Optical Gravitational Lensing Experiment Research Articles

Semi-supervised classification and clustering analysis for variable stars

ABSTRACT The immense amount of time series data produced by astronomical surveys has called for the use of machine learning algorithms to discover and classify several million celestial sources. In the case of variable stars, supervised learning approaches have become commonplace. However, this needs a considerable collection of expert-labelled light curves to achieve adequate performance, which is costly to construct. To solve this problem, we introduce two approaches. First, a semi-supervised hierarchical method, which requires substantially less trained data than supervised methods. Second, a clustering analysis procedure that finds groups that may correspond to classes or subclasses of variable stars. Both methods are primarily supported by dimensionality reduction of the data for visualization and to avoid the curse of dimensionality. We tested our methods with catalogues collected from the Optical Gravitational Lensing Experiment (OGLE), the Catalina Sky Survey (CSS), and the Gaia survey. The semi-supervised method reaches a performance of around 90 per cent for all of our three selected catalogues of variable stars using only $5{{\ \rm per\ cent}}$ of the data in the training. This method is suitable for classifying the main classes of variable stars when there is only a small amount of training data. Our clustering analysis confirms that most of the clusters found have a purity over 90 per cent with respect to classes and 80 per cent with respect to subclasses, suggesting that this type of analysis can be used in large-scale variability surveys as an initial step to identify which classes or subclasses of variable stars are present in the data and/or to build training sets, among many other possible applications.

The OGLE Collection of Variable Stars: Nearly 66,000 Mira Stars in the Milky Way

We present a collection of 65,981 Mira-type variable stars found in the Optical Gravitational Lensing Experiment (OGLE) project database. Two-thirds of our sample (40,356 objects) are located in the Galactic bulge fields, whereas 25,625 stars are in the Galactic disk. The vast majority of the collection (47,532 objects) comprises new discoveries. We provide basic observational parameters of the Mira variables: equatorial coordinates, pulsation periods, I-band and V-band mean magnitudes, I-band brightness amplitudes, and identifications in other catalogs of variable stars. We also provide the I-band and V-band time-series photometry collected since 1997 during the OGLE-II, OGLE-III, and OGLE-IV phases. The classical selection process, i.e., being mostly based on the visual inspection of light curves by experienced astronomers, has led to the high purity of the catalog. As a result, this collection can be used as a training set for machine-learning classification algorithms. Using overlapping areas of adjacent OGLE fields, we estimate the completeness of the catalog to be about 96%. We compare and discuss the statistical features of Miras located in different regions of the Milky Way. We show examples of stars that change their type over time, from a semiregular variable to Mira and vice versa. This data set is perfectly suited to studying the three-dimensional structure of the Milky Way, and it may help to explain the puzzle of the X-shaped bulge.

The VVV survey: Long-period variable stars

Context. Long-period variable stars (LPVs) are pulsating red giants, primarily in the asymptotic giant branch phase, and they include both Miras and semi-regular variables (SRVs). Their period-age and period-luminosity relations enable us to trace different stellar populations, as they are intrinsically very bright and cover a wide range in distances and ages. Aims. The purpose of this study is to establish a census of LPV stars in a region close to the Galactic center, using the six-year database of the Vista Variables in the Vía Láctea (VVV) ESO Public Survey, as well as to describe the methodology that was employed to search for and characterize LPVs using VVV data. Near-IR surveys such as VVV provide a unique opportunity to probe the high-extinction innermost regions of the Milky Way. The detection and analysis of the intrinsically bright Miras in this region could provide us with an excellent probe of the properties of the Milky Way far behind its bulge. Methods. We used point-spread function photometry for all available Ks-band images in ten VVV tiles, covering 16.4 deg2 in total, overlapping fields observed in the course of the Optical Gravitational Lensing Experiment (OGLE)-III survey. We designed a method to select LPV candidates, and we used the known variables from OGLE-III and other known variables from the literature to test our approach. The reduced χ2 statistic, along with the flux-independent index K(fi), were used in our analysis. The Lomb-Scargle period search method, Fourier analysis, template fitting, and visual inspection were then performed to refine our sample and characterize the properties of the stars included in our catalog. Results. A final sample of 130 Mira candidates, of which 129 are new discoveries, was thus obtained, with periods in the range between about 80 and 1400 days. Moreover, a sample of 1013 LPV candidates is also presented, whose periods are however not sufficiently constrained by the available data. A fraction of the latter may eventually turn out to be SRVs. Ages are measured for these stars based on a reassessment of the period-age relations available in the literature. The Miras in our catalog include 18 stars satisfying the requirements to serve as reliable distance indicators and which are not saturated in the VVV Ks-band images. Their distances are accordingly derived and discussed. A number of objects that are seemingly placed far behind the Milky Way’s bulge was detected.

