Search

Classical novae are known to demonstrate a supersoft X-ray source (SSS) state following outbursts, which is associated with residual thermonuclear burning on the white dwarf (WD) surface. During its all-sky survey (eRASS1), the eROSITA telescope onboard the Spectrum-Roentgen-Gamma observatory discovered a bright new SSS, whose position is consistent with the known classical nova AT 2018bej in the Large Magellanic Cloud. There were two eROSITA spectra obtained during eRASS1 and eRASS2 monitoring epochs and one XMM-Newton grating spectrum close to the eRASS1 epoch. We aim to describe the eROSITA and XMM-Newton spectra of AT 2018bej with our local thermodynamic equilibrium (LTE) atmosphere models. We focused on the evolution of the hot WD properties between the eRASS1 and eRASS2 epochs, especially on the change of the carbon abundance. A grid of LTE model atmosphere spectra were calculated for different values of the effective temperature (from $T_{\rm eff}= 525$ to $700\,\rm kK$), surface gravity (six values) and chemical composition with five different values of carbon and nitrogen abundances. Both eRASS1 and XMM $0.3-0.6$ keV spectral analyses yield a temperature of the WD of $T_{\rm eff}{\sim}\,600\, \rm kK$ and a WD radius of $8000-8700\,\rm km$. Simultaneous fitting of the eROSITA spectra for two epochs (eRASS1 and eRASS2) with a common WD mass parameter demonstrates a decrease in $T_{\rm eff}$ accompanied by an increase in the WD radius and a decrease in the carbon abundance. However, these changes are marginal and coincide within errors. The derived WD mass is estimated to be $1.05-1.15\, M_\odot$. We traced a minor evolution of the source on a half-year timescale accompanied by a decrease in carbon abundance and concluded that LTE model atmospheres can be used to analyse the available X-ray spectra of classical novae during their SSS stage., Comment: 12 pages, 16 figures, accepted for publication to A&A

For the precise spectral analysis of hot stars, advanced stellar-atmosphere models that consider deviations from the local thermodynamic equilibrium are mandatory. This requires accurate atomic data to calculate all transition rates and occupation numbers for atomic levels in the considered model atoms, not only for a few prominent lines exhibited in an observation. The critical evaluation of atomic data is a challenge because it requires precise laboratory measurements. Ultraviolet spectroscopy of hot stars with high resolving power provide such "laboratory" spectra. We compare observed, isolated lines of the iron group (here calcium to nickel) with our synthetic line profiles to judge the accuracy of the respective oscillator strengths. This will verify them or yield individual correction values to improve the spectral analysis, i.e., the determination of, e.g., effective temperature and abundances. To minimize the error propagation from uncertainties in effective temperature, surface gravity (g), and abundance determination, we start with a precise reanalysis of three hot subdwarf stars, namely EC 11481-2303, Feige 110, and PG 0909+276. Then, we measure the abundances of the iron-group elements individually. Based on identified, isolated lines of these elements, we compare observation and models to measure their deviation in strength (equivalent width). For EC 11481-2303 and Feige 110, we confirmed the previously determined effective temperatures and log g values within their error limits. For all three stars, we fine-tuned all metal abundances to achieve the best reproduction of the observation. For more than 450 isolated absorption lines of the iron group, we compared modeled and observed line strengths. We selected strong, reliable isolated absorption lines, which we recommend to use as reference lines for abundance determinations in related objects., Comment: 13 pages, 12 figures

Supersoft X-ray sources (SSSs) are accreting white dwarfs (WDs) with stable or recurrent thermonuclear burning on their surfaces. High-resolution X-ray spectra of such objects are rather complex, often consist of several components, and are difficult to interpret accurately. The main emission source is the hot surface of the WD and the emergent radiation can potentially be described by hot WD model atmospheres. We present a new set of such model atmosphere spectra computed in the effective temperature range from $100\rm\,kK$ to $1000\rm\,kK$, for eight values of surface gravity and three different chemical compositions. These compositions correspond to the solar one as well as to the Large and Small Magellanic Clouds, with decreased heavy element abundances, at one-half and one-tenth of the solar value. The presented model grid covers a broad range of physical parameters and, thus, it can be applied to a wide range of objects. It is also publicly available in XSPEC~format. As an illustration, we applied it here for the interpretation of \textit{Chandra} and XMM grating spectra of two classical SSSs, namely, CAL 83 (RX J0543.5$-$6823) and RX J0513.9$-$6951. The obtained effective temperatures and surface gravities of $T_{\rm eff} \approx 560$ kK, $\log g \approx 8.6-8.7$, and $T_{\rm eff} \approx 630\,{\rm kK}, \log g \approx 8.5-8.6$, respectively, are in a good agreement with previous estimations for both sources. The derived WD~mass estimations are within $1.1-1.4\,M_\odot$ for CAL 83 and $1.15-1.4\,M_\odot$ for RX J0513.9$-$6951. The mass of the WD in CAL $83$ is consistent with the mass predicted from the respective model of recurrent thermonuclear burning., Comment: 12 pages, 15 figures, accepted for publication to A&A

