6]Astropy Collaboration, Price-Whelan, A. M., Lim, P. L., et al. 2022, ApJ, 935, 167, doi: 10.3847/1538-4357/ac7c74DOI
  7. [7]Atkinson, J. W., Collett, J. L., Marconi, A., et al. 2005, MNRAS, 359, 504, doi: 10.1111/j.1365-2966.2005.08904.xDOI
  8. [8]Balcells, M., Graham, A. W., Domínguez-Palmero, L., & Peletier, R. F. 2003, ApJL, 582, L79, doi: 10.1086/367783DOI
  9. [9]Balcells, M., Graham, A. W., & Peletier, R. F. 2007, ApJ, 665, 1084, doi: 10.1086/519752DOI
  10. [10]Balser, D. S., Wenger, T. V., Goss, W. M., Johnson, K. E., & Kepley, A. A. 2017, ApJ, 844, 73, doi: 10.3847/1538-4357/aa7a01DOI
  11. [11]Bauer, F. E., Brandt, W. N., & Lehmer, B. 2003, AJ, 126, 2797, doi: 10.1086/379139DOI
  12. [12]Bekki, K., & Freeman, K. C. 2003, MNRAS, 346, L11, doi: 10.1046/j.1365-2966.2003.02725.x
  13. [13]Bekki, K., & Graham, A. W. 2010, ApJL, 714, L313, doi: 10.1088/2041-8205/714/2/L313DOI
  14. [14]Binney, J., & Mamon, G. A. 1982, MNRAS, 200, 361, doi: 10.1093/mnras/200.2.361DOI
  15. [15]Böker, T., Forster-Schreiber, N. M., & Genzel, R. 1997, AJ, 114, 1883, doi: 10.1086/118624
  16. [16]Böker, T., Laine, S., van der Marel, R. P., et al. 2002, AJ, 123, 1389, doi: 10.1086/339025DOI
  17. [17]Böker, T., Sarzi, M., McLaughlin, D. E., et al. 2004, AJ, 127, 105, doi: 10.1086/380911
  18. [18]Böker, T., van der Marel, R. P., & Vacca, W. D. 1999, AJ, 118, 831, doi: 10.1086/300985DOI
  19. [19]Bradley, L., Sipőcz, B., Robitaille, T., et al. 2024,, 2.0.2 Zenodo, doi: 10.5281/zenodo.13989456
  20. [20]Bruzual, G., & Charlot, S. 2003, MNRAS, 344, 1000, doi: 10.1046/j.1365-8711.2003.06897.xDOI
  21. [21]Cappellari, M. 2002, MNRAS, 333, 400, doi: 10.1046/j.1365-8711.2002.05412.xDOI
  22. [22]Cappellari, M. 2008, MNRAS, 390, 71, doi: 10.1111/j.1365-2966.2008.13754.xDOI
  23. [23]Cappellari, M. 2017, MNRAS, 466, 798, doi: 10.1093/mnras/stw3020DOI
  24. [24]Cappellari, M. 2020, MNRAS, 494, 4819, doi: 10.1093/mnras/staa959DOI
  25. [25]Cappellari, M. 2023, MNRAS, 526, 3273, doi: 10.1093/mnras/stad2597DOI
  26. [26]Cappellari, M. 2025, MNRAS, 544, 1432, doi: 10.1093/mnras/staf1726DOI
  27. [27]Cappellari, M. 2026, in Encyclopedia of Astrophysics, Volume 4, ed. I. Mandel, Vol. 4 (Amsterdam, The Netherlands: Elsevier), 122–152, doi: 10.1016/B978-0-443-21439-4.00109-7DOI
  28. [28]Cappellari, M., & Emsellem, E. 2004, PASP, 116, 138, doi: 10.1086/381875DOI
  29. [29]Cappellari, M., Neumayer, N., Reunanen, J., et al. 2009, MNRAS, 394, 660, doi: 10.1111/j.1365-2966.2008.14377.xDOI
  30. [30]Cappellari, M., Scott, N., Alatalo, K., et al. 2013, MNRAS, 432, 1709, doi: 10.1093/mnras/stt562DOI
  31. [31]Carson, D. J., Barth, A. J., Seth, A. C., et al. 2015, AJ, 149, 170, doi: 10.1088/0004-6256/149/5/170DOI
  32. [32]Chattopadhyay, D., Stegmann, J., Antonini, F., Barber, J., & Romero-Shaw, I. M. 2023, MNRAS, 526, 4908, doi: 10.1093/mnras/stad3048DOI