The OGLE Collection of Variable Stars: One Thousand Heartbeat Stars in the Galactic Bulge and Magellanic Clouds

Abstract We present a collection of 991 heartbeat star (HBS) candidates found in the Optical Gravitational Lensing Experiment (OGLE) project data archive. We discuss the selection process of the HBS candidates and the structure of the catalog itself. It consists of 512 stars located toward the Galactic bulge, 439 stars located in the Large Magellanic Cloud, and 40 in the Small Magellanic Cloud. The collection contains two large groups of HBSs with different physical properties. The main distinction between the two groups is the evolutionary status of the primary star. The first group of about 100 systems contains a hot main-sequence or a Hertzsprung-gap primary star, while the second group of about 900 systems includes a red giant. For each star, we provide two-decade-long time-series photometry, in the Cousins I- and Johnson V-band filters, obtained by the OGLE project. We also present basic observational information as well as orbital parameters derived from the light-curve modeling.

NGC 2004 #115: a black hole imposter containing three luminous stars

ABSTRACT NGC 2004 #115 is a recently identified black hole (BH) candidate in the Large Magellanic Cloud (LMC) containing a B star orbiting an unseen companion in a 2.9 d orbit and a Be star tertiary. We show that the unseen companion is not a 25 M⊙ BH, but a $(2\!-\!3)\, {\rm M}_{\odot }$ luminous star. Analysing the Optical Gravitational Lensing Experiment (OGLE) and MAssive Compact Halo Object (MACHO) light curves of the system, we detect ellipsoidal variability with amplitude 10 times larger than would be expected if the companion were a 25 M⊙ BH, ruling out the low inclination required for a massive companion. The light curve also shows a clear reflection effect that is well modelled with a $2.5\, {\rm M}_{\odot }$ main-sequence secondary, ruling out a lower mass BH or neutron star companion. We consider and reject models in which the system is a binary containing a stripped star orbiting the Be star: only a triple model with an outer Be star can explain both the observed light curve and radial velocities. Our results imply that the B star, whose slow projected rotation velocity and presumed tidal synchronization were interpreted as evidence for a low inclination (and thus a high companion mass), is far from being tidally synchronized: despite being in a 2.9 d orbit that is fully or nearly circularized (e < 0.04), its surface rotation period appears to be at least 20 d. We offer cautionary notes on the interpretation of dormant BH candidates in binaries.

Accretion Disks and Long Cycles in β Lyrae-Type Binaries

In order to inquire about the nature of the accretion disks formed around the more massive companion in binaries with β Lyrae-type light curves, we review literature presenting some physical and observational properties of these systems. In addition, we inspect the photometric time series of three representative eclipsing systems obtained by the Optical Gravitational Lensing Experiment (OGLE) project during the last decades and compare them with β Lyrae. All these three systems show indications of being semidetached with a more massive B-type component and in a mass transfer stage. They also show long photometric cycles, and two of them show changes in the orbital light curve that can be interpreted in terms of structural changes of the accretion disks, eventually driven by variations in the mass transfer rate.

NANOGrav hints on planet-mass primordial black holes

Recently, the North American Nanohertz Observatory for Gravitational Waves (NANOGrav) claimed the detection of a stochastic common-spectrum process of the pulsar timing array (PTA) time residuals from their 12.5 year data, which might be the first detection of the stochastic background of gravitational waves (GWs). We show that the amplitude and the power index of such waves imply that they could be the secondary GWs induced by the peaked curvature perturbation with a dust-like post inflationary era with −0.091 ≲ w ≲ 0.048. Such stochastic background of GWs naturally predicts substantial existence of planet-mass primordial black holes (PBHs), which can be the lensing objects for the ultrashort-timescale microlensing events observed by the Optical Gravitational Lensing Experiment (OGLE).

