Search

Many complex fluids can be described by continuum hydrodynamic field equations, to which noise must be added in order to capture thermal fluctuations. In almost all cases, the resulting coarse-grained stochastic partial differential equations carry a short-scale cutoff -- which is also reflected in numerical discretisation schemes. We draw together our recent findings concerning the construction of such schemes and the interpretation of their continuum limits, focusing for simplicity on models with a purely diffusive scalar field, such as `Model B' which describes phase separation in binary fluid mixtures. We address the requirement that the steady state entropy production rate (EPR) must vanish for any stochastic hydrodynamic model in thermal equilibrium. Only if this is achieved can the given discretisation scheme be relied upon to correctly calculate the nonvanishing EPR for `active field theories' in which new terms are deliberately added to the fluctuating hydrodynamic equations that break detailed balance. To compute the correct probabilities of forward and time-reversed paths (whose ratio determines the EPR) we must make a careful treatment of so-called `spurious drift' and other closely related terms that depend on the discretisation scheme. We show that such subtleties can arise not only in the temporal discretisation (as is well documented for stochastic ODEs with multiplicative noise) but also from spatial discretisation, even when noise is additive, as most active field theories assume. We then review how such noise can become multiplicative, via off-diagonal couplings to additional fields that encode thermodynamically the underlying chemical processes responsible for activity. In this case the spurious drift terms need careful accounting, not just to evaluate correctly the EPR, but also to numerically implement the Langevin dynamics itself., Comment: 27 pages, 4 figures

The exploration of the universe has recently entered a new era thanks to the multi-messenger paradigm, characterized by a continuous increase in the quantity and quality of experimental data that is obtained by the detection of the various cosmic messengers (photons, neutrinos, cosmic rays and gravitational waves) from numerous origins. They give us information about their sources in the universe and the properties of the intergalactic medium. Moreover, multi-messenger astronomy opens up the possibility to search for phenomenological signatures of quantum gravity. On the one hand, the most energetic events allow us to test our physical theories at energy regimes which are not directly accessible in accelerators; on the other hand, tiny effects in the propagation of very high energy particles could be amplified by cosmological distances. After decades of merely theoretical investigations, the possibility of obtaining phenomenological indications of Planck-scale effects is a revolutionary step in the quest for a quantum theory of gravity, but it requires cooperation between different communities of physicists (both theoretical and experimental). This review is aimed at promoting this cooperation by giving a state-of-the art account of the interdisciplinary expertise that is needed in the effective search of quantum gravity footprints in the production, propagation and detection of cosmic messengers., Comment: This review was written by participants of the COST Action CA18108. Further information on the review can be found at https://qg-mm.unizar.es/review/, updated to published version

Multi-access Edge Computing (MEC) is expected to act as the enabler for the integration of 5G (and future 6G) communication technologies with cloud-computing-based capabilities at the edge of the network. This will enable low-latency and context-aware applications for users of such mobile networks. In this paper we describe the implementation of a MEC model for the Simu5G simulator and illustrate how to configure the environment to evaluate MEC applications in both simulation and real-time emulation modes., Comment: Published in: M. Marek, G. Nardini, V. Vesely (Eds.), Proceedings of the 8th OMNeT++ Community Summit, Virtual Summit, September 8-10, 2021

These are the Proceedings of the 8th OMNeT++ Community Summit, which was held virtually on September 8-10, 2021.

