What is Strong Lensing?
Strong gravitational lenses are rare systems where a massive foreground object (e.g. a galaxy or a cluster) creates multiple images of one or more higher redshift sources (e.g. galaxies or quasars). See on the right an example of Einstein ring, which is created when lens and source are approximately aligned. Strong lenses are useful for a wide range of cosmological and astrophysical studies. As first, they can act as a cosmic telescope, providing magnified images of otherwise unresolved high-z sources, and can provide cosmological constraints on the dark energy equation of state and precision measurements of the Hubble constant. The information obtained from strong lensing also allows us to study the mass distribution in the inner regions of lens galaxies, e.g. the fraction of dark matter in their central regions, the slope of their inner mass density profile, and their dark matter substructures, putting constraints on the Initial Mass Function (IMF), when combined with dynamical and stellar population synthesis analyses.
My work is focussing on the whole assembly line of gravitational lensing: to search new candidates, arcs and lensed quasars, in wide-field surveys (e.g., KiDS@VST) and perform their spectroscopic follow-up, preparing ourself for the future, for the enormous amount of data coming from Euclid and LSST
Finding strong lensing arcs with CNNs
We have applied a morphological classification based on Convolutional Neural Networks (CNNs) for recognizing strong gravitational lenses, in wide field surveys. We have trained with a dataset composed of real galaxies from the Kilo Degree Survey (KiDS) and simulated lensed sources. One CNN is trained with single r-band galaxy images, hence basing the classification mostly on the morphology. While the other CNN is trained on g-r-i composite images, relying mostly on colours and morphology. We present a new sample of strong gravitational lens candidates, selected from 904 square degree of Data Release 4 of KiDS, i.e. the `Lenses in the Kilo-Degree Survey' (LinKS) sample. A straightforward application of our procedure to future Euclid data can select a sample of about 3000 lens candidates with less than 10 per cent expected false positives and requiring minimal human intervention.
Information and images about the sample of candidates are provided at the LinKS webpage.
Our list of papers:
We were also part of the Strong lens finding challenge, whose results were discussed in the paper: The strong gravitational lens finding challenge (Metcalf et al. 2019, A&A, 625, 119).
Finding strong lensed quasars
We have also started a systematic census of strong lensed quasars using KiDS data, within the KiDS-SQuaD (KiDS Strongly lensed Quasar Detection) project. Lensed quasars are Strong lensing events with a quasar as a lensed source. Our search has first started using the following methods: 1) multiplet detection in KiDS and/or Gaia, 2) direct modeling of KiDS cutouts and iii) positional offsets between different surveys (KiDS-vs-Gaia, Gaia-vs-2MASS). We have also used a decision trees based classifier to separate galaxies from quasars. We have built and made available to the community the KiDS Bright EXtraGalactic Objects catalogue (KiDS-BEXGO), created to find gravitational lenses. This is made of about ∼6 millions of sources classified as quasars (∼200,000) and galaxies (∼5.7 millions), up to r < 22.
Our list of papers:
We have also reported the discovery of a new quadruple lensed quasar in the Research note KiDS0239-3211: A New Gravitational Quadruple Lens Candidate: Sergeyev et al. 2018, RNAAS, 2, 189 (independently also discovered with CNNs and included in the LinKS sample)
The next step of the strong lensing assembly line consists to measure the spectroscopic redshifts of lenses and sources, validating the discovered lens candidates. Thus, we have planned a multi-facility campaign to perform the spectroscopic follow-ups of lens candidates. In particular, we are collecting data from the program: Gotta catch'Em All: the spectroscopic follow-up of strong gravitational lenses from KiDS and KABS surveys using the SALT telescope. The program is ongoing, and renewed for a third semester.
