EDP Sciences logo
Subscriber Authentication Point
Search
Menu
Home All issues Volume 694 (February 2025) A&A, 694 (2025) A301 Full HTML
Open Access
Issue
A&A
Volume 694, February 2025
Article Number A301
Number of page(s) 11
Section Extragalactic astronomy
DOI https://doi.org/10.1051/0004-6361/202452640
Published online 20 February 2025

© The Authors 2025

Licence Creative CommonsOpen Access article, published by EDP Sciences, under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

This article is published in open access under the Subscribe to Open model. Subscribe to A&A to support open access publication.

1. Introduction

Ultraluminous X-ray sources (ULXs) are a distinct subclass of X-ray binaries characterized by an isotropic X-ray luminosity (LX) exceeding 1039 erg s−1, which surpasses the Eddington limit for a 10 M⊙ black hole (BH). There is a growing consensus that ULXs are predominantly powered by supercritical accretion onto neutron stars (NSs) or stellar-mass BHs, as discussed in recent reviews (Kaaret et al. 2017; Fabrika et al. 2021; King et al. 2023; Pinto & Walton 2023). However, some studies propose the existence of intermediate-mass BHs within the range 102 to 104 M⊙ accreting at sub-Eddington rates (Roberts 2007; Sutton et al. 2012; Caballero-García et al. 2013; Pasham & Strohmayer 2013). In the case of supercritical accretion, geometric beaming and/or super-Eddington accretion onto stellar-mass compact objects could explain their high luminosities (King 2009; Middleton et al. 2015; Walton et al. 2018). Temporal variability is a key characteristic of ULXs; quasi-periodic oscillations and coherent pulsations are observed, providing insights into the nature of the compact object (Agrawal & Nandi 2015; Bachetti et al. 2014; Carpano et al. 2018; Rodríguez Castillo et al. 2020). Furthermore, the detection of cyclotron resonance scattering features in the X-ray spectrum of a ULX suggests the presence of a NS candidates (Brightman et al. 2018; Kong et al. 2022).

Moreover, identifying optical candidates and analyzing their properties to constrain the age, mass, and possible spectral type of the donor star makes important contributions to ULX studies. The optical emission observed in ULXs can arise from the accretion disk, the donor star, or a combination of the two, though many studies suggest that this emission is often dominated by reprocessed radiation from an irradiated accretion disk (Tao et al. 2011; Soria et al. 2012; Sutton et al. 2014; Ambrosi & Zampieri 2018; Yao & Feng 2019; Allak et al. 2022a; Allak 2024). Long-term optical light curves have revealed sinusoidal modulations, indicating super-orbital or orbital periods for some optical counterparts (Liu et al. 2009; Allak et al. 2022a; Allak 2022). Identifying donor stars in ULXs is challenging due to the faintness of their apparent magnitudes and their locations in crowded, star-forming regions, necessitating high spatial resolution detectors such as the James Webb Space Telescope (JWST; Allak 2024). Some ULX counterparts, particularly those bright in the near-infrared (NIR) band, have been suggested to be red supergiants (RSGs), though recent studies indicate that many NIR counterparts may not be sufficiently red to be RSGs, possibly due to unresolved sources in distant galaxies (Heida et al. 2014; Allak 2023). Additionally, ULXs without detectable optical counterparts in previous studies may still have counterparts in JWST images, possibly hidden by hot dust or circumbinary disks (Chandar et al. 2020; Allak 2024).

This study aims to identify the IR and optical counterparts of the eight ULXs identified by Ma et al. (2023) in NGC 1559. NGC 1559 is a barred spiral galaxy located at a distance of 12.6 Mpc, as determined using the Tully–Fisher relation (Tully et al. 2013). Although a wide range of distance values (9–23 Mpc) have been determined for this galaxy using different methods in the NED1, we adopted the value of 12.6 Mpc for comparison with the X-ray analysis in Ma et al. (2023). The galaxy is also notable for having hosted four supernovae in the last 40 years, including SN 1986L, which has been a primary focus of Chandra observations. The galaxy, classified as SB(s)cd, features fragmented spiral arms with high star formation rates but likely lacks an active galactic nucleus (AGN; de Vaucouleurs et al. 1991). In this work we investigated the potential donor stars of ULXs by utilizing data from JWST and the Hubble Space Telescope (HST), with a focus on their masses and ages, as well as the emission mechanisms within these systems. Additionally, we conducted a detailed study of the long-term variability and energy spectra of ULXs in NGC 1559 using Chandra and Swift/XRT.

The paper is organized as follows: Sect. 2 presents the properties of the galaxy NGC 1559, along with the target ULXs, and provides details of the observations. Data reduction and analysis of X-ray and multiwavelength observations are discussed in Sect. 3. Section 4 presents results and discusses the properties of the ULXs. Finally, Sect. 5 summarizes the main findings of this study.

2. Target sources and observations

2.1. Ultraluminous X-ray sources in NGC 1559

The primary target sources of this study are the eight ULXs (X-1, X-3, X-5, X-6, X14, X-17, X-18, and X-24) displayed in Fig. 1. In Ma et al. (2023), ULXs were defined with the criterion LX > 1039 erg s−1, which we adopted in this study. These sources exhibit X-ray luminosities ranging from a minimum of 1039 erg s−1 to of a maximum 7.98 × 1039 erg s−1 in the (0.3–7) keV energy band and, along with various spectral properties, as detailed in their work. Below, we summarize their characteristics.

thumbnail Fig. 1.

False-color Chandra image of the galaxy NGC 1559. The energy ranges used are highlighted. The image has been smoothed using a 5 arcsec Gaussian, and the ULXs are indicated with white circles.

X-3, X-5, X-14, and X-17 all display hard spectra with photon indices (Γ) between 1.47 and 1.85, indicative of nonthermal emission consistent with the presence of a compact object such as a BH or NS. Among these sources, X-3 has the lowest LX value at ∼2 × 1039 erg s−1, while X-17 has the highest value at ∼8 × 1039 erg s−1. In contrast, X-6 and X-18 present softer spectra with Γ = 3.16 and 2.34, respectively, suggesting the presence of a thermal component or distinct accretion mechanisms. X-6, with a luminosity of ∼1 × 1039 erg s−1, stands out due to its particularly soft spectrum, while X-18, at ∼3.6 × 1039 erg s−1, may represent a different emission mechanism linked to its accretion environment.