Multiwavelength Properties of Miras

Abstract We comprehensively study the variability of Miras in the Large Magellanic Cloud by simultaneously analyzing light curves in 14 bands in the range of 0.5–24 μm. We model over 20 yr long, high-cadence I-band light curves collected by the Optical Gravitational Lensing Experiment (OGLE) and fit them to light curves collected in the remaining optical/near-infrared/mid-infrared bands to derive both the variability amplitude ratio and phase lag as a function of wavelength. We show that the variability amplitude ratio declines with increasing wavelength for both oxygen-rich (O-rich) and carbon-rich (C-rich) Miras, while the variability phase lag increases slightly with increasing wavelength. In a significant number of Miras, mostly the C-rich ones, the spectral energy distributions (SEDs) require the presence of a cool component (dust) in order to match the mid-IR data. Based on SED fits for a golden sample of 140 Miras, we calculated synthetic period–luminosity relations (PLRs) in 42 bands for the existing and future sky surveys that include OGLE, the VISTA Near-Infrared YJK s Survey of the Magellanic Clouds System, Legacy Survey of Space and Time, Gaia, Spitzer, the Wide-field Infrared Survey Explorer, the James Webb Space Telescope, the Nancy Grace Roman Space Telescope (formerly WFIRST), and the Hubble Space Telescope. We show that the synthetic PLR slope decreases with increasing wavelength for both the O-rich and C-rich Miras in the range of 0.1–40 μm. Finally, we show the location and motions of Miras on the color–magnitude and color–color diagrams.

Model for the long and orbital brightness variability of the β Lyrae type binary OGLE-BLG-ECL-157529

Context. Some close binaries of the β Lyrae type show photometric cycles longer than the orbital one, which are possibly related to changes in their accretion disks. Aims. We aim to understand the short- and long-scale morphologic changes observed in the light curve of the eclipsing system OGLE-BLG-ECL-157529. In particular, we want to shed light on the contribution of the disk variability to these changes, especially those related to the long cycle, occurring on timescales of hundreds of days. Methods. We studied I-band Optical Gravitational Lensing Experiment (OGLE) photometric times series spanning 18.5 years, constructing disk models by analyzing the orbital light curve at 52 different consecutive epochs. An optimized simplex algorithm was used to solve the inverse problem by adjusting the light curve with the best stellar-orbital-disk parameters for the system. We applied an analysis of principal components to the parameters to evaluate their dependence and variability. We constructed a description of the mass transfer rate in terms of disk parameters. Results. We find that the overall light variability can be understood in terms of a variable mass transfer rate and variable accretion disk. The system brightness at orbital phase 0.25 follows the long cycle and is correlated with the mass transfer rate and the disk thickness. The long-cycle brightness variations can be understood in terms of differential occultation of the hotter star by a disk of variable thickness. Our model fits the overall light curve during 18.5 years well, including epochs of reversal of main and secondary eclipse depths. The disk radius cyclically change around the tidal radius, decoupled from changes in the mass transfer rate or system brightness, suggesting that viscous delay might explain the non-immediate response. Although the disk is large and fills a large fraction of the hot star Roche lobe, Lindblad resonance regions are far beyond the disk, excluding viscous dissipation as a major source of photometric variability.

Classification of OGLE Eclipsing Binary Stars Based on Their Morphology Type with Locally Linear Embedding

Abstract The Optical Gravitational Lensing Experiment (OGLE) continuously monitors hundreds of thousands of eclipsing binaries in the Galactic bulge field and the Magellanic Clouds. These objects have been classified into major morphological subclasses, such as contact, noncontact, ellipsoidal, and cataclysmic variables, both by matching the light curves with predefined templates and by visual inspections. Here we present the result of a machine-learned automatic classification based on the morphology of light curves inspired by the classification of eclipsing binaries observed by the original Kepler mission. We similarly use a dimensionality reduction technique with locally linear embedding to map the high dimension of the data set into a low-dimensional embedding parameter space, while keeping the local geometry and the similarities of the neighboring data points. After three consecutive steps, we assign one parameter to each binary star, which scales well with the “detachness,” i.e., the sum of the relative radii of the components. This value is in good agreement with the morphology types listed in the OGLE catalog and, along with the orbital periods, can be used to filter any morphological subtypes based on the similarity of light curves. Our open-source pipeline can be applied in a fully automatic way to any other large data set to classify binary stars.