Our list of papers:
Ultra-compact massive galaxies
Massive Early-Type Galaxies (ETGs) are believed to form through a two-phase formation scenario (Oser et al. 2010). An intense and fast dissipative series of processes forms their central mass bulk at z > 2 generating, after the star formation quenches, a compact massive quiescent galaxy with size a factor of ∼4 smaller than local massive galaxies. These are the so-called red nuggets. Then, a second phase, dominated by mergers and gas inflows, is responsible for their dramatic structural evolution and size growth (e.g. van Dokkum et al. 2010). Nevertheless, a small fraction of red nuggets survives intact until the present day Universe, without experiencing any merger or interaction, massive (M★ > 8 × 1010 M☉) and compact (Re < 1.5 kpc): the so-called Relic Galaxies (Trujillo et al. 2009). These old Ultra Compact Massive Galaxies (UCMGs) are supposedly made of "in situ only" pristine stellar populations. As such they provide a unique opportunity to track the formation of this specific galaxy stellar component, which is mixed with the accreted one in normal massive ETGs. Relic galaxies are the perfect systems to study in great detail the processes that shaped the mass assembly of massive galaxies in the high-z universe. In the local universe one true relic has been found and studied in great details (NGC 1277, Trujillo et al. 2009, 2014; Ferré-Mateu et al. 2017). Even if UCMGs remain extremely rare at very low redshifts, their number is increasing at larger redshifts (0.2 < z < 0.7), thus it is crucial to increase their number statistics in this redshift range.
With KiDS we are concentrating on the search of the UCMGs (Tortora et al. 2016, 2018). This analysis requires large samples of UCMGs, discovered thanks to the large areas observed, and spectroscopic follow-ups to determine the spectroscopic redshifts, provide their validation, and study the systematic sources in the searching strategy. KiDS offers the opportunity to observe large areas of the sky (1350 sq. deg. at the end of the survey) with a good seeing (FWHM ∼ 0.65, on average, for r-band). Among the others, the KiDS image quality makes the data very suitable for measuring structural parameters of galaxies (we have studied the size evolution of ETGs and LTGs in Roy et al. 2018), including compact ones. We have performed the first census of UCMGs in KiDS in Tortora et al. (2016). Then, we have upgraded analysis and results in a second paper of the series (Tortora et al. 2018), selecting UCMGs within 333 sq. deg. of the third data release of KiDS, discussing the first results from our spectroscopic campaign and plotting the abundances in terms of redshift. We find ∼3 UCMG candidates per square degrees, which corresponds to ∼1 per cent of the whole galaxy population at M★ > 8 × 1010 M☉. The number density of UCMGs is reduced of about ten times among redshift z = 0.5 and z = 0, which is a large variation if compared with normal-sized galaxies, suggesting a size-dependent evolution. In the attached figure (extracted from Tortora et al. 2018) I present the number density as a function of redshift, and the results are compared with other works in the literature. Read Tortora et al. (2018) for more details.
We have started a multi-site and multi-facility spectroscopic campaign in the North and South hemisphere, to cover the whole KiDS area during the entire solar year. The multi-site approach allows us to cover the two KiDS patches (KiDS-North from La Palma and KiDS-South from Chile), while the multi-facility allows us to optimize the exposure time according to the target brightness (ranging from r ∼ 18.5 to 20.5). We have planned to observe our UCMG candidates at 3-4m and 8-10m class telescopes (for brighter and fainter targets, respectively). In Tortora et al. (2018) we have presented the first 28 UCMG candidates observed with NTT and TNG, and we present the data for further 33 galaxies, observed with TNG and INT, in Scognamiglio et al. (in preparation).
We will also obtain higher-resolution spectra for a subsample of UCMGs using X-Shooter, which will allow us to measure the stellar population parameters and the Initial Mass Function. We aim at proving that the age of these systems is comparable with the age of the Universe, confirming their relic nature of relic galaxies. We will also derive estimates of the metal content and velocity dispersion, and constraints on the Initial Mass Function, to test the two-phase scenario, and to understand the origin of such rare and peculiar objects, tracing the evolutionary history of the elliptical galaxies.
Our list of papers:
Dark matter and Initial Mass Function
In the concordance cosmological model, structures form hierarchically (in a bottom-up fashion), starting from the amplification via gravitational instability of primordial small density fluctuations in the dark matter (DM). Within these potential wells, the gas is falling down, condensates and starts producing stars. These structures further evolve hierarchically, galaxy merging induces the size evolution that is observed in the most massive galaxies. It is therefore natural to aim at quantifying the amount of DM in galaxies. This task can be addressed when stellar population information are joined with dynamics and/or gravitational lensing.