  33. [33]Ciambur, B. C. 2015, ApJ, 810, 120, doi: 10.1088/0004-637X/810/2/120DOI
  34. [34]Ciambur, B. C. 2016, PASA, 33, e062, doi: 10.1017/pasa.2016.60DOI
  35. [38]Davis, B. L., Graham, A. W., & Cameron, E. 2018, ApJ, 869, 113, doi: 10.3847/1538-4357/aae820DOI
  36. [39]Davis, T. A., Nguyen, D. D., Seth, A. C., et al. 2020, MNRAS, 496, 4061, doi: 10.1093/mnras/staa1567DOI
  37. [40]Debattista, V. P., Mayer, L., Carollo, C. M., et al. 2006, ApJ, 645, 209, doi: 10.1086/504147DOI
  38. [41]den Brok, M., Seth, A. C., Barth, A. J., et al. 2015a, ApJ, 809, 101, doi: 10.1088/0004-637X/809/1/101DOI
  39. [42]den Brok, M., Seth, A. C., Barth, A. J., et al. 2015b, ApJ, 809, 101, doi: 10.1088/0004-637X/809/1/101DOI
  40. [43]Dominiak, P., Bureau, M., Liang, F.-H., et al. 2026, MNRAS, 548, stag546, doi: 10.1093/mnras/stag546DOI
  41. [45]Emsellem, E., Monnet, G., & Bacon, R. 1994, A&A, 285, 723
  42. [46]Ferrarese, L., & Ford, H. 2005, SSRv, 116, 523, doi: 10.1007/s11214-005-3947-6DOI
  43. [47]Ferrarese, L., & Merritt, D. 2000, ApJL, 539, L9, doi: 10.1086/312838DOI
  44. [48]Gebhardt, K., Bender, R., Bower, G., et al. 2000, ApJL, 539, L13, doi: 10.1086/312840DOI
  45. [49]Gebhardt, K., Lauer, T. R., Kormendy, J., et al. 2001, AJ, 122, 2469, doi: 10.1086/323481DOI
  46. [50]Genzel, R., & Townes, C. H. 1987, ARA&A, 25, 377, doi: 10.1146/annurev.aa.25.090187.002113DOI
  47. [51]González Delgado, R. M., Pérez, E., Cid Fernandes, R., & Schmitt, H. 2008, AJ, 135, 747, doi: 10.1088/0004-6256/135/3/747DOI
  48. [52]Graham, A. W. 2008, ApJ, 680, 143, doi: 10.1086/587473DOI
  49. [53]Graham, A. W. 2016a, in IAU Symposium, Vol. 312, Star Clusters and Black Holes in Galaxies across Cosmic Time, ed. Y. Meiron, S. Li, F.-K. Liu, & R. Spurzеm, 269–273, doi: 10.1017/S1743921315008017
  50. [54]Graham, A. W. 2016b, in Astrophysics and Space Science Library, Vol. 418, Galactic Bulges, ed. E. Laurikainen, R. Peletier, & D. Gadotti, 263, doi: 10.1007/978-3-319-19378-6_11DOI
  51. [55]Graham, A. W. 2020, MNRAS, 492, 3263, doi: 10.1093/mnras/stz3547DOI
  52. [56]Graham, A. W. 2023a, MNRAS, 522, 3588, doi: 10.1093/mnras/stad1124DOI
  53. [57]Graham, A. W. 2023b, MNRAS, 518, 6293, doi: 10.1093/mnras/stac3173DOI
  54. [58]Graham, A. W. 2026, MNRAS, 549, stag782, doi: 10.1093/mnras/stag782DOI
  55. [59]Graham, A. W., Chilingarian, I., Nguyen, D. D., et al. 2025, PASA, 42, e068, doi: 10.1017/pasa.2025.10035DOI
  56. [60]Graham, A. W., Erwin, P., Caon, N., & Trujillo, I. 2001, ApJL, 563, L11, doi: 10.1086/338500DOI
  57. [61]Graham, A. W., & Guzmán, R. 2003, AJ, 125, 2936, doi: 10.1086/374992DOI
  58. [62]Graham, A. W., & Sahu, N. 2023, MNRAS, 518, 2177, doi: 10.1093/mnras/stac2019DOI
  59. [63]Graham, A. W., & Soria, R. 2019, MNRAS, 484, 794, doi: 10.1093/mnras/stv3068