The VMC Survey – XLII. Near-infrared period–luminosity relations for RR Lyrae stars and the structure of the Large Magellanic Cloud

ABSTRACT We present results from an analysis of ∼29 000 RR Lyrae stars located in the Large Magellanic Cloud (LMC). For these objects, near-infrared time-series photometry from the VISTA survey of the Magellanic Clouds system (VMC) and optical data from the Optical Gravitational Lensing Experiment (OGLE) IV survey and the Gaia Data Release 2 catalogue of confirmed RR Lyrae stars were exploited. Using VMC and OGLE IV magnitudes we derived period–luminosity (PL), period–luminosity–metallicity (PLZ), period–Wesenheit (PW), and period–Wesenheit–metallicity (PWZ) relations in all available bands. More that 7000 RR Lyrae were discarded from the analysis because they appear to be overluminous with respect to the PL relations. The $PL_{K_{\mathrm{s}}}$ relation was used to derive individual distance to ${\sim}22\, 000$ RR Lyrae stars, and study the three-dimensional structure of the LMC. The distribution of the LMC RR Lyrae stars is ellipsoidal with the three axis S1 = 6.5 kpc, S2 = 4.6 kpc, and S3 = 3.7 kpc, inclination i = 22 ± 4° relative to the plane of the sky and position angle of the line of nodes θ = 167 ± 7° (measured from north to east). The north-eastern part of the ellipsoid is closer to us and no particular associated substructures are detected and neither any metallicity gradient.

The Sun’s distance from the Galactic Centre and mid-plane, and the Galactic old bulge’s morphology: 715 VVV Type II Cepheids

ABSTRACT A statistical method is employed in tandem with new VISTA Variables in the Via Lactea (VVV) near-infrared observations to determine the Sun’s distance from the Galactic Centre (r0, GC), the Sun’s height from the local mid-plane (z0), and to likewise infer the shape of the Galactic ∼10 Gyr old bulge. Specifically, the conclusions stem from an investigation of 715 high-latitude (|b| > 1°) and centrally symmetric concentrated Type II Cepheids (T2Cs) recently identified in the VVV survey by Braga et al. The analysis yields r0 = 8.35 ± 0.10 kpc and z0 = 10 ± 2 pc. The T2Cs distribution within the effective bulge radius rbulge = 2–3 kpc is an ellipsoid exhibiting axial ratios of ≈1:0.7:0.6, with the major axis inclined at an angle θ ≈ −3° to the Sun–GC sightline. T2Cs do not trace a prominent barred structure at distances >1 kpc from the GC. A key conclusion is that analyses of independent optical and infrared Optical Gravitational Lensing Experiment (OGLE) and VVV observations yield consistent results (e.g. r0 > 8.0 kpc and both observations display a comparable shape of an ellipsoid), thus providing a constrained and reduced systematic uncertainty.

Measuring stellar atmosphere parameters using follow-up polarimetric microlensing observations

ABSTRACT We present an analysis of potential follow-up polarimetric microlensing observations to study the stellar atmospheres of distant stars. First, we produce synthetic microlensing events using the Galactic model, stellar population and interstellar dust toward the Galactic bulge. We simulate the polarization microlensing light curves and pass them through the instrument specifications of the FOcal Reducer and low dispersion Spectrograph (FORS2) polarimeter at the Very Large Telescope (VLT), and then analyse them. We find that the accuracy of the VLT allows us to constrain the atmospheres of cool red giant branch (RGB) stars. Assuming that about 3000 microlensing events are detected per year by the Optical Gravitational Lensing Experiment (OGLE) telescope, we expect to detect almost 20, 10, 8 and 5 polarization microlensing events for the following four criteria: being three consecutive polarimetric data points above the baseline with 1σ, 2σ, 3σ and 4σ, respectively, in the polarimetric light curves. We generalize the covariance matrix formulation and present the combination of polarimetric and icinformation that leads us to measure the scattering optical depth of the atmosphere and the inner radius of the stellar envelope of RGB stars. These two parameters could determine the dust opacity of the atmosphere of cool RGB source stars and the radius where dust can be formed.