One of the remaining two ULXs, X-1, has an LX of ∼4 × 1039 erg s−1 and Γ is 1.8, indicating a hard X-ray spectrum. The source also shows long-term variability on a timescale of 10 to 100 days. The isolated location of X-1 within NGC 1559 reduces the risk of contamination from nearby sources. The other ULX,X-24, is particularly noteworthy due to its LX of ∼3.7 × 1039 erg s−1 and the detection of a periodicity around 7500 s. This periodicity is indicative of an orbital period within a compact binary system, suggesting that X-24 may host a stellar-mass BH or NS. The quasi-sinusoidal variation observed in the light curve strengthens the hypothesis that X-24 represents a unique case of a ULX within NGC 1559. In this study, X-1 and X-24 are given priority because of their distinct characteristics. X-1 displays high luminosity and variability, while X-24 exhibits a notable periodicity.

2.2. X-ray and multiwavelength observations of NGC 1559

NGC 1559 was observed by Swift/XRT a total of 56 times between 2005 and 2016 (target IDs 30252, 34132, 35896, and 92169), with exposure times ranging from 25.1 s to 5302.3 s. Additionally, the galaxy was observed for 46 ks by the Chandra Advanced CCD Imaging Spectrometer (ACIS) in 2016 (Obs ID 16745). These observations provide comprehensive data for studying the X-ray sources in NGC 1559, with a particular focus on the potential variability of ULXs over time. Besides the Chandra and Swift/XRT observations, NGC 1559 was also observed by XMM-Newton with a 14.4 ks exposure. However, this observation was excluded from the current study, as no new findings were obtained beyond those reported in the study by Ma et al. (2023).

The NGC 1559 was observed by HST employing the Wide Field Camera 3 (WFC3)/UVIS1 in 2015, 2017, and 2019. Moreover, the Gaia2 source catalog was used for astrometry calculations in this study. The JWST observed the spiral galaxy NGC 1559 from June to September 2023. The galaxy was imaged using the JWST/NIRCam with filters F090W, F150W, F277W, F300M, and F335M, as well as the JWST/MIRI with F770W and F2100W filters. We noted that the ULXs in this galaxy are located in highly crowded regions, where sources that are distinguishable in NIRCam images often blend in JWST/MIRI images. The details of all the HST/WFC3/UVIS1 observations are given in Table A.1. Red-green-blue images of the galaxy from HST and JWST are displayed in Fig. 2. Moreover, between 2017 and 2021, NGC 1559 was observed 13 times using the HST/WFC3/IRF160W filter (Proposal IDs 15145 and 16250). These observations enabled us to investigate long-term IR variations for counterparts.

thumbnail Fig. 2.

Red-green-blue images of NGC 1559 from HST (left) and JWST (right). The filters used for the HST image are F814W, F555W, and F438W, and for JWST they are F355M, F300W, and F275W, respectively. The ULXs are marked with red circles on both images. Note that the X-1 source is outside the area shown in the JWST image.

3. Data reduction and analysis

3.1. Source detection and photometry

For all drizzled images HST/WFC3/UVIS1 of this galaxy, point-like sources were detected with the daofind task, and aperture photometry of these sources was performed using the DAOPHOT package Stetson (1987) in IRAF3 (Image Reduction and Analysis Facility). To perform aperture photometry, 3 pixels (0 . ″ $ {{\overset{\prime\prime}{.}}} $15) aperture radius and for the background, nine pixels were chosen. The Vega magnitudes were derived from instrumental magnitudes using zero-point magnitudes taken from the WFC3/UVIS1 and IR was taken from the study Calamida et al. (2022). To derive the aperture corrections, an approach similar to that of Allak (2023) was followed. Moreover, to determine the positions and photometry of the sources, the photometry tools provided by the PHOTUTILS V1.82 package4 Astropy were used. For the JWST data, the process included background estimation, source detection, and photometric analysis, following the methodology outlined in Allak (2023). Given the significant variation in background levels across the images, the PHOTUTILS background task was applied to estimate local background levels. Each detected source was identified as being more than 3σ above the local background. Aperture photometry was then conducted using a circular aperture with a 3-pixel radius, with the background contribution subtracted based on an annulus positioned nine pixels away from the source center.

3.2. Astrometry and determination of counterparts

Precise astrometry is necessary to determine both the NIR and optical counterparts of the eight ULXs, whose positions are shown in Fig. 1. Optical counterparts of ULXs can be investigated using HST and Chandra observations, owing to their excellent spatial resolution. The reference sources were searched by comparing Chandra image with HST observations. For this, 32 X-ray sources were detected in Chandra ACIS-S image by using wavdetect tool in Chandra Interactive Analysis of Observations (CIAO)5. Wavelets of 2, 4, 8, and 16 pixels were used, along with a detection threshold of 10−6. Only 27 out of 32 X-ray sources matched the HST images. The 27 X-ray sources were compared with optical point sources, and only one reference source, ULX X-5, was found for astrometric calculations. Therefore, Gaia Data Release 3 (DR3)6 optical source catalog was compared with the Chandra X-ray sources since the single source is less reliable for astrometric calculations. Taking account of the same shift directions, including the ULX X-5, three reference sources were found between both Chandra-Gaia. The astrometric offsets between Chandra and Gaia were found as − 0 . ″ 33 ± 0 . ″ 17 $ -0{{{\overset{\prime\prime}{.}}}}33 \pm 0{{{\overset{\prime\prime}{.}}}}17 $ for RA and − 0 . ″ 27 ± 0 . ″ 01 $ -0{{{\overset{\prime\prime}{.}}}}27 \pm 0{{{\overset{\prime\prime}{.}}}}01 $ for Dec with 1σ errors.

Furthermore, to find the counterparts of the ULXs, Gaia sources were compared with the sources identified in the HST images, and at least eight references with matching positions were identified. Positions of all the references used for astrometric calculations are displayed on the HST F555W image in Fig. 3. Astrometric offsets between Gaia and HST were found as − 0 . ″ 006 ± 0 . ″ 001 $ -0{{{\overset{\prime\prime}{.}}}}006 \pm 0{{{\overset{\prime\prime}{.}}}}001 $ for RA and − 0 . ″ 010 ± 0 . ″ 001 $ -0{{{\overset{\prime\prime}{.}}}}010 \pm 0{{{\overset{\prime\prime}{.}}}}001 $ for Dec with 1σ errors. Following our study of Allak (2022), the positional error radius was derived as 0.38 arcsec at a 90% confidence level. This radius corresponds to 23.33 pc based on the adopted distance of NGC 1559 (1 arcsec  ≃  61 pc; Tully et al. 2013). Moreover, relative astrometry was performed between NIRCam and HST images to determine the NIR counterparts using Gaia DR3 source catalog. After adequate reference sources were identified, offsets between JWST/NIRCam – HST/UVIS1 were derived as − 0 . ″ 023 $ -0{{{\overset{\prime\prime}{.}}}}023 $ and 0 . ″ 023 $ 0{{{\overset{\prime\prime}{.}}}}023 $, respectively at 1σ confidence level. Additionally, corrections were made by performing relative astrometry both between the filters of HST and within the filters of JWST. Finally, the X-ray positions were corrected on the basis of all these shifts. The Chandra coordinates and corrected coordinates of the ULXs are given in Table A.2.

thumbnail Fig. 3.