  60. [64]Graham, A. W., & Spitler, L. R. 2009, MNRAS, 397, 2148, doi: 10.1111/j.1365-2966.2009.15118.xDOI
  61. [65]Greene, J. E., Peng, C. Y., Kim, M., et al. 2010, ApJ, 721, 26, doi: 10.1088/0004-637X/721/1/26DOI
  62. [66]Haario, H., Saksman, E., & Tamminen, J. 2001, Bernoulli, 7, 223
  63. [67]Häberle, M., Neumayer, N., Seth, A., et al. 2024, Nature, 631, 285, doi: 10.1038/s41586-024-07511-zDOI
  64. [68]Han, J. J., El-Badry, K., Lucchini, S., et al. 2025, ApJ, 982, 188, doi: 10.3847/1538-4357/adb967DOI
  65. [69]Häring, N., & Rix, H.-W. 2004, ApJL, 604, L89, doi: 10.1086/383567
  66. [70]Harris, C. R., Millman, K. J., van der Walt, S. J., et al. 2020, Nature, 585, 357, doi: 10.1038/s41586-020-2649-2DOI
  67. [71]Hernquist, L. 1992, ApJ, 400, 460, doi: 10.1086/172009DOI
  68. [73]Hunter, J. D. 2007, Computing In Science & Engineering, 9, 90, doi: 10.1109/MCSE.2007.55DOI
  69. [74]Husser, T.-O., Wende-von Berg, S., Dreizler, S., et al. 2013, A&A, 553, A6, doi: 10.1051/0004-6361/201219058DOI
  70. [75]Ibata, R. A., Bellazzini, M., Malhan, K., Martin, N., & Bianchini, P. 2019, Nature Astronomy, 3, 667, doi: 10.1038/s41550-019-0751-xDOI
  71. [76]Karachentsev, I. D., Sharina, M. E., Dolphin, A. E., et al. 2003, A&A, 404, 93, doi: 10.1051/0004-6361:20030170DOI
  72. [77]Kormendy, J., & Ho, L. C. 2013, ARA&A, 51, 511, doi: 10.1146/annurev-astro-082708-101811DOI
  73. [78]Krajnović, D., Cappellari, M., de Zeeuw, P. T., & Copin, Y. 2006, MNRAS, 366, 787, doi: 10.1111/j.1365-2966.2005.09902.xDOI
  74. [79]Lablanche, P.-Y., Cappellari, M., Emsellem, E., et al. 2012, MNRAS, 424, 1495, doi: 10.1111/j.1365-2966.2012.21346.x
  75. [80]Larkin, J., Barczys, M., Krabbe, A., et al. 2006, New Astron Rev, 50, 362, doi: 10.1016/j.newar.2006.02.005
  76. [82]Lockhart, K. E., Do, T., Larkin, J. E., et al. 2019, AJ, 157, 75, doi: 10.3847/1538-3881/aaf64eDOI
  77. [83]Lodato, G., & Natarajan, P. 2006, MNRAS, 371, 1813, doi: 10.1111/j.1365-2966.2006.10801.xDOI
  78. [84]Madau, P., & Rees, M. J. 2001, ApJL, 551, L27, doi: 10.1086/319848DOI
  79. [86]Magorrian, J., Tremaine, S., Richstone, D., et al. 1998, AJ, 115, 2285, doi: 10.1086/300353DOI
  80. [87]Mak, D. S. S., Pun, C. S. J., & Kong, A. K. H. 2008, ApJ, 686, 995, doi: 10.1086/591482
  81. [88]Marconi, A., Schreier, E. J., Koekemoer, A., et al. 2000, ApJ, 528, 276, doi: 10.1086/308168DOI
  82. [89]Massari, D., Koppelman, H. H., & Helmi, A. 2019, A&A, 630, L4, doi: 10.1051/0004-6361/201936135DOI
  83. [90]Merritt, D., Ferrarese, L., & Joseph, C. L. 2001, Science, 293, 1116, doi: 10.1126/science.1063896DOI
  84. [91]Milosavljević, M., & Merritt, D. 2001, ApJ, 563, 34, doi: 10.1086/323830DOI
  85. [92]Mitzkus, M., Cappellari, M., & Walcher, C. J. 2017, MNRAS, 464, 4789, doi: 10.1093/mnras/stw2677DOI
