I am interested in many topics in observational and theoretical astrophysics. Below is a brief description of some of the topics of my past and current research, from exotic stellar binaries to planet formation to quasars to exo-planets to black hole jets to cosmology.

Projects are available for current and prospective graduate and undergraduates students to work in two areas:

(1) With LIGO's recent discovery of neutron star and black hole mergers and upcoming time-variability surveys like LSST, the exploration of the variable universe is the emerging frontier of Galactic astrophysics. Our group is particularly interested in stellar evolution of binary systems and resulting production of stellar remnants -- white dwarfs, neutron stars and black holes. We are taking advantage of cutting-edge emerging astrometric (Gaia), spectroscopic (LAMOST; SDSS), and time-domain data (WISE, TESS, etc) to look for exotic products of close binary evolution, to study the dynamics of stellar binaries in the Galaxy and to probe star and planet formation (see example papers below).

(2) In extragalactic astrophysics, the co-evolution of supermassive black holes and their host galaxies is now considered to be a major component of galaxy formation. With my group, I have led an extensive observational program on discovery and characterization of galactic winds powered by supermassive black holes. This long-sought process is now considered the key missing piece in our understanding of galaxy formation, and the removal of gas from galaxies due to black-hole-driven winds likely limited the maximal mass of galaxies in the universe. We will be continuing to explore the co-evolution of galaxies and black holes in multiple approved JWST programs and with upcoming spectroscopic surveys like PFS.

Do not hesitate to contact me -- email is best; make sure to include your CV.

Binaries of compact stellar remnants. Ultracompact binaries containing stellar remnants -- white dwarfs, neutron stars and black holes -- are at the origin of some of the most exotic and important transient events in the universe. One of the most enduring mysteries in astrophysics is that of progenitors of type Ia supernovae, stellar explosions that serve as cosmological standard candles and were used in discovering the dark energy. While it is generally accepted that type Ia supernovae result from an explosion of a white dwarf, the nature of the companion that provides the mass accreted on the white dwarf before the explosion remains unknown. Even the phase preceding the formation of ultracompact binaries -- so-called "common envelope" evolution -- is shrouded in mystery and has rarely been observed.

Example papers by our group: In our group, we are interested in all stages of formation and evolution of binaries containing stellar remnants. We are actively involved in SDSS-V which is currently underway and will obtain the largest set of spectra of white dwarfs. Vedant Chandra, an undergraduate student in our group, was the lead author of the first scientific paper from SDSS-V, that on our discovery of a 99-min white dwarf-white dwarf binary.

[Top figure: modeling of the binary orbit of the 99-min double-lined spectroscopic binary of two white dwarfs; Bottom figure: the masses of companions, periods and expected gravitational wave strain for known ultracompact binaries. Credit: V.Chandra]

Formation and evolution of stellar binaries. Our group is making use of two observational techniques enabled by the high-quality astrometric Gaia data. One is kinematic age measurements. When stars are born in the Galaxy, they have near-circular Galactic orbits with low velocity dispersions relative to each other. As they age, they experience gravitational perturbations from other stars and inhomogeneities within the Galaxy, such as molecular clouds. As a result, the velocity dispersion grows as a function of age. Another novel probe of stellar evolution enabled by Gaia is the co-moving companions method. Because of the availability of distances and proper motions, loosely bound companions to single stars or short-period binaries can be identified in large numbers for the first time.

Example papers by our group: In Lifetime of short-period binaries, we obtain two new and unexpected results: (1) short-period binaries take 1-4 Gyr to form and (2) they vanish from the population at 10 Gyr, before the stars reach the end of their main-sequence lives. These are the first measurements of the timescales associated with orbital evolution. In this paper, we do not find an enhancement of wide binary companions to hot-jupiter hosts and all but rule out the Kozai oscillations as the origin of hot jupiters. In this paper, we find a strong and non-monotonic dependence of the wide binary fraction on metallicity, providing novel probes of star formation scenarios and even of stellar migration in the Galactic disk.

[Shown in the figure is the short-period binary fraction as a function of stellar kinematics, which is proxy for stellar age. Credit: H.-C.Hwang]

Precision modeling of white dwarfs and their equation of state. Millions of stellar spectra -- including hunders of thousands white dwarfs -- will be taken by ongoing and upcoming Galactic surveys. Analysis tools must be adjusted to deal with this volume of data and adapted to quickly search for rare and unusual objects. Our white dwarf spectroscopic analysis pipeline WDTools starts with sparse grids of public theoretical spectra and uses neural networks to interpolate over them. It can then generate theoretical spectra for any sets of temperatures and surface gravities in negligible amount of time and then fit a white dwarf spectrum with accuracy and precision which are as good as those of the underlying theoretical models.