Positions of the Chandra (red circle) and Gaia (green box) reference sources used in astrometric calculations shown in the HST image. The solid yellow circle (radius of 1 arcmin) indicates the area within which reference sources were searched. Here the y- and x-axes indicate declination (Dec) and right ascension (RA), respectively.

As a result of astrometric calculations, a unique optical counterpart was determined for ULXs X-1, X-14, X-18, and X-24 while no optical source was detected at the corrected X-ray position of the ULX X-6. The remaining ULXs, X-3 and X-17, likely have multiple sources that exhibit unresolved or extended morphologies; therefore, they were not included in detailed analyses. A bright source was detected at the position of the ULX X-5, which matches the position of Gaia DR3 source (source ID: 4676459695824016128) and 2MASS (Two Micron All Sky Survey) (J04173037-6247258; one of the sources used for astrometric calculations) was not considered an optical counterpart. The corrected positions of ULXs and Chandra source detection error ellipses are shown on the HST images, Fig. 4, and photometric results are given Table 1. In the case of the JWST NIR counterparts, sources matching the optical counterparts of X-14 and X-24 were identified, which were also identified in the HST/WFC3/IRF160W images. Since the X-1 source was out of the field in both JWST and HST IR observations, it is unclear whether it has an IR counterpart. No detectable source was identified at the position of the remaining ULXs. NIR counterparts were not identified in the JWST F770W and F2100W images due to the band feature and poor pixel resolution (0.111 pixels/arcsec); thus, these images were only used to assess the location of ULXs.

thumbnail Fig. 4.

HST F555W image of corrected X-ray positions (solid red circles) and Chandra source detection error ellipses (dashed blue ellipses) of the ULXs. Since X-1 is not observed in the HST F555W filter, its position is shown on the F438W image. The numbers shown for the color bars are in units of electrons s−1. Solid green circles within the error radii of X-5 indicate Gaia sources. All panels are to the same scale, with north up and east to the left.

Table 1.

Vega magnitudes of the optical and IR counterparts of four ULXs.

Additionally, the FX/Fopt ratio is a useful diagnostic tool for identifying whether a source could be an AGN, based on the relative flux of its X-ray emission (FX) compared to its optical emission (Fopt; Maccacaro et al. 1982; Stocke et al. 1991; Aird et al. 2010). Although there are no simultaneous optical and X-ray observations when considering the closest dates (within a 1-year interval), the high ratios for the counterparts of X-1, X-14, X-18, and X-24 exceed the value suggested for AGN or BL Lac types. Furthermore, as ULXs are found in crowded regions, it is likely that we detected false positive counterparts. To assess the number of false positive counterparts (NFP), the false positive rates were derived using a method similar to the study of Allak (2024) (and reference therein). This approach involves calculating the probability of random matches within a specified positional error radius, allowing for the estimation of false positive identifications for a given set of counterparts. The average false positive rate is ∼5.7% for optical counterparts and ∼4.2% for NIR counterparts indicating that the counterparts determined in the error radius in this study are reliable.

3.3. Variability of the counterparts

Given the relatively sufficient multi-epoch HST and JWST observations of the NGC 1559 galaxy (see Table A.1), we could place constraints on the variability of the observed emission from the counterparts. Therefore, through multiwavelength observations of NGC 1559, we focused on the optical and IR variability of ULXs X-14, X-18, and X-24, which have unique optical counterparts. For ULX X-14, no variability was observed across HST optical, IR, and JWST NIR bands. No variation was observed in the optical emission of X-18, and no sources were detected at the position of ULX X-18 in the HST IR or JWST/NIRCam images. The optical emission of the counterpart of X-24 remained nearly constant but notable (≤0.2 mag) nonperiodic variations were observed in the light curve of HST IR (F160W) and JWST NIR (F277W) observations (September 2017 to October 2021), the long-term light curves for the X-14 and X-24 sources are displayed in Fig. 5.

thumbnail Fig. 5.

Long-term light curves for X-14 (above) and X-24 (below) using 13 observations taken with the HST F160W IR images. The dashed red lines represent the average magnitudes derived from these observations.

3.4. SEDs and CMDs

Spectral energy distributions (SEDs) of the optical counterparts were constructed to obtain the spectral characteristics of the counterparts using flux derived from values given in Table 1. The wavelengths of the filters are selected as the pivot wavelength, obtained from PYSYNPHOT7. We attempted to constrain the nature of possible donor stars by fitting optical SEDs with a blackbody or power-law (F ∝ λα) model. Physically acceptable parameters were obtained only for X-1, X-18, and X-24. The SEDs of X-1 and X-18 were adequately fitted by power-law models with spectral indices (α) of −0.66 ± 0.11 and −2.1 ± 0.13, respectively, in the case of X-24, the SED is adequately represented by a blackbody model with a temperature of 7000 K. The number of degrees of freedom for X-18 and X-24 is three and for X-1 is two. Optical SEDs for counterparts X-1, X-18, and X-24 are displayed in Fig. 6.

thumbnail Fig. 6.

SEDs of four counterparts of X-1, X-18, and X-24 (from top to bottom). The power-law and blackbody models for ULXs X-1, X-18, and X-24 are shown with dashed red, blue, and green lines, respectively. All data are shown with filled black circles. The data errors, taken as a systematic 0.05 mag, match the symbol size.