  86. [93]Neumayer, N., Seth, A., & Böker, T. 2020, A&A Rv, 28, 4, doi: 10.1007/s00159-020-00125-0DOI
  87. [94]Neumayer, N., & Seth, A. C. 2011, ApJ, 743, 120, doi: 10.1088/0004-637X/743/2/120
  88. [97]Ngo, H. N., Nguyen, D. D., Le, T. Q. T., et al. 2025c, ApJ, 992, 211, doi: 10.3847/1538-4357/ae0455DOI
  89. [98]Nguyen, D. D., Cappellari, M., & Pereira-Santaella, M. 2023, MNRAS, 526, 3548, doi: 10.1093/mnras/stad2860DOI
  90. [99]Nguyen, D. D., Seth, A. C., Reines, A. E., et al. 2014, ApJ, 794, 34, doi: 10.1088/0004-637X/794/1/34DOI
  91. [100]Nguyen, D. D., Seth, A. C., den Brok, M., et al. 2017, ApJ, 836, 237, doi: 10.3847/1538-4357/aa5cb4DOI
  92. [101]Nguyen, D. D., Seth, A. C., Neumayer, N., et al. 2018, ApJ, 858, 118, doi: 10.3847/1538-4357/aabe28DOI
  93. [102]Nguyen, D. D., Seth, A. C., Neumayer, N., et al. 2019, ApJ, 872, 104, doi: 10.3847/1538-4357/aafe7aDOI
  94. [103]Nguyen, D. D., den Brok, M., Seth, A. C., et al. 2020, ApJ, 892, 68, doi: 10.3847/1538-4357/ab77aaDOI
  95. [104]Nguyen, D. D., Izumi, T., Thater, S., et al. 2021, MNRAS, 504, 4123, doi: 10.1093/mnras/stab1002DOI
  96. [105]Nguyen, D. D., Bureau, M., Thater, S., et al. 2022, MNRAS, 509, 2920, doi: 10.1093/mnras/stab3016DOI
  97. [106]Nguyen, D. D., Cappellari, M., Ngo, H. N., et al. 2025a, AJ, 170, 124, doi: 10.3847/1538-3881/ade9baDOI
  98. [107]Nguyen, D. D., Ngo, H. N., Le, T. Q. T., et al. 2025b, A&A, 698, L9, doi: 10.1051/0004-6361/202554672DOI
  99. [109]Nguyen, D. D., Cappellari, M., Le, T. Q. T., et al. 2026a, MNRAS, 546, stag238, doi: 10.1093/mnras/stag238DOI
  100. [110]Nguyen, D. D., Ngo, H. N., Cappellari, M., et al. 2026b, ApJ, 999, 97, doi: 10.3847/1538-4357/ae32f2DOI
  101. [111]Nguyen, D. D., Le, T. N., Cappellari, M., et al. 2026c, ApJ, 1003, 98, doi: 10.3847/1538-4357/acef40
  102. [112]Pechetti, R., Seth, A., Kamann, S., et al. 2022, ApJ, 924, 48, doi: 10.3847/1538-4357/ac339fDOI
  103. [113]Pfeffer, J., Lardo, C., Bastian, N., Saracino, S., & Kamann, S. 2021, MNRAS, 500, 2514, doi: 10.1093/mnras/staa3407DOI
  104. [114]Pilyugin, L. S., Tautvaišienė, G., & Lara-López, M. A. 2023, A&A, 676, A57, doi: 10.1051/0004-6361/202346503DOI
  105. [115]Rantala, A., Rawlings, A., Naab, T., Thomas, J., & Johansson, P. H. 2024, MNRAS, 535, 1202, doi: 10.1093/mnras/stae2424DOI
  106. [117]Saglia, R. P., Opitsch, M., Erwin, P., et al. 2016, ApJ, 818, 47, doi: 10.3847/0004-637X/818/1/47DOI
  107. [118]Saha, A., Claver, J., & Hoessel, J. G. 2002, AJ, 124, 839, doi: 10.1086/341650
  108. [119]Sahu, N., Graham, A. W., & Davis, B. L. 2019a, ApJ, 876, 155, doi: 10.3847/1538-4357/ab0f32DOI
  109. [120]Sahu, N., Graham, A. W., & Davis, B. L. 2019b, ApJ, 887, 10, doi: 10.3847/1538-4357/ab50b7DOI
  110. [121]Sandage, A., Binggeli, B., & Tammann, G. A. 1985, AJ, 90, 1759, doi: 10.1086/113875DOI