I have provided a strong contribution to this field of research during the past years, deriving constraints on the central DM content, the total mass density slope and the Initial Mass Function (IMF).
We have analyzed DM fraction within the effective radius as a function of mass, effective radius, average galaxy density and galaxy age, studying the impact of IMF and total mass profile in both local (Tortora et al. 2009, 2012, 2013, 2014a, 2016; Napolitano, Romanowsky & Tortora 2010) and higher-z (Cardone et al. 2009, Cardone & Tortora 2010, Cardone et al. 2011, Tortora et al. 2010, 2014b, 2018) early-type galaxies (ETGs).
We have faced the problem of the IMF (Tortora et al. 2013, 2014a), constraining the stellar mass normalization using the Jeans equations. We show that the IMF is not universal in ETGs, and changes in terms of velocity dispersion and mass. And we also demonstrate that IMF is not universal if the gravity framework is changed (MOND, Tortora et al. 2014; Emergent gravity, Tortora et al. 2018). In Tortora et al. (2016) we have provided the first complete analysis of DM content and IMF for dwarf ellipticals.
In Tortora et al. (2014a) we have found that the slope of the total mass distribution in the central regions of ETGs is not universal and is varying with stellar mass, which may be related to a varying role of dissipation and galaxy mergers with galaxy mass (as pointed out by simulations, Remus et al. 2013). We already found some indication of such "non-homology" in the total mass density profile in Tortora et al. (2009), by comparing observations with N-body simulation predictions. Combining the results for massive ETGs and dwarf ellipticals, we show that DM fraction follows a U-shape trend with stellar mass, with a minimum at M★ ≈ 3 × 1010 M☉. The same mass scale is emerging from the analysis of colour gradients (Tortora et al. 2010) and total mass density slopes (Tortora et al. 2019). In Tortora et al. (2019) we also show that the total mass density slope of late-type galaxies is steepening as a function of mass. The difference of the slope in dwarf ellipticals and late-type galaxies can be understood by stellar feedback from a more prolonged star formation period in the latter systems, causing a transformation of the initial steep density cusp to a more shallow profile. These results are summarized in the figure on the right (from Tortora et al. 2019). By studying the evolution with redshift of the central DM content of very massive ETGs, in Tortora et al. (2014b, 2018), we show that ETGs are larger and more DM dominated at lower redshift (if the IMF is left fixed) and find that minor mergers can explain these results.
List of papers:
Color, stellar population and M/L gradients
The different processes which drive the galaxy evolution might rule the star formation at the global galaxy scale, or act at subgalactic scales (e.g. the nuclear regions versus outskirts) such that they are expected to introduce a gradient in the main stellar properties with the radius, that shall leave observational signatures in galaxy colours.
In Tortora et al. (2010) we have investigated colour and stellar population gradients for a sample of local SDSS galaxies. The colour (and metallicity) gradients of late-type galaxies (LTGs) decrease systematically with mass while the trend for early-type galaxies (ETGs) inverts near a mass of M★ ∼ 3 × 1010 M☉: systematically decreasing and increasing at lower and higher masses, respectively. The overall observational picture is interpreted in the context of differential feedback efficiency of supernovae at low masses and galaxy mergers, AGN feedback at large masses (see a cartoon in the picture on the right, where colour-mass plane is plotted, and main physical processes in place are shown). In Tortora et al. (2011), from the fitted synthetic spectra we have also derived the stellar mass-to-light (M/L) gradients and we have discussed the trends as a function of colour gradients and mass. We find that M/L gradients are tightly correlated with colour gradients, and generally follow patterns of variation with stellar mass and galaxy type found for colour and metallicty gradients. M/L gradients in LTGs would have a larger effect on dark matter inferences, while ETGs have, generally, very shallow gradients.
In Tortora & Napolitano (2012) we have investigated the role of the galaxy mergers in massive galaxies. For central galaxies in groups, we find that both optical colour and M/L gradients are shallower in central galaxies residing in denser environments. On the other hand, satellites do not show any differences in terms of the environment. In central galaxies, we show that the observed trends can be explained with the occurrence of dry mergings (more numerous in denser environments), which produce shallower colour gradients because of more uniform metallicity distributions due to the mixing of stellar populations.