We then applied WDTools to 3,000 white dwarfs with high-quality Gaia data and SDSS spectra. Curiously, this population appears to be redshifted on average, but not because of any bulk expansion of the white dwarf population relative to the Sun -- this is the manifestation of the gravitational redshift as photons climb out of the white dwarf gravitational potential on their way from the stellar surface to the observer. We then measured the mass-dependent gravitational redshift by averaging out random motions within each white dwarf mass bin. This is the first time the effect has been measured across a wide range of masses and with a large sample of stars and found to be in excellent agreement with the theoretical models of white dwarf structure. This work was subject of press releases and is described here and here.

Example papers by our group: Here we present the WDTools and here we measure white dwarf gravitational redshifts.

[Shown in the figure is the fitting of Balmer absorption lines performed by WDTools and the gravitational redshift as a function of surface gravity compared to theoretical models. Credit: V.Chandra]

Time-domain astrophysics -- the study of how astronomical objects change and evolve as a function of time -- is an emerging frontier of astrophysical research. Opening the discovery space of astronomical variability will be one of the top priorities for the field in the next decade, with several all-sky surveys currently in operation (e.g., WISE, Gaia, TESS, ZTF) or coming online shortly (e.g., the Large Synoptic Survey Telescope). In our group, we are exploring variability of binary stars, stellar remnants, young stellar objects and other Galactic populations using a variety of datasets and theoretical modeling.

Example papers by our group: On Mira variables as distance calibrators, On the origin of Orion fingers. Descriptions of ongoing undergraduate projects in this area can be found here.

[Shown in the figure is the lightcurve (brightness vs time) of a giant outburst of a young stellar object seen in the optical (blue) and in the infrared (red). Credit: N.L.Zakamska]

Dual super-massive black holes. The search for sub-kpc dual supermassive black holes and for gravitationally bound black holes is of immense interest in many areas of astrophysics, from galaxy formation to studies of active galactic nuclei (AGN) to searches for low-frequency gravitational waves. It is expected that following the merger of two galaxies, their two central supermassive black holes evolve into a bound binary via dynamical friction and interactions with gas and stars. This binary may eventually merge through gravitational radiation to become a single black hole. While some kpc-scale dual AGN have been found, currently there is no systematic way to uncover dual AGN at sub-kpc scales. We are developing a new method to use Gaia to discover sub-kpc dual AGN over the entire sky. Specifically, if two AGN are variable, even though they are not individually resolved, Gaia is able to detect the shift of their joint light centroid, which manifests as high astrometric noise.

Example papers by our group and by our collaborators: The fundamentals of varstrometry technique; Constraints on actively accreting black holes offset from their hosts' nuclei. Multiple follow-up programs are ongoing.

[Shown in the figure is the main principle behind the varstrometry technique which identifies double quasars by their variability-induced astrometric noise. Credit: H.-C. Hwang]

Quasar feedback is one of the major puzzles in galaxy formation theory. The word "feedback" implies that the quasar must have a strong effect on its large-scale environment: its entire host galaxy or even the inter-galactic matter. Quasar feedback can provide a natural explanation for the upper limit on the mass of galaxies in the local universe; for the black-hole / bulge correlations; and for the similarity between the black hole accretion history and the star formation history of the Universe. The only problem is that the direct observations of this phenomen have been extremely hard to come by!

Our group obtained ground-breaking observations of powerful quasar winds, and we are now using Magellan, Gemini, Herschel, EVLA, Hubble, Chandra and ALMA data to further characterize their phenomenology and physical conditions. We are studying this phenomenon both at low redshifts, where high-quality observations can yield a detailed physical description of this process, and at high redshifts, when quasar winds made the greatest impact on the formation of massive galaxies. My group, my collaborators and I published dozens of papers on this topic. Furthermore, in collaboration with Prof. Veilleux (UMd College Park) and my former postdoc Dr. Wylezalek we have an approved JWST program to further study quasar winds.

More about feedback can be found on this page for the workshop on feedback which took place at JHU in Fall 2013.

[Picture: example kinematic maps of quasar winds (intensity, radial velocity, FWHM); credit: Liu, Zakamska et al. MNRAS, 2013a, 2013b, 2014 based on Gemini GMOS data]

The puzzle of radio emission from radio quiet quasars. The most extended, powerful and beautiful sources in the radio sky are due to relativistic jets launched by supermassive black holes in centres of galaxies, but only a minority of active black holes produce these structures. The majority of black holes in the universe are "radio-quiet" and do not have extended bright jets, but they are not necessarily radio silent. Many of them appear as weak point sources in deep radio surveys, such as FIRST, and likely dominate number counts of radio sources above 0.1 mJy.

Among actively accreting black holes -- quasars -- only about 10% show extended jets and are "radio-loud". The nature of the radio emission in the remaining 90% of objects is not known. In 2014, we discovered a relationship between the radio emission of quasars and their extended gas kinematics and hypothesized that radio emission can be a bi-product of winds launched by the quasars and now propagating through the galactic interstellar medium, driving shocks and accelerating relativistic particles, which would produce radio emission. We have several papers discussing this scenario. In a recent paper, Hwang et al. 2018 confirm that quasars with the most extreme outflows produce radio emission in the "radio-intermediate" range, which is consistent with the wind hypothesis.