Studying the stellar population around the ULX provides valuable insights into the characteristics of the donor star. Stars in the same region likely formed at similar times, allowing more reliable estimates of the properties of donor candidates. Using color-magnitude diagrams (CMDs) of nearby stars, this approach helps constrain the age and mass of the donor stars. Optical photometric and SED results suggested that their donor stars could dominate the optical emission for X-14 and X-24. Therefore, two CMDs were derived for the optical counterparts of X-14 and X-24: F438W (B) versus F438W−F555W (B–V) and F555W (V) versus F555W−F814W (V–I) (see Fig. 7). Isochrones were obtained using solar metallicity of 0.02 and PARSEC models (Bressan et al. 2012). The distance modulus was calculated as 30.5 mag, corresponding to the adopted distance of 12.6 Mpc.

thumbnail Fig. 7.

HST/WFC3 CMDs for counterparts of X-14 (filled star) and X-24 (filled triangle). The isochrones were corrected for an extinction of AV = 0.04 mag.

3.5. Swift/XRT data analysis of X-1

The extensive set of Swift/XRT observations allows us to conduct a detailed study of both the long-term variability and energy spectra of X-1, as it is the only isolated source without contamination among the eight ULXs (see Fig. 8). Although the study by Ma et al. (2023) mentioned the X-ray properties of X-1, we performed a more detailed analysis of this source, emphasizing its multiwavelength characteristics and long-term X-ray evolution. By Utilizing 56 Swift/XRT photon-counting (PC) mode observations, the temporal variability and spectral characteristics were examined. To obtain the source count rate, we used HEASoft together with Swift/XRT data products generator8 (Evans et al. 2007, 2009). For the long-term X-ray light curve and the time-averaged spectrum of X-1, circular extraction regions with radii of 15 arcsec for the source and 45 arcsec for the background were used. The source energy spectra were grouped using the FTOOLS grppha with at least 15 counts per energy bin. Spectral fits were performed on all available 0.3–10 keV spectral data to determine the best-fitting model for the time-averaged spectrum using XSPEC v12.13. A hydrogen column density (NH) component (tbabs) was included in all cases.

thumbnail Fig. 8.

Stack Swift/XRT image of the galaxy NGC 1559. The image is smoothed with a 2 arcsec Gaussian. The ULXs are indicated with red circles.

Using all Swift/XRT observations, a long-term light curve of X-1 was constructed, displayed in Fig. 9. The ratio of maximum to minimum count rates for X-1 was determined to be one order of magnitude in the range 0.3–10 keV. Since the Swift/XRT observations are not continuous (e.g., observation dates 2005, 2007, 2016), the light curve was divided into three epochs: 2005 (epoch-1), 2007 (epoch-2), and 2016 (epoch-3) to investigate possible periodic variations. As seen in panel C of Fig. 9, a variation worth considering is noticeable only for epoch-3. Although all observations were used to investigate periodic variation, none was detected. Using the Lomb–Scargle periodogram (Lomb 1976; Scargle 1982), only in epoch-3, a low amplitude periodic modulation of 130.5 days was detected with a false alarm probability of ∼10−4. To investigate whether this low-amplitude periodicity was due to random variability, following the procedures outlined in Allak (2022), a total of 2000 simulated red-noise time series were generated for each filter using Monte Carlo simulations, ensuring they had the same number of sampling points and parameter values. This approach provides a robust assessment of significance against potential false detections. To model the red-noise continuum, the Lomb–Scargle periodogram was applied to each simulated light curve, using a power-law index ranging from −1 to −2 and employing a multifrequency approach. According to Monte Carlo analysis, the identified period of 130.5 days exhibits a significance level close to 2σ (see Fig. 9), which makes the reliability of this period for the X-1 system questionable.

thumbnail Fig. 9.

Light-curves of the X-1. Panel a: Long-term Swift/XRT light curve of X-1. The epoch-1 (panel b), epoch-2, and epoch-3 (panel c) observations are indicated by red-, blue-, and black-filled circles, respectively. Panel d: Lomb–Scargle periodograms of X-1. The periodic signal that peaks at 130.5 days (dashed red line) was found for epoch-3.

The ULX X-1 is well-suited for applying a time-averaged spectrum due to its long-term variability and the low signal-to-noise ratio often present in individual observations. Consequently, the time-averaged spectrum for X-1 was obtained using data from 56 Swift/XRT observations and thoroughly analyzed to identify the well-fitted spectral models. The spectrum of X-1, along with the instrument response and ancillary files, was generated based on Swift/XRT observations. Due to the averaged source counts, the energy spectrum was grouped using FTOOLS grppha, with at least 15 counts per energy bin. Spectral fits were then conducted over the 0.3–10 keV range to identify the best-fitting model for the time-averaged spectrum, utilizing XSPEC v12.13. In all cases, a hydrogen column density (NH) component (tbabs) was applied as part of the spectral modeling. Among the models tested, the power-law and diskbb models were found to be the best fits for the energy spectrum of X-1. The detailed parameters for these models are presented in Table 2 and the time-averaged spectrum is displayed in Fig. 10.

Table 2.

Power-law and diskbb model parameters of X-1 from the Swift/XRT time-averaged spectrum.

thumbnail Fig. 10.

Swift/XRT time-averaged energy spectrum of X-1 with diskbb (above) and power-law (below) models.

4. Results and discussion

Following astrometric corrections, ULX counterparts of the eight previously identified ULXs were determined based on HST and JWST observations. Accordingly, a unique optical counterpart was identified for X-1, X-14, X-18, and X-24, while no source was detected for X-6, even at the 1σ confidence level. The presence of multiple optical and NIR sources within the astrometric error radius of both X-3 and X-17 makes it difficult to identify the most likely counterparts, so these sources were not discussed further. For X-5, a unique source was detected that is too extended to be an optical counterpart; it was matched with a Gaia and 2MASS source (see Sect. 3.2). On the other hand, X-1 is not covered by the JWST or HST IR observation fields, and no IR counterpart(s) was detected within the error radius of X-18. For the remaining ULXs, X-14 and X-24, unique IR counterparts matching their optical counterparts were identified. The results obtained in this study and their discussion are given individually for each ULX as follows.

4.1. ULX X-1

ULX X-1 stands out as the one source among the eight that is isolated and free from contamination by nearby sources in all observations. The long-term light curve showed significant count rate variability (one order of magnitude) consistent with the results of Ma et al. (2023). Additionally, low amplitude, low confidence (≃2σ) potential periodic variability of 130.5-d was identified only in 2016 observations (epoch-3; see Fig. 9).