  111. [122]Schinnerer, E., Böker, T., & Meier, D. S. 2003, ApJL, 591, L115, doi: 10.1086/377118DOI
  112. [123]Schlafly, E. F., & Finkbeiner, D. P. 2011, ApJ, 737, 103, doi: 10.1088/0004-637X/737/2/103DOI
  113. [124]Schödel, R., Feldmeier, A., Kunneriath, D., et al. 2014, A&A, 566, A47, doi: 10.1051/0004-6361/201423481DOI
  114. [125]Scott, N., & Graham, A. W. 2013, ApJ, 763, 76, doi: 10.1088/0004-637X/763/2/76DOI
  115. [126]Serabyn, E., & Lacy, J. H. 1985, ApJ, 293, 445, doi: 10.1086/163250DOI
  116. [127]Seth, A. C., Blum, R. D., Bastian, N., Caldwell, N., & Debattista, V. P. 2008, ApJ, 678, 116, doi: 10.1086/533955
  117. [128]Seth, A. C., Dalcanton, J. J., Hodge, P. W., & Debattista, V. P. 2006, AJ, 132, 2539, doi: 10.1086/508994DOI
  118. [129]Seth, A. C., van den Bosch, R., Mieske, S., et al. 2010, Nature, 466, 1077, doi: 10.1038/nature09255
  119. [130]Silge, J. D., & Gebhardt, K. 2003, AJ, 125, 2809, doi: 10.1086/375324DOI
  120. [131]Simon, D. A., Cappellari, M., & Hartke, J. 2024, MNRAS, 527, 2341, doi: 10.1093/mnras/stad3309DOI
  121. [132]Smith, M. D., Bureau, M., Davis, T. A., et al. 2019, MNRAS, 485, 4359, doi: 10.1093/mnras/stz625DOI
  122. [133]Thater, S., Lyubenova, M., Fahrion, K., et al. 2023, A&A, 675, A18, doi: 10.1051/0004-6361/202245362DOI
  123. [134]Thater, S., Krajnović, D., Weilbacher, P. M., et al. 2022, MNRAS, 509, 5416, doi: 10.1093/mnras/stab3210DOI
  124. [135]Tremaine, S. D., Ostriker, J. P., & Spitzer, Jr., L. 1975, ApJ, 196, 407, doi: 10.1086/153422DOI
  125. [139]Van Rossum, G., & Drake, F. L. 2009, Python 3 Reference Manual (Scotts Valley, CA: CreateSpace)
  126. [140]Verro, K., Trager, S. C., Peletier, R. F., et al. 2022, A&A, 660, A34, doi: 10.1051/0004-6361/202142388DOI
  127. [141]Virtanen, P., Gommers, R., Oliphant, T. E., et al. 2020, Nature Methods, 17, 261, doi: 10.1038/s41592-019-0686-2DOI
  128. [142]Volonteri, M., Lodato, G., & Natarajan, P. 2008, MNRAS, 383, 1079, doi: 10.1111/j.1365-2966.2007.12589.xDOI
  129. [143]Wizinowich, P. L., Le Mignant, D., Bouchez, A. H., et al. 2006, PASA, 118, 297, doi: 10.1086/499290DOI
  130. [145]Zaw, I., Rosenthal, M. J., Katkov, I. Y., et al. 2020, ApJ, 897, 111, doi: 10.3847/1538-4357/ab9944DOI
  131. [146]Zhang, C., Wang, J., Tan, Q.-H., et al. 2024, ApJL, 967, L25, doi: 10.3847/2041-8213/ad4a69DOI
  132. [147]Zinnecker, H., Keable, C. J., Dunlop, J. S., Cannon, R. D., & Griffiths, W. K. 1988, in IAU Symposium, Vol. 126, The Harlow-Shapley Symposium on Globular Cluster Systems in Galaxies, ed. J. E. Grindlay & A. G. D. Philip, 603

Paper details

vlnc:W_5uzef655
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 0

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Physics and Astronomy

Not recorded for this paper

View statement
Looking for related papers…
Contents