The steepening of metallicity gradients with mass in low mass galaxies is also confirmed by the analysis of a sample of group and cluster satellite galaxies from a N-body-hydrodynamical simulation in Tortora et al (2011). Environment processes are important, we find that dwarf galaxies in clusters have steeper negative gradients with respect to the dwarfs in groups. Finally, Tortora et al (2013) have analyzed the evolution of colour gradients predicted by the hydrodynamical models of ETGs in Pipino et al. (2008), reproducing fairly well the chemical abundance pattern, the metallicity and colour gradients of ETGs at z < 1.
Alternatives to dark matter
Dark matter is one of the biggest puzzles in the modern astrophysics and cosmology, since it is thought to dominate the mass density of galaxies and clusters of galaxies in the Universe, but is elusive, interacting very weakly with visible matter and has not yet detected by any experiment. Thus, alternative ways to solve the missing mass problem have been suggested.
We have analyzed early-type galaxies (ETGs) within the MOdified Newtonian Dynamics (MOND) and the new revolutionary proposition by Verlinde (2016), i.e. Emergent Gravity. In Tortora et al. (2014) and Tortora et al. (2018), we show, for the first time, and with unique and systematic analyses, that within a modified gravity scenario, IMF is not universal, as in the standard gravity. I have also provided a contribution in the study of the weak-field limit of f(R) theories. In this regime, though an additional Yukawa term in the gravitational potential modifies dynamics with respect to the standard Newtonian limit of General Relativity, the motion of massless particles (e.g. photons) results unaffected thanks to suitable cancellations in the post-Newtonian limit (Lubini et al. 2011). We have also successfully fitted the observed velocity dispersion profiles of three elliptical galaxies with a Yukawa-like potential (Napolitano et al. 2012).
I have been recently involved in the PRIN-SKA project ESKAPE-HI, a national funded project which aims at exploring stellar and gas properties in galaxy populations from low to high redshift, in terms of mass and environment, and pave the way for SKA HI surveys. I am working on both predictions for HI detection in SKA and scaling relations at low- and high-redshift.
Metallicity and gas content are intimately related in the baryonic exchange cycle of galaxies. To quantify this relation and obtain information about physical processes shaping it, in Ginolfi et al. (2019) we have collected "MAGMA" (Metallicity And Gas for Mass Assembly), a sample of ∼400 local galaxies, which covers an unprecedented range in parameter space: it spans more that 5 orders of magnitudes in stellar mass, star formation rate and gas mass, and almost a factor 2 in metallicity. We find that the relations between M★, SFR, Metallicity (Z) and Mgas (including HI and H2 gas) require only two dimensions to describe the hypersurface. In particular, to accommodate the curvature in the M★-Z, we have applied a piecewise 3D PCA that successfully predicts observed metallicities to an accuracy of ∼0.07 dex. The break mass used in the piecewise PCA, i.e. Mbreak ∼ 2 × 1010 M☉, is the value which minimizes the scatter. This value is similar to the characteristic mass which emerges from other observational probes (colour gradients, Tortora et al. 2010; dark matter fraction and mass density slope, Tortora et al. 2016, 2019). We also present a new relation to express Mgas as a linear combination of M★ and SFR, to an accuracy of ∼0.2 dex. For the first time on a statistically significant sample, we also quantify Mgas as a function of M★, evaluating the effect of gas on the mass-metallicity relation. Finally, we derive the metallicity-loading and mass-loading factors for the outflows produced by the MAGMA galaxies, finding that the metal-retention efficiency is not constant with M★ (in the figure the metal-loading factor is plotted as a function of circular velocity, Ginolfi et al. 2019). We show that metals are expelled more efficiently from low-mass galaxies than from massive ones. Our analysis shows clearly that gas content and outflows driven by star formation shape the Mass-metallicity relation. Such a dependence on the depth of the potential well is also shaping other scaling relations: a) the steepening of colour profiles with stellar mass (more negative gradients in more massive galaxies) found for LTGs in Tortora et al. (2010) and b) the steepening of total mass density profiles (more negative total mass density slopes in more massive galaxies) in Tortora et al. (2019).