The nature of this variability can be interpreted in the context of super-orbital periods observed in ULXs (e.g., Brightman et al. 2019, 2020). Super-orbital periods in ULXs are mainly driven by precessing accretion disks, which can be affected by gravitational interactions with companion stars and varying mass transfer rates from those stars (King et al. 2023). The low false alarm probability value of ∼10−4 indicates that this periodic variation is unlikely to result from random variability. However, the Monte Carlo simulations evaluating this variation at the 3σ confidence level do not provide robust support for its classification as a real periodic variation. If the variation were random, it would align with the irregular and sometimes periodic fluctuations observed in AGNs (Buisson et al. 2017). Additional observations are needed to confirm whether the periodic variation is physically meaningful.

The ULX X-1 in NGC 1559 displays spectral features that strongly suggest accretion onto a compact object, likely a stellar-mass BH or NS, with super-Eddington accretion processes playing a significant role. The power-law model applied to the X-ray spectrum reveals an α = 1.60, indicating a hard X-ray spectrum typically associated with nonthermal emission, likely from high-energy processes occurring near the compact object. The parameters are also consistent with the values obtained by Ma et al. (2023). Furthermore, the calculated unabsorbed LX ∼ 7.87 × 1039 ergs s−1 in the 0.3–10 keV energy band. Additionally, when modeled with a diskbb, the inner disk temperature of 1.69 keV was found, pointing to a hot accretion disk, typically associated with very high accretion rates. This model also yields an LX ∼ 6.75 × 1039 erg s−1 in the 0.3–10 keV energy range. The X-ray energy spectrum fits the diskbb and power-law models, with nearly identical fit statistics for both. Therefore, it is challenging to distinguish between the two models. On the other hand, a hard power-law and a thermal disk component may suggest a complex accretion environment, where super-Eddington accretion may be accompanied by outflows or relativistic jets (King et al. 2001). These findings suggest that this ULX is powered by a compact stellar object, where the quite high LX and spectral features are driven by super-Eddington accretion onto a BH or NS.

To further investigate the long-term spectral evolution of ULX X-1, including a possible spectral state transition, temporal variations of the hard and soft count rates and a hardness-intensity diagram were plotted using Swift/XRT observations (see Fig. 11). The hardness ratios (hard/soft) were calculated based on the hard and soft count rates in the energy ranges of 1–4 keV and 0.3–1 keV, respectively. The sinusoidal-like variation observed in epoch-3 for X-1 in the 0.3–10 keV energy range was not evident in the soft band while a similar variation was observed in the hard band (see Fig. 11). The precession of the magnetic axis of the NS could also produce periodic modulations in the X-ray emission (Link 2003; Wasserman 2003; Townsend & Charles 2020). Due to the insufficient availability of high-quality data, it remains uncertain whether X-1 undergoes spectral state transitions (see panel c of the Fig. 11). If the source indeed displays spectral state transitions, the observed variability could originate from high-energy photons emitted from the innermost regions of the accretion disk (Kotze & Charles 2012) or from the jet precession of the BH (Abraham 2018; Cui et al. 2023). According to this scenario, the X-ray energy spectrum is more appropriately described by the diskbb model (with Tin = 1.69 keV).

thumbnail Fig. 11.

X-ray soft (0.3–1 keV) and hard (1–4 keV) count rates vs. time diagrams (upper two panels) and the hardness-intensity diagram (lower panel) of X-1.

In the case of the optical counterpart of X-1, the optical SED was adequately fitted by a power-law model with a spectral index of α ∼ −0.66. This suggests that the nonthermal emission from the accretion disk or jets is dominant, providing a clue that the accretion disk significantly contributes to the optical emission (Tao et al. 2011; Allak et al. 2022b). We emphasize that the possible donor star of X-1 is quite faint in the UVIS1 bands (> 24.3 mag). We calculated its absolute magnitude in the F606W filter to be ∼ − 6, which falls within the range of known optical candidates for ULXs ( − 3 < Mv < −8) (Fabrika et al. 2015; Vinokurov et al. 2018). Although the nature of the donor star candidate cannot be further discussed due to the lack of IR wavelength data, it could still have a sufficiently large mass to be detectable at this distance (12.6 Mpc), suggesting that the system may be a high-mass X-ray binary (HMXB; Chandar et al. 2020).

4.2. ULX X-14

The availability of data from different epochs in the same filters (e.g., JWST F275W and HST F555W) for the counterparts enabled us to place constraints on the optical and IR variability. Accordingly, it was determined that the optical emission of X-14 was not significantly changed (in all cases ≤0.1 mag). Moreover, we found that the (N)IR counterpart matching the optical counterpart shows no variability in emission in either the HST IR or JWST/NIRCam observations. This suggests that the donor star is the primary contributor to the optical and IR emissions. Moreover, the constant emission at the IR wavelength may indicate the absence of a variable jet. However, an SED could be created using optical and IR data, but a model with physically meaningful parameters could not be fitted. In addition, the CMDs plotted with the donor star candidate of X-14 and the surrounding stellar candidates indicate that the most likely age of the donor candidate is 7 Myr, with a mass of ≃18 M⊙. The donor candidate with an MF555W ≃ −8 indicates that it is an HMXB at this distance of the NGC 1559.

4.3. ULX X-18

A possible donor star of X-18, similar to X-1, is faint in the optical bands but relatively bright in the UV band with calculated magnitudes of MF555W ≃ −5 and MF275W ≃ −6. Furthermore, no counterpart was detected in the NIRCam and IR bands within the astrometric error radius. The SED of X-18 is well represented by a power-law model with an index of ≃ − 2, which suggests that the accretion disk may dominate the observed optical emission. The absence of a NIR counterpart and the interpretation of the SED may be attributable to various scenarios, such as the possibility that in ULXs strong winds or super-Eddington accretion could disperse the cold material around the system, or that the optical emission likely originates from the hot accretion disk, making the IR counterpart undetectable due to disk dominance. If we consider the latter possibility, X-18 could be an HMXB and could potentially include a blue supergiant donor star.

4.4. ULX X-24

The candidate donor star of X-24 is faint in the UVIS1 bands with an MF555W ≃ −5, but bright in the IR, with an MF335M ≃ −12 in the NIRCam. The optical emission remained nearly constant (with variations of ≤0.1 mag in all cases), while a noticeable variation was detected in the IR emission (see Fig. 5). This stability in the optical could suggest that it originates from a region possibly unaffected by the processes driving the variability in the IR and X-ray bands. The IR variability may be linked to mechanisms such as synchrotron emission from jets or thermal emission from surrounding dust, potentially influenced by periodic changes in the X-ray emission associated with the orbital dynamics (Dudik et al. 2016; Lau et al. 2019).