The purpose of this work is to spectroscopically analyse the classical Cepheid V708 Car. A preliminary check of the spectrum of V708 Car showed that this is a lithium-rich supergiant. We also found that V708 Car has an unusual chemical composition in that the abundances of various elements correlate with their condensation temperatures. We tried to find an explanation of this feature, which is unusual for classical Cepheids. For the spectroscopic analysis, we used methods based on the assumption of local and non-local thermodynamic equilibrium. We determined the fundamental parameters of our program star V708 Car. This long-period Cepheid has a mass of about 12 M$_{\odot}$. We derived the abundances of 27 chemical elements in this star. They are clearly correlated with their condensation temperature: the higher the condensation temperature, the lower the abundance (there are exceptions for sodium and barium, however). We explain this peculiar chemical composition of the V708 Car atmosphere by the gas-dust separation in the envelope of this star. A similar mechanism leads to the observed peculiarities of the chemical composition of $\lambda$ Boo, W Vir, and asymptotic giant branch stars., Comment: Accepted for publication in A&A

The EPOS4 project is an attempt to construct a realistic model for describing relativistic collisions of different systems, from proton-proton ($pp$) to nucleus-nucleus ($AA$), at energies from several TeV per nucleon down to several GeV. We argue that a parallel scattering formalism (as in EPOS4) is relevant for primary scatterings in AA collisions above 4 GeV, whereas sequential scattering (cascade) is appropriate below. We present briefly the basic elements of EPOS4, and then investigate heavy ion collisions from 62.4 GeV down to 4 GeV, to understand how physics changes with energy, studying in particular the disappearance of the fluid component at low energies., Comment: 20 pages, 31 figures. arXiv admin note: text overlap with arXiv:2310.09380

While looking for recently discovered distant Cepheids with an interesting chemical composition, we noticed one star (OGLE GD-CEP-1353) with extremely large equivalent widths of spectral lines of heavy elements. The aim of this work is to perform an abundance analysis, and to find a possible explanation for the found chemical anomaly. Quantitative analysis of the equivalent widths and synthetic spectrum synthesis were used to derive abundances in this star. Both local and nonlocal thermodynamic equilibrium (LTE and NLTE) approximations were used in our analysis. Abundances of 28 chemical elements from carbon to thorium were derived. While light and iron peak elements show abundances typical for distant Cepheids (located in the outer disk), the s-process elements are overabundant about one dex. r-process elements are slightly less overabundant. This makes the star a unique Cepheid of our Galaxy., Comment: Accepted for publication in A&A

Ultrarelativistic heavy-ion collisions will first realize many nucleon-nucleon scatterings, happening instantaneously and therefore necessarily in parallel, due to the short collision time. An appropriate quantum mechanical tool to treat that problem is S-matrix theory, and it has been known for a long time how to derive a simple geometric probabilistic picture, still widely used, and here the Abramovskii-Gribov-Kancheli (AGK) theorem plays a crucial role. All this is done in a scenario where energy conservation is not taken care of, but this is needed, in particular for Monte Carlo simulations. When introducing energy-momentum sharing properly, the AGK theorem does not apply anymore, nor do simple geometric concepts such as binary scaling. I will discuss this (very serious) problem, and how it can be solved, in the EPOS4 framework. When connecting the multiple Pomeron approach (for parallel scatterings) and perturbative QCD, one is actually forced to implement in a very particular way saturation scales, in order to get an approach free of contradictions. One recovers a generalized AGK theorem (gAGK), valid at large $p_{t}$ (larger than the relevant saturation scales). I discuss how gAGK is related to factorization (in proton-proton scatterings) and binary scaling (in heavy-ion collisions). I will show some applications, using this new approach as an initial condition for hydrodynamical evolutions, for heavy-ion collisions at $\sqrt{s_{NN}}$ of 5.02 TeV and 200 GeV, to get some idea about the energy dependence., Comment: 53 pages, 73 figures