Our previous largest sample of obscured quasars included nearly 1000 objects and demonstrated that they were at least as common as the `normal' unobscured quasars at low (z < 0.7) redsfhits. This work was subject of a press release, and Reinabelle Reyes, a Princeton student who worked with me on this project, received the Chambliss Award Honorable Mention for our AAS poster. The most recent sample by Yuan et al. 2016 is the largest optical catalog of type 2 quasars at z < 1 (nearly 3000 objects), with all gas kinematic measurements publicly available.

[Picture: three obsured quasars; credit HST/SDSS/Nadia Zakamska]

My paper "H_2 emission arises outside photodissociation regions in ultraluminous infrared galaxies" came out in Nature and demonstrated that ULIRGs have more warm molecular gas emission than would be expected based on their star formation rate. With undergraduate student Matthew Hill, we demonstrated that the excess emission likely originates in neutral interstellar gas shock-heated by outflows, which are driven by the supernovae explosions or an active nucleus. We present the analysis of the relationship between the molecular hydrogen and the AGN activity in the largest sample of Spitzer galaxies in Lambrides et al. 2018.

[Picture: ULIRG NGC 6240 optical+IR which shows strong anomalous molecular hydrogen emission; credit NASA/JPL/Caltech/STScI/ESA]

[Picture: outflows in ULIRGs are driven by supernovae associated with recent star-formation activity or an active nucleus, distinguishable by the outflow velocity. From Hill and Zakamska, MNRAS 2014.]

Accretion of matter onto compact objects (neutron stars or black holes) is frequently accompanied by collimated relativistic outflows. My collaborators and I found a new class of self-similar analytical solutions for the structure of pressurized relativistic jets with toroidal magnetic fields. As the jet material expands, it runs into the ambient medium, resulting in a pile-up of material along the jet boundary. The magnetic field acts as a collimating force and produces a magnetic pinch along the axis of the jet. We constructed models which take into account these forces, and then used these models to predict the intensity and polarization of the observed synchrotron emission based on the physical properties of the jet, such as its Lorenz-factor, energy flux and magnetization. This work demonstrated that projection effects and the emissivity pattern of the jet have a strong effect on the observed polarization signal and provided an explanation for puzzling polarization patterns seen in some BL Lac objects.

[Picture: one of our models; credit: Zakamska, Begelman, Blandford, 2008]

More recently, with collaborators M.Pan and E.Ford, I investigated observational biases in the measurements of exoplanet orbital parameters -- especially eccentricity -- obtained from radial velocity observations. We created a mock catalog of radial velocity data, choosing input planet masses, orbital periods, and observing patterns from actual radial velocity surveys and varying input eccentricities. We then analyzed the simulated data sets using Markov chain Monte-Carlo simulations and compared calculated orbital parameters with the input values. We found a significant bias in the determination of small eccentricities in radial velocity surveys. Since eccentricity is positive definite, eccentricities of planets on nearly circular (e < 0.05) orbits are preferentially overestimated. Though the extrasolar planet catalogs report eccentricities below 0.05 for just 19% of single-planet systems, we estimate that the true fraction of e < 0.05 orbits is about 35%. These methods can be applied to multi-planet systems and to other types of planetary surveys.

[Picture: some results of our simulations; credit: Zakamska, Pan, Ford, 2011]

Whether or not the solar system contains as yet undiscovered massive planets or possesses a distant stellar companion has been the subject of intensive research. One of the ways to constrain the mass and position of a putative companion is to constrain the acceleration of the known solar system barycenter. In my work with S.Tremaine, I used pulsar timing data to determine limits on such acceleration. The constraint can be obtained by comparing the observed orbital period decay of binary pulsars to the value expected from general relativity. An independent constraint can be obtained in a statistical sense from millisecond pulsars. While we did not find any massive companions, the sensitivity in accelerations achieved using pulsar timing is comparable to the acceleration in the Galactic potential at the position of the Sun. As of this writing (fall 2020), the topic is fashionable again, thanks to the discovery of gravitational waves, high-precision astrometry from Gaia and ongoing work on improving pulsar timing arrays.

[Picture: one of our sky maps showing acceleration limits; credit: Zakamska, Tremaine, 2005]

The mechanism of powering the X-ray emission of hot gas in clusters of galaxies is largely unknown. One possibility is the transport of thermal energy from the outer regions of the cluster into the center via thermal conduction. With R.Narayan, I constructed models of intracluster gas incorporating thermal conduction and showed that these models provided a good description of observed temperature and electron density profiles. We further showed that thermal conduction was a viable energy source for about half of the clusters in our study, whereas another half required additional energy sources, probably AGNs, in their centers. We also showed that conduction may prevent the gas from becoming thermally unstable.

[Picture: our models for one of the clusters; credit: Zakamska, Narayan, 2003]