OGLE-GAL-ACEP-091: The First Known Multi-mode Anomalous Cepheid

Abstract Anomalous Cepheids (ACs) are metal-deficient, core-helium-burning pulsating stars with masses in the range 1.2–2.2 . Until recently, all known ACs were pure single-mode pulsators. The first candidate for an AC pulsating in more than one radial mode—OGLE-GAL-ACEP-091—was recently identified in the Milky Way based on the photometric database of the Optical Gravitational Lensing Experiment (OGLE) survey. We analyze this object showing that it is actually a triple-mode pulsator. Its position in the Petersen diagram, the light-curve morphology quantified by Fourier coefficients, and absolute magnitudes derived from the Gaia parallax are consistent with the assumption that OGLE-GAL-ACEP-091 is an AC. Our grid of linear pulsation models indicates that OGLE-GAL-ACEP-091 is a 1.8 star with a metallicity of about [Fe/H] = −0.5 dex.

Verification of Einstein’s Formula for Gravitational Deflection of Light Using Observations of Galactic Microlensing

General relativity (GR) has a solid experimental base. However, the emergence of new experimental capabilities and independent observational information stimulates continuing tests of general relativity. The purpose of this work is to evaluate the potential of gravitational microlensing of distant sources on the stars of our Galaxy and to verify Einstein’s formula of gravitational refraction. This effect has been repeatedly tested in the Solar System in high-accuracy experiments with the propagation of radio waves, when the measurements are most effective for the distances from the signal trajectory to the Sun on the order of several solar radii. In the case of galactic microlensing, a quite different type of observational data and other characteristic distances are used that are determined in the high magnification events by the Einstein ring radii, which is typically of the order of 1 AU. Although the gravitational deflections of light by stars are very small and currently practically inaccessible by direct measurements, nonetheless, due to the large distances to the microlenses, the radiation flux from the source in strong microlensing events can increase several times. To verify Einstein’s formula, a more general dependence of the beam deflection angle $$\alpha \propto 1/{{p}^{{1 + \varepsilon }}}$$ on its impact distance p relative to the deflector is considered and, accordingly, the equations of gravitational lensing are modified. The challenge is to limit e based on observational data. The Early Warning System data obtained in 2018 within the Optical Gravitational Lensing Experiment (OGLE) ( http://ogle.astrouw.edu.pl/ogle4/ ews/2019/ews.html ) was used. A sample of 100 light curves from the data obtained by the OGLE group in 2018 was formed. Each light curve was fitted as part of a modified model of gravitational lensing with parameter e. As a result, 100 values of e and estimates of their variances were obtained. It was found that the mean value of e does not contradict GR within the limits of a one percent standard deviation. In the future, using a larger number of light curves will allow one to hope for a significant decrease in the error of e due to statistical averaging.

Verification of Einstein’s formula for gravitational deflection of light using observations of galactic microlensing

The potential of the gravitational microlensing inside our Galaxy for testing the Einstein formula for the gravitational light deflection is discussed. For this purpose, the lens mapping is modified by introducing parameter eps, which characterizes the deviation from this formula. An example of such deviation described by a simple power law is analyzed. We formed a sample of 100 microlensing light curves using the data of the Optical Gravitational Lensing Experiment (OGLE) for 2018. The resulting eps value does not contradict the General Relativity within 1 percent errors.

Microlensing Optical Depth and Event Rate in the OGLE-IV Galactic Plane Fields

Abstract Searches for gravitational microlensing events are traditionally concentrated on the central regions of the Galactic bulge but many microlensing events are expected to occur in the Galactic plane, far from the Galactic Center. Owing to the difficulty in conducting high-cadence observations of the Galactic plane over its vast area, which are necessary for the detection of microlensing events, their global properties were hitherto unknown. Here, we present results of the first comprehensive search for microlensing events in the Galactic plane. We searched an area of almost 3000 square degrees along the Galactic plane ( , 0° < l < 50°, 190° < l < 360°) observed by the Optical Gravitational Lensing Experiment (OGLE) during 2013–2019 and detected 630 events. We demonstrate that the mean Einstein timescales of Galactic plane microlensing events are on average three times longer than those of Galactic bulge events, with little dependence on the Galactic longitude. We also measure the microlensing optical depth and event rate as a function of Galactic longitude and demonstrate that they exponentially decrease with the angular distance from the Galactic Center (with the characteristic angular scale length of 32°). The average optical depth decreases from 0.5 × 10−6 at l = 10° to 1.5 × 10−8 in the Galactic anticenter. We also find that the optical depth in the longitude range 240° < l < 330° is asymmetric about the Galactic equator, which we interpret as a signature of the Galactic warp.