Deep Language Models (DLMs) provide a novel computational paradigm for understanding the mechanisms of natural language processing in the human brain. Unlike traditional psycholinguistic models, DLMs use layered sequences of continuous numerical vectors to represent words and context, allowing a plethora of emerging applications such as human-like text generation. In this paper we show evidence that the layered hierarchy of DLMs may be used to model the temporal dynamics of language comprehension in the brain by demonstrating a strong correlation between DLM layer depth and the time at which layers are most predictive of the human brain. Our ability to temporally resolve individual layers benefits from our use of electrocorticography (ECoG) data, which has a much higher temporal resolution than noninvasive methods like fMRI. Using ECoG, we record neural activity from participants listening to a 30-minute narrative while also feeding the same narrative to a high-performing DLM (GPT2-XL). We then extract contextual embeddings from the different layers of the DLM and use linear encoding models to predict neural activity. We first focus on the Inferior Frontal Gyrus (IFG, or Broca's area) and then extend our model to track the increasing temporal receptive window along the linguistic processing hierarchy from auditory to syntactic and semantic areas. Our results reveal a connection between human language processing and DLMs, with the DLM's layer-by-layer accumulation of contextual information mirroring the timing of neural activity in high-order language areas.

We have searched for lines of the radioactive element technetium (Tc) in the spectrum of Przybylski's star (HD101065). The nuclei of this chemical element are formed in the slow process of capturing thermalized neutrons. The possible lines of Tc~I are heavily blended. We have synthesized the profile of one resonance line at 4297.06 \AA~, which is also a part of the complex blend, and we arrived at a decision that it is not visible in the spectrum (as was first noted by Ryabchikova), casting doubt on the existence of technetium in the atmosphere of the Przybylski's star. Therefore, based on our calculated combined profile, which has been adjusted to the observed blend profile at ~4297.2 A (that may possibly contain the resonance technetium line 4297.06 A), we reduce the maximum technetium abundance to $\log\epsilon$(Tc/H) = 2.5. This value can be considered only as an upper limit of the technetium abundance in the Przybylski's star., Comment: 4 pages, 3 figures

Nearly ten years ago, Kang, Ma, Qiu, and Sterman derived an evolution equation for a $Q\bar{Q}$ pair fragmenting into a quarkonium. In this study we explore the consequence of this evolution for the color-evaporation model, focusing on $J/\psi$ transverse-momentum ($p_t$) distributions in proton-proton collisions. We show that, as expected, it softens the spectrum obtained by fixed-order calculations. While next-to-leading-order calculations strongly overestimate data at large $p_t$, ours, including the (approximate) $Q\bar{Q}$ evolution and next-to-leading-order cross sections computed with Madgraph, are in good agreement with experiments. Since our study with the color-evaporation model shows a significant effect of the $Q\bar{Q}$ evolution at large $p_t$, a determination of scale-dependent long-distance-matrix elements of non-relativistic QCD could be necessary. To describe data at small and intermediate $p_t$, we use the $k_t$-factorization approach, and we argue that quarkonia data could help constrain unintegrated parton densities., Comment: 24 pages, 16 figures

The multiplicity dependence of multistrange hadron yields in proton-proton and lead-lead collisions at several TeV allows one to study the transition from very big to very small systems, in particular, concerning collective effects. I investigate this, employing a core-corona approach based on new microcanonical hadronization procedures in the EPOS4 framework, as well as new methods allowing one to transform energy-momentum flow through freeze-out surfaces into invariant-mass elements. I try to disentangle effects due to ``canonical suppression'' and ``core-corona separation'', which will both lead to a reduction of the yields at low multiplicity., Comment: 27 pages, 32 figures