The optical SED of X-24 is modeled as a blackbody with a temperature of 7000 K, suggesting that the donor star may be the source of the optical emission. Moreover, as seen in Fig. 12, X-24 is located in a dense region, likely composed of dust and gas, in the NIRCam image, which could be the source of the NIR emission observed from the position of donor star candidate. The HST Hα (F657N) image also reveals the presence of an H II region at the same location (see Fig. 12). Additionally, the excess observed in the F814W image (see Fig. 6) may suggest that the system is surrounded by gas and dust, or potentially a circumbinary disk (Allak 2023, and references therein). Due to the variable NIR emission, only the SED of three simultaneous filters (F277W, F300M, and F335M) fits a blackbody model with a temperature of 300 K (see Fig. 13). This temperature is not high enough to come from the donor star. Therefore, it is likely that the observed NIR emission is due to gas and dust in the vicinity of the system, while the variation in the NIR emission could be attributed to the presence of a variable jet (Lau et al. 2019 and references therein).

thumbnail Fig. 12.

JWST/NIRCam image of the region around ULX X-24, in the broadband filters F335M (red), F300M (green), and F277W (blue). The position of X-4 is also shown on the HST UVIS F657N image in the upper-right corner. White and green bars indicate the position of the donor star candidate of X-24.

thumbnail Fig. 13.

NIR SED of the donor star candidate for X-24. The dashed red line indicates a blackbody temperature of 300 K. The errors, taken as a systematic 0.05 mag, match the symbol size.

According to the results obtained from the CMDs, the possible age of X-24 was found as 12 Myr, and according to this age, the mass of the donor candidate was derived as 12 M⊙. Considering the optical absolute magnitudes as well as the mass of the donor star candidate, the system is most likely a supergiant in HMXB (Straizys & Kuriliene 1981; Schmidt-Kaler 1982). Although the donor candidate could appear as an RSG based on its colors, the temperature obtained from the optical SED could be considered quite high for an RSG (Davies et al. 2013) but relatively low for an HMXB supergiant (e.g., OB type). Furthermore, this relatively low temperature (7000 K) could be considered an additional indication of gas and dust in the surroundings, which may obscure the optical emission and result in a lower observed temperature.

Ma et al. (2023) interpreted a 7500 s periodic modulation from X-ray data to be an orbital period for X-24. They suggest that this observation is consistent with a scenario involving Roche-lobe overflow in a binary star system. In this context, material from the donor star is transferred to the compact object, likely a BH, leading to strong X-ray emissions. They have claimed that the donor star of X-24 could be a low-mass X-ray binary that is losing mass through Roche-lobe overflow. Furthermore, the periodic behavior and high X-ray luminosity indicate that X-24 may be a stellar-mass BH within a compact binary system. However, in contrast to them, the multiwavelength properties indicated that the possible donor of X-24 could be HMXB. Given that the ULX X-24 is an HMXB supergiant, regardless of the mass of a compact object, the 7500 s orbital period presents a physically unlikely scenario for Roche lobe overflow. For instance, For a compact object mass of 3 M⊙, the Roche lobe radius-to-separation ratio (RL/a), calculated using the Eggleton equation (Eggleton 1983), is approximately 0.5, whereas for 80 M⊙, it drops to about 0.2. With the fixed parameters of donor mass of 12 M⊙ and a 7500 s orbital period, the separation remains too small for the donor to fill its Roche lobe. Thus, the periodic variation is more likely a quasi-periodic oscillation, spin period, or random variability, rather than an orbital period.

5. Summary and conclusions

We conducted an X-ray and multiwavelength analysis of eight ULXs in NGC 1559. We summarize our key findings as follows:

  1. Using astrometric calculations, we identified unique optical counterparts for X-1, X-14, X-18, and X-24. From IR observations with JWST, we found only two ULXs (X-14 and X-24) that matched their optical counterparts. For the remaining ULXs, X-3 and X-17, multiple counterparts were found for both optical and IR images. Also, a source was matched with the position of X-5 from the Gaia and 2MASS catalogs, while no counterpart(s) were found for X-6 in either the HST or JWST images.

  2. The models of the X-ray spectrum of X-1 did not allow us to determine whether its compact object is a BH or a NS. The long-term X-ray light curve of X-1 shows a count rate variation of almost one order of magnitude. Additionally, the optical counterpart of X-1 was detected very faintly in the UVIS1 bands. The constructed SED suggests that the donor star candidate is not the dominant source of the observed optical emission.

  3. No variation was observed in the X-14 emission from either the optical or NIR counterparts. The main source of the observed emission is likely the donor star candidate. CMDs indicate that this donor candidate could be a component of the HMXB system

  4. Although the stable optical emission of X-18 indicates a significant contribution from the donor star, its SED suggests that nonthermal processes from the accretion disk could also contribute to the optical emission. No NIR counterpart was detected for X-18, possibly due to surrounding dust or gas being dispersed by strong winds or super-Eddington disk accretion processes.

  5. For the donor candidate of X-24, the constant optical emission, along with the source SED that is well fitted by a blackbody at 7000 K, suggests that the observed emission may originate from the donor star. Additionally, the variable IR emission detected for this ULX may indicate the influence of a variable jet effect.

  6. In Ma et al. (2023), X-24 was reported to have an orbital period of 7500 s based on X-ray analysis. However, our analysis of CMDs suggests that X-24 has a supergiant donor star. Therefore, we believe the reported period is too short to be the orbital period of an HMXB system.

1

NASA/IPAC Extragalactic Database.

Acknowledgments

This paper was supported by the Scientific and Technological Research Council of Türkiye (TÜBİTAK) through project number 122C183. AA acknowledges support provided by the TÜBİTAK through project number 124F004. We would like to thank the reviewer for the thoughtful comments and efforts towards improving our manuscript.