Optical and X-ray study of the peculiar high-mass X-ray binary XMMU J010331.7−730144

ABSTRACT For a long time XMMU J010331.7−730144 was proposed as a high-mass X-ray binary candidate based on its X-ray properties, however, its optical behaviour was unclear – in particular previous observations did not reveal key Balmer emission lines. In this paper, we report on optical and X-ray variability of the system. XMMU J010331.7–730144 has been monitored with the Optical Gravitational Lensing Experiment (OGLE) in the I and V bands for the past 9 yr where it has shown extremely large amplitude outbursts separated by long periods of low-level flux. During its most recent optical outburst we obtained spectra with the Southern African Large Telescope (SALT) where, for the first time, the H α line is seen in emission, confirming the Be nature of the optical companion. The OGLE colour–magnitude diagrams also exhibit a distinct loop that is explained by changes in mass-loss from the Be star and mass outflow in its disc. In the X-rays, XMMU J010331.7−730144 has been monitored by the Neil Gehrels Swift Observatory through the S-CUBED programme. The X-ray flux throughout the monitoring campaign shows relatively low values for a typical Be/X-ray binary system. We show, from the analysis of the optical data, that the variability is due to the Be disc density and opacity changing rather than its physical extent as a result of efficient truncation by the NS. The relatively low X-ray flux can then be explained by the neutron star normally accreting matter at a low rate due to the small radial extent of the Be disc.

Swift J004427.3−734801 – a probable Be/white dwarf system in the Small Magellanic Cloud

ABSTRACT Swift J004427.3−734801 is an X-ray source in the Small Magellanic Cloud (SMC) that was first discovered as part of the Swift S-CUBED programme in 2020 January. It was not detected in any of the previous 3 yr worth of observations. The accurate positional determination from the X-ray data has permitted an optical counterpart to be identified that has the characteristics of an O9V−B2III star. Evidence for the presence of an infrared excess and significant I-band variability strongly suggests that this is an OBe-type star. Over 17 yr worth of optical monitoring by the OGLE (Optical Gravitational Lensing Experiment) project reveals periods of time in which quasi-periodic optical flares occur at intervals of ∼21.5 d. The X-ray data obtained from the S-CUBED project reveal a very soft spectrum, too soft to be that from accretion on to a neutron star or black hole. It is suggested here that this is a rarely identified Be star–white dwarf binary in the SMC.

Confirmation of monoperiodicity above 20 s for two blue large-amplitude pulsators

ABSTRACT Blue large-amplitude pulsators (BLAPs) are a new class of pulsating variable stars. They are located close to the hot subdwarf branch in the Hertzsprung–Russell diagram and have spectral classes of late O or early B. Stellar evolution models indicate that these stars are likely radially pulsating, driven by iron group opacity in their interiors. A number of variable stars with a similar driving mechanism exist near the hot subdwarf branch with multiperiodic oscillations caused by either pressure (p) or gravity (g) modes. No multiperiodic signals were detected in the OGLE (Optical Gravitational Lensing Experiment) discovery light curves since it would be difficult to detect short-period signals associated with higher order p modes with the OGLE cadence. Using the RISE instrument on the Liverpool Telescope, we produced high-cadence light curves of two BLAPs, OGLE-BLAP-009 (mv = 15.65 mag) and OGLE-BLAP-014 (mv = 16.79 mag), using a 720 nm longpass filter. Frequency analysis of these light curves identifies a primary oscillation with a period of 31.935 ± 0.0098 min and an amplitude from a Fourier series fit of 0.236 mag for BLAP-009. The analysis of BLAP-014 identifies a period of 33.625 ± 0.0214 min and an amplitude of 0.225 mag. Analysis of the residual light curves reveals no additional short-period variability down to an amplitude of 15.20 ± 0.26 mmag for BLAP-009 and 58.60 ± 3.44 mmag for BLAP-014 for minimum periods of 20 and 60 s, respectively. These results further confirm that the BLAPs are monoperiodic.