We recently introduced new concepts, implemented in EPOS4, which allow to consistently accommodate factorization and saturation in high energy proton-proton and nucleus-nucleus collisions, in a rigorous parallel scattering framework. EPOS4 has a modular structure and in this paper, we present in detail how the "single scattering module" (the main EPOS4 building block) is related to perturbative QCD, and how these calculations are performed, with particular care being devoted to heavy flavor contributions. We discuss similarities and differences compared to the usual pQCD approach based on factorization., Comment: 36 pages, 39 figures

Type Ia supernovae are cosmic distance indicators, and the main source of iron in the Universe, but their formation paths are still debated. Several dozen supersoft X-ray sources, in which a white dwarf accretes hydrogen-rich matter from a non-degenerate donor star, have been observed and suggested as Type Ia supernovae progenitors. However, observational evidence for hydrogen, which is expected to be stripped off the donor star during the supernova explosion, is lacking. Helium-accreting white dwarfs, which would circumvent this problem, have been predicted for more than 30 years, also including their appearance as supersoft X-ray sources, but have so far escaped detection. Here we report a supersoft X-ray source with an accretion disk whose optical spectrum is completely dominated by helium, suggesting that the donor star is hydrogen-free. We interpret the luminous and supersoft X-rays as due to helium burning near the surface of the accreting white dwarf. The properties of our system provides evidence for extended pathways towards Chandrasekhar mass explosions based on helium accretion, in particular for stable burning in white dwarfs at lower accretion rates than expected so far. This may allow to recover the population of the sub-energetic so-called Type Iax supernovae, up to 30% of all Type Ia supernovae, within this scenario.

The Southern African Large Telescope (SALT) survey of helium-rich hot subdwarfs aims to explore evolutionary pathways amongst groups of highly-evolved stars. The selection criteria mean that several hot white dwarfs and related objects have also been included. This paper reports the discovery and analysis of eight new very hot white dwarf and pre-white dwarf stars with effective temperatures exceeding 100,000 K. They include two PG1159 stars, one DO white dwarf, three O(He) and two O(H) stars. One of the O(H) stars is the central star of a newly-discovered planetary nebula, the other is the hottest `naked' O(H) star. Both of the PG1159 stars are GW Vir variables, one being the hottest GW Vir star measured and a crucial test for pulsation stability models. The DO white dwarf is also the hottest in its class., Comment: 10 pages, 13 figures

One of the models explaining the high luminosity of pulsing ultra-luminous X-ray sources (pULXs) was suggested by Mushtukov et al. (2015). They showed that the accretion columns on the surfaces of highly magnetized neutron stars can be very luminous due to opacity reduction in the high magnetic field. However, a strong magnetic field leads also to amplification of the electron-positron pairs creation. Therefore, increasing of the electron and positron number densities compensates the cross-section reduction, and the electron scattering opacity does not decrease with the magnetic field magnification. As a result, the maximum possible luminosity of the accretion column does not increase with the magnetic field. It ranges between 10$^{40} - 10^{41}$ erg s$^{-1}$ depending only slightly on the magnetic field strength., Comment: 3 pages, 2 figures, subm. to Proc. IAU Symp. 363, poster presentation

Using the EPOS3 model with UrQMD to describe the hadronic phase, we study the production of short-lived hadronic resonances and the modification of their yields and $p_{T}$ spectra in p-Pb collisions at $\sqrt{s_{NN}}$ = 5.02 TeV. High-multiplicity p-Pb collisions exhibit similar behavior to mid-peripheral Pb-Pb collisions at LHC energies, and we find indications of a short-lived hadronic phase in p-Pb collisions that can modify resonance yields and $p_{T}$ spectra through scattering processes. The evolution of resonance production is investigated as a function of the system size, which is related to the lifetime of the hadronic phase, in order to study the onset of collective effects in p-Pb collisions. We also study hadron production separately in the core and corona parts of these collisions, and explore how this division affects the total particle yields as the system size increases., Comment: 12 pages, 10 figures, to be submitted to Physical Review C