References

  1. Abraham, Z. 2018, Nat. Astron., 2, 443 [NASA ADS] [CrossRef] [Google Scholar]
  2. Agrawal, V. K., & Nandi, A. 2015, MNRAS, 446, 3926 [NASA ADS] [CrossRef] [Google Scholar]
  3. Aird, J., Nandra, K., Laird, E. S., et al. 2010, MNRAS, 401, 2531 [Google Scholar]
  4. Allak, S. 2022, MNRAS, 517, 3495 [NASA ADS] [CrossRef] [Google Scholar]
  5. Allak, S. 2023, MNRAS, 526, 5765 [NASA ADS] [CrossRef] [Google Scholar]
  6. Allak, S. 2024, MNRAS, 527, 2599 [Google Scholar]
  7. Allak, S., Akyuz, A., Akkaya Oralhan, İ., et al. 2022a, MNRAS, 510, 4355 [NASA ADS] [CrossRef] [Google Scholar]
  8. Allak, S., Akyuz, A., Sonbas, E., & Dhuga, K. S. 2022b, MNRAS, 515, 3632 [NASA ADS] [CrossRef] [Google Scholar]
  9. Ambrosi, E., & Zampieri, L. 2018, MNRAS, 480, 4918 [Google Scholar]
  10. Bachetti, M., Harrison, F. A., Walton, D. J., et al. 2014, Nature, 514, 202 [NASA ADS] [CrossRef] [Google Scholar]
  11. Bressan, A., Marigo, P., Girardi, L., et al. 2012, MNRAS, 427, 127 [NASA ADS] [CrossRef] [Google Scholar]
  12. Brightman, M., Harrison, F. A., Fürst, F., et al. 2018, Nat. Astron., 2, 312 [CrossRef] [Google Scholar]
  13. Brightman, M., Harrison, F. A., Bachetti, M., et al. 2019, ApJ, 873, 115 [Google Scholar]
  14. Brightman, M., Earnshaw, H., Fürst, F., et al. 2020, ApJ, 895, 127 [Google Scholar]
  15. Buisson, D. J. K., Lohfink, A. M., Alston, W. N., & Fabian, A. C. 2017, MNRAS, 464, 3194 [Google Scholar]
  16. Caballero-García, M. D., Belloni, T., & Zampieri, L. 2013, MNRAS, 436, 3262 [CrossRef] [Google Scholar]
  17. Calamida, A., Bajaj, V., Mack, J., et al. 2022, AJ, 164, 32 [NASA ADS] [CrossRef] [Google Scholar]
  18. Carpano, S., Haberl, F., Maitra, C., & Vasilopoulos, G. 2018, MNRAS, 476, L45 [NASA ADS] [CrossRef] [Google Scholar]
  19. Chandar, R., Johns, P., Mok, A., et al. 2020, ApJ, 890, 150 [NASA ADS] [CrossRef] [Google Scholar]
  20. Cui, Y., Hada, K., Kawashima, T., et al. 2023, Nature, 621, 711 [NASA ADS] [CrossRef] [Google Scholar]
  21. Davies, B., Kudritzki, R.-P., Plez, B., et al. 2013, ApJ, 767, 3 [Google Scholar]
  22. de Vaucouleurs, G., de Vaucouleurs, A., Corwin, Herold G., Jr., et al. 1991, Third Reference Catalogue of Bright Galaxies (Springer Science& Business Media) [Google Scholar]
  23. Dudik, R. P., Berghea, C. T., Roberts, T. P., et al. 2016, ApJ, 831, 88 [NASA ADS] [CrossRef] [Google Scholar]
  24. Eggleton, P. P. 1983, ApJ, 268, 368 [Google Scholar]
  25. Evans, P. A., Beardmore, A. P., Page, K. L., et al. 2007, A&A, 469, 379 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  26. Evans, P. A., Beardmore, A. P., Page, K. L., et al. 2009, MNRAS, 397, 1177 [Google Scholar]
  27. Fabrika, S., Ueda, Y., Vinokurov, A., Sholukhova, O., & Shidatsu, M. 2015, Nat. Phys., 11, 551 [NASA ADS] [CrossRef] [Google Scholar]
  28. Fabrika, S. N., Atapin, K. E., Vinokurov, A. S., & Sholukhova, O. N. 2021, Astrophys. Bull., 76, 6 [NASA ADS] [CrossRef] [Google Scholar]
  29. Heida, M., Jonker, P. G., Torres, M. A. P., et al. 2014, MNRAS, 442, 1054 [NASA ADS] [CrossRef] [Google Scholar]
  30. Kaaret, P., Feng, H., & Roberts, T. P. 2017, ARA&A, 55, 303 [Google Scholar]
  31. King, A. R. 2009, MNRAS, 393, L41 [NASA ADS] [Google Scholar]
  32. King, A. R., Davies, M. B., Ward, M. J., Fabbiano, G., & Elvis, M. 2001, ApJ, 552, L109 [NASA ADS] [CrossRef] [Google Scholar]
  33. King, A., Lasota, J.-P., & Middleton, M. 2023, New A Rev., 96, 101672 [NASA ADS] [CrossRef] [Google Scholar]
  34. Kong, L.-D., Zhang, S., Zhang, S.-N., et al. 2022, ApJ, 933, L3 [NASA ADS] [CrossRef] [Google Scholar]
  35. Kotze, M. M., & Charles, P. A. 2012, MNRAS, 420, 1575 [NASA ADS] [CrossRef] [Google Scholar]
  36. Lau, R. M., Heida, M., Walton, D. J., et al. 2019, ApJ, 878, 71 [NASA ADS] [CrossRef] [Google Scholar]
  37. Link, B. 2003, in Radio Pulsars, eds. M. Bailes, D. J. Nice, & S. E. Thorsett, ASP Conf. Ser., 302, 241 [NASA ADS] [Google Scholar]
  38. Liu, J., Bregman, J. N., & McClintock, J. E. 2009, ApJ, 690, L39 [NASA ADS] [CrossRef] [Google Scholar]
  39. Lomb, N. R. 1976, Ap&SS, 39, 447 [Google Scholar]
  40. Ma, C.-H., Li, K.-L., Chu, Y.-H., & Kong, A. K. H. 2023, ApJ, 956, 41 [NASA ADS] [CrossRef] [Google Scholar]
  41. Maccacaro, T., Feigelson, E. D., Fener, M., et al. 1982, ApJ, 253, 504 [NASA ADS] [CrossRef] [Google Scholar]
  42. Middleton, M. J., Walton, D. J., Fabian, A., et al. 2015, MNRAS, 454, 3134 [Google Scholar]
  43. Pasham, D. R., & Strohmayer, T. E. 2013, ApJ, 774, L16 [NASA ADS] [CrossRef] [Google Scholar]
  44. Pinto, C., & Walton, D. J. 2023, High-Resolution X-ray Spectroscopy: Instrumentation, Data Analysis, and Science (Singapore: Springer Nature Singapore) [Google Scholar]
  45. Roberts, T. P. 2007, Ap&SS, 311, 203 [NASA ADS] [CrossRef] [Google Scholar]
  46. Rodríguez Castillo, G. A., Israel, G. L., Belfiore, A., et al. 2020, ApJ, 895, 60 [Google Scholar]
  47. Scargle, J. D. 1982, ApJ, 263, 835 [Google Scholar]
  48. Schmidt-Kaler, T. 1982, Bulletin d’Information du Centre de Donnees Stellaires, 23, 2 [Google Scholar]
  49. Soria, R., Kuntz, K. D., Winkler, P. F., et al. 2012, ApJ, 750, 152 [NASA ADS] [CrossRef] [Google Scholar]
  50. Stetson, P. B. 1987, PASP, 99, 191 [Google Scholar]
  51. Stocke, J. T., Morris, S. L., Gioia, I. M., et al. 1991, ApJS, 76, 813 [Google Scholar]
  52. Straizys, V., & Kuriliene, G. 1981, Ap&SS, 80, 353 [CrossRef] [Google Scholar]
  53. Sutton, A. D., Roberts, T. P., Walton, D. J., Gladstone, J. C., & Scott, A. E. 2012, MNRAS, 423, 1154 [NASA ADS] [CrossRef] [Google Scholar]
  54. Sutton, A. D., Done, C., & Roberts, T. P. 2014, MNRAS, 444, 2415 [NASA ADS] [CrossRef] [Google Scholar]
  55. Tao, L., Feng, H., Grisé, F., & Kaaret, P. 2011, ApJ, 737, 81 [NASA ADS] [CrossRef] [Google Scholar]
  56. Townsend, L. J., & Charles, P. A. 2020, MNRAS, 495, 139 [Google Scholar]
  57. Tully, R. B., Courtois, H. M., Dolphin, A. E., et al. 2013, AJ, 146, 86 [NASA ADS] [CrossRef] [Google Scholar]
  58. Vinokurov, A., Fabrika, S., & Atapin, K. 2018, ApJ, 854, 176 [NASA ADS] [CrossRef] [Google Scholar]
  59. Walton, D. J., Bachetti, M., Fürst, F., et al. 2018, ApJ, 857, L3 [Google Scholar]
  60. Wasserman, I. 2003, MNRAS, 341, 1020 [NASA ADS] [CrossRef] [Google Scholar]
  61. Yao, Y., & Feng, H. 2019, ApJ, 884, L3 [NASA ADS] [CrossRef] [Google Scholar]