Stratospheric balloons offer accessible and affordable platforms for observations in atmosphere-constrained wavelength ranges. At the same time, they can serve as an effective step for technology demonstration towards future space applications of instruments and other hardware. The Stratospheric UV Demonstrator of an Imaging Observatory (STUDIO) is a balloon-borne platform and mission carrying an imaging micro-channel plate (MCP) detector on a 0.5 m aperture telescope. STUDIO is currently planned to fly during the summer turnaround conditions over Esrange, Sweden, in the 2022 season. For details on the ultraviolet (UV) detector, see the contribution of Conti et al. to this symposium. The scientific goal of the mission is to survey for variable hot compact stars and flaring M-dwarf stars within the galactic plane. At the same time, the mission acts as a demonstrator for a versatile and scalable astronomical balloon platform as well as for the aforementioned MCP instrument. The gondola is designed to allow the use of different instruments or telescopes. Furthermore, it is designed to serve for several, also longer, flights, which are envisioned under the European Stratospheric Balloon Observatory (ESBO) initiative. In this paper, we present the design and current status of manufacturing and testing of the STUDIO platform. We furthermore present the current plans for the flight and observations from Esrange., Comment: Proc. SPIE Conf., Astronomical Telescopes and Instrumentation, 14-18 December 2020

We investigate the possible existence of graviballs, a system of bound gravitons, and show that two gravitons can be bound together by their gravitational interaction. This idea connects to black hole formation by a high-energy $2\to N$ scattering and to the gravitational geon studied by Brill and Hartle. Our calculations rely on the formalism and techniques of quantum field theory, specifically on low-energy quantum gravity. By solving numerically the relativistic equations of motion, we have access to the space-time dynamics of the (2-gravitons) graviball formation. We argue that the graviball is a viable dark matter candidate and we compute the associated gravitational lensing., Comment: 25 pages, 6 figures, 2 videos as ancillary files. Includes two new sections

In this work is presented several results for cross sections of exotic particles production in electron-positron collisions in future linear electron-positron colliders. It is shown that the central production of exotic leptons, dilatons, radions, axions and technipions in this type of collision can be measurable for the predicted kinematic parameters for linear colliders., Comment: 14 pages, 7 figures

Accurate atomic data is an essential ingredient for the calculation of reliable non-local thermodynamic equilibrium (NLTE) model atmospheres that are mandatory for the spectral analysis of hot stars. We aim to search for and identify for the first time spectral lines of copper (atomic number Z = 29) and indium (Z = 49) in hot white dwarf (WD) stars and to subsequently determine their photospheric abundances. Oscillator strengths of Cu IV - VII were calculated to include radiative and collisional bound-bound transitions of Cu in our NLTE model-atmosphere calculations. Oscillator strengths of In IV - VI were compiled from the literature. We newly identified 1 Cu IV, 51 Cu V, 2 Cu VI, and 5 In Vlines in the ultraviolet (UV) spectrum of DO-type WD RE 0503-289. We determined the photospheric abundances of 9.3 X 10**-5 (mass fraction, 132 times solar) and 3.0 X 10**-5 (56 600 times solar), respectively; we also found Cu overabundances in the DA-type WD G191-B2B (6.3 X 10**-6, 9 times solar). All identified Cu IV - VI lines in the UV spectrum of RE 0503-289 were simultaneously well reproduced with our newly calculated oscillator strengths. With the detection of Cu and In in RE 0503-289, the total number of trans-iron elements (Z > 28) in this extraordinary WD reaches an unprecedented number of 18., Comment: 32 pages, 7 figures

We have identified 484 lines of the trans-iron elements (TIEs) Zn, Ga, Ge, Se, Br, Kr, Sr, Zr, Mo, In, Te, I, Xe, and Ba, for the first time in the ultraviolet spectrum of a DAO-type WD, namely BD$-22{^\circ}3467$, surrounded by the ionized nebula Abell 35. Our TIE abundance determination shows extremely high overabundances of up to five dex -- a similar effect is already known from hot, H-deficient (DO-type) white dwarfs. In contrast to these where a pulse-driven convection zone has enriched the photosphere with TIEs during a final thermal pulse and radiative levitation has established the extreme TIE overabundances, {here the extreme TIE overabundances are exclusively driven by radiative levitation on the initial stellar metallicity. The very low mass ($0.533^{+0.040}_{-0.025}\,M_\odot$) of BD$-22{^\circ}3467$ implies that a third dredge-up with enrichment of s-process elements in the photosphere did not occur in the AGB precursor., Comment: 34 pages, 19 Figures