Appendix A: Observations and positions of the ULXs

Table A.1.

Log of HST optical and JWST IR observations.

Table A.2.

Chandra and corrected X-ray coordinates of the ULXs.

All Tables

Table 1.

Vega magnitudes of the optical and IR counterparts of four ULXs.

In the text
Table 2.

Power-law and diskbb model parameters of X-1 from the Swift/XRT time-averaged spectrum.

In the text
Table A.1.

Log of HST optical and JWST IR observations.

In the text
Table A.2.

Chandra and corrected X-ray coordinates of the ULXs.

In the text

All Figures

thumbnail Fig. 1.

False-color Chandra image of the galaxy NGC 1559. The energy ranges used are highlighted. The image has been smoothed using a 5 arcsec Gaussian, and the ULXs are indicated with white circles.
In the text
thumbnail Fig. 2.

Red-green-blue images of NGC 1559 from HST (left) and JWST (right). The filters used for the HST image are F814W, F555W, and F438W, and for JWST they are F355M, F300W, and F275W, respectively. The ULXs are marked with red circles on both images. Note that the X-1 source is outside the area shown in the JWST image.
In the text
thumbnail Fig. 3.

Positions of the Chandra (red circle) and Gaia (green box) reference sources used in astrometric calculations shown in the HST image. The solid yellow circle (radius of 1 arcmin) indicates the area within which reference sources were searched. Here the y- and x-axes indicate declination (Dec) and right ascension (RA), respectively.
In the text
thumbnail Fig. 4.

HST F555W image of corrected X-ray positions (solid red circles) and Chandra source detection error ellipses (dashed blue ellipses) of the ULXs. Since X-1 is not observed in the HST F555W filter, its position is shown on the F438W image. The numbers shown for the color bars are in units of electrons s−1. Solid green circles within the error radii of X-5 indicate Gaia sources. All panels are to the same scale, with north up and east to the left.
In the text
thumbnail Fig. 5.

Long-term light curves for X-14 (above) and X-24 (below) using 13 observations taken with the HST F160W IR images. The dashed red lines represent the average magnitudes derived from these observations.
In the text
thumbnail Fig. 6.

SEDs of four counterparts of X-1, X-18, and X-24 (from top to bottom). The power-law and blackbody models for ULXs X-1, X-18, and X-24 are shown with dashed red, blue, and green lines, respectively. All data are shown with filled black circles. The data errors, taken as a systematic 0.05 mag, match the symbol size.
In the text
thumbnail Fig. 7.

HST/WFC3 CMDs for counterparts of X-14 (filled star) and X-24 (filled triangle). The isochrones were corrected for an extinction of AV = 0.04 mag.
In the text
thumbnail Fig. 8.

Stack Swift/XRT image of the galaxy NGC 1559. The image is smoothed with a 2 arcsec Gaussian. The ULXs are indicated with red circles.
In the text
thumbnail Fig. 9.

Light-curves of the X-1. Panel a: Long-term Swift/XRT light curve of X-1. The epoch-1 (panel b), epoch-2, and epoch-3 (panel c) observations are indicated by red-, blue-, and black-filled circles, respectively. Panel d: Lomb–Scargle periodograms of X-1. The periodic signal that peaks at 130.5 days (dashed red line) was found for epoch-3.
In the text
thumbnail Fig. 10.

Swift/XRT time-averaged energy spectrum of X-1 with diskbb (above) and power-law (below) models.
In the text
thumbnail Fig. 11.

X-ray soft (0.3–1 keV) and hard (1–4 keV) count rates vs. time diagrams (upper two panels) and the hardness-intensity diagram (lower panel) of X-1.
In the text
thumbnail Fig. 12.

JWST/NIRCam image of the region around ULX X-24, in the broadband filters F335M (red), F300M (green), and F277W (blue). The position of X-4 is also shown on the HST UVIS F657N image in the upper-right corner. White and green bars indicate the position of the donor star candidate of X-24.
In the text
thumbnail Fig. 13.

NIR SED of the donor star candidate for X-24. The dashed red line indicates a blackbody temperature of 300 K. The errors, taken as a systematic 0.05 mag, match the symbol size.
In the text

Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.

Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.

Initial download of the metrics may take a while.