Stellar post asymptotic giant branch (post-AGB) evolution can be completely altered by a final thermal pulse (FTP) which may occur when the star is still leaving the AGB (AFTP), at the departure from the AGB at still constant luminosity (late TP, LTP) or after the entry to the white-dwarf cooling sequence (very late TP, VLTP). Then convection mixes the He-rich material with the H-rich envelope. According to stellar evolution models the result is a star with a surface composition of $\mathrm{H}\approx\,20\,$% by mass (AFTP), $\approx 1\,$% (LTP), or (almost) no H (VLTP). Since FTP stars exhibit intershell material at their surface, spectral analyses establish constraints for AGB nucleosynthesis and stellar evolution. We performed a spectral analysis of the so-called hybrid PG 1159-type central stars (CS) of the planetary nebulae Abell 43 and NGC7094 by means of non-local thermodynamical equilibrium models. We confirm the previously determined effective temperatures of $T_\mathrm{eff} = 115\,000\pm 5\,000\,$K and determine surface gravities of $\log (g\,/\,\mathrm{cm/s^2}) = 5.6\pm 0.1$ for both. From a comparison with AFTP evolutionary tracks, we derive stellar masses of $0.57^{+0.07}_{-0.04}\,M_\odot$ and determine the abundances of H, He, and metals up to Xe. Both CS are likely AFTP stars with a surface H mass fraction of $0.25 \pm 0.03$ and $0.15 \pm 0.03$, respectively, and a Fe deficiency indicating subsolar initial metallicities. The light metals show typical PG 1159-type abundances and the elemental composition is in good agreement with predictions from AFTP evolutionary models. However, the expansion ages do not agree with evolution timescales expected from the AFTP scenario and alternatives should be explored., Comment: 27 pages, 22 figures

There is a striking paucity of hydrogen-rich (DA) white dwarfs (WDs) relative to their hydrogen-deficient (non-DA) counterparts at the very hot end of the WD cooling sequence. The three hottest known DAs (surface gravity log g $\geq$ 7.0) have effective temperatures around Teff = 140,000 K, followed by only five objects in the range 104,000 - 120,000 K. They are by far outnumbered by forty non-DAs with Teff = 100,000 - 250,000 K, giving a DA/non-DA ratio of 0.2. In contrast, this ratio is the inverse of that for the cooler WDs. One reason for this discrepancy could be uncertainties in the temperature determination of hot DAs using Balmer-line spectroscopy. Recent investigations involving metal-ionization balances in ultraviolet (UV) spectra indeed showed that the temperatures of some DAs were underestimated, but the paucity of extremely hot DAs prevailed. Here we present the results of a UV spectral analysis of one of the three hottest DAs, PG0948+534. We find that its temperature was strongly overestimated by recent Balmer line analyses. We correct it downward to 105,000 $\pm$ 5000 K, aggravating the hot-DA paucity. The Balmer-line problem encountered previously is not resolved by our non-LTE line-blanketed model atmospheres. We speculate that it might be related to the possible presence of a magnetosphere. This is supported by the V-band variability that shows a period of P=3.45 d (amplitude 0.19 mag), which we interpret as the star's rotation period. The metal abundances in PG0948+534 are affected by atomic diffusion and we conclude that the onset of diffusion in hot DAs occurs when they cool below Teff about 105,000 K. We discuss the possibility that the paucity of very hot DAs is a consequence of their fast evolutionary rate., Comment: Accepted for publication in MNRAS

The presence of ultra-high excitation (UHE) absorption lines (e.g., O VIII) in the optical spectra of several of the hottest white dwarfs poses a decades-long mystery and is something that has never been observed in any other astrophysical object. The occurrence of such features requires a dense environment with temperatures near $10^6$K, by far exceeding the stellar effective temperature. Here we report the discovery of a new hot wind white dwarf, GALEXJ014636.8+323615. Astonishingly, we found for the first time rapid changes of the equivalent widths of the UHE features, which are correlated to the rotational period of the star ($P=0.242035$d). We explain this with the presence of a wind-fed circumstellar magnetosphere in which magnetically confined wind shocks heat up the material to the high temperatures required for the creation of the UHE lines. The photometric and spectroscopic variability of GALEXJ014636.8+323615 can then be understood as consequence of the obliquity of the magnetic axis with respect to the rotation axis of the white dwarf. This is the first time a wind-fed circumstellar magnetosphere around an apparently isolated white dwarf has been discovered and finally offers a plausible explanation of the ultra hot wind phenomenon., Comment: Published in MNRAS Letters