We present a study of the relation between X-rays and ultraviolet emission in quasars for a sample of broad-line, radio-quiet objects obtained from the cross-match of the Sloan Digital Sky Survey DR14 with the latest Chandra Source Catalog 2.0 (2,332 quasars) and the Chandra COSMOS Legacy survey (273 quasars). The non-linear relation between the ultraviolet (at 2500 A, $L_{O}$) and the X-ray (at 2 keV, $L_{X}$) emission in quasars has been proved to be characterised by a smaller intrinsic dispersion than the observed one, as long as a homogeneous selection, aimed at preventing the inclusion of contaminants in the sample, is fulfilled. By leveraging on the low background of Chandra, we performed a complete spectral analysis of all the data available for the SDSS-CSC2.0 quasar sample (i.e. 3,430 X-ray observations), with the main goal of reducing the uncertainties on the source properties (e.g. flux, spectral slope). We analysed whether any evolution of the $L_{X}-L_{O}$ relation exists by dividing the sample in narrow redshift intervals across the redshift range spanned by our sample, $z \simeq 0.5-4$. We find that the slope of the relation does not evolve with redshift and it is consistent with the literature value of $0.6$ over the explored redshift range, implying that the mechanism underlying the coupling of the accretion disc and hot corona is the same at the different cosmic epochs. We also find that the dispersion decreases when examining the highest redshifts, where only pointed observations are available. These results further confirm that quasars are `standardisable candles', that is we can reliably measure cosmological distances at high redshifts where very few cosmological probes are available., Comment: 20 pages, 14 figures, accepted for publication in A&A

We present the first high signal-to-noise broadband X-ray spectrum of the radio-quiet type-2 Seyfert ESO 033-G002, combining data from $XMM$-$Newton$ and $NuSTAR$. The nuclear X-ray spectrum is complex, showing evidence for both neutral and ionised absorption, as well as reflection from both the accretion disc and more distant material, but our broadband coverage allows us to disentangle all of these different components. The total neutral column during this epoch is $N_{\rm{H}} \sim 5-6 \times 10^{22}$ cm$^{-2}$, consistent with the optical classification of ESO 033-G002 as a type-2 Seyfert but not so large as to prevent us from robustly determining the properties of the innermost accretion flow. The ionised absorption - dominated by lines from Fe XXV and Fe XXVI - reveals a moderately rapid outflow ($v_{\rm{out}} \sim 5400$ km s$^{-1}$) which has a column comparable to the neutral absorption. We find the disc reflection from the innermost regions to be extreme, with a reflection fraction of $R_{\rm{frac}} \sim 5$. This requires strong gravitational lightbending and, in turn, both an extremely compact corona (within $\sim$2 $R_{\rm{G}}$ of the black hole) and a rapidly rotating black hole ($a^* > 0.96$). Despite this tight size constraint, with a temperature of $kT_{\rm{e}} = 40-70$ keV the X-ray corona in ESO 033-G002 appears similar to other AGN in terms of its placement in the compactness-temperature plane, consistent with sitting close to the limit determined by runaway pair production. Finally, combining X-ray spectroscopy, timing and updated optical spectroscopy, we also estimate the mass of the black hole to be $\log[M_{\rm{BH}} / M_{\odot}] \sim 7.0 - 7.5$., Comment: 15 pages, 8 figures; published in MNRAS

The science objectives of the LISA mission have been defined under the implicit assumption of a 4 yr continuous data stream. Based on the performance of LISA Pathfinder, it is now expected that LISA will have a duty cycle of $\approx 0.75$, which would reduce the effective span of usable data to 3 yr. This paper reports the results of a study by the LISA Science Group, which was charged with assessing the additional science return of increasing the mission lifetime. We explore various observational scenarios to assess the impact of mission duration on the main science objectives of the mission. We find that the science investigations most affected by mission duration concern the search for seed black holes at cosmic dawn, as well as the study of stellar-origin black holes and of their formation channels via multi-band and multi-messenger observations. We conclude that an extension to 6 yr of mission operations is recommended., Comment: 50 pages, 19 figures, 5 tables. Matches version published in GERG

We analyse the properties of the CIV broad emission line in connection with the X-ray emission of 30 bright SDSS quasars at z~3.0-3.3 with pointed XMM-Newton observations, which were selected to test the suitability of AGN as cosmological tools. In our previous work, we found that a large fraction (~25%) of the quasars in this sample are X-ray underluminous by factors of >3-10. As absorbing columns of >10$^{23}$ cm$^{-2}$ can be safely ruled out, their weakness is most likely intrinsic. Here we explore possible correlations between the UV and X-ray features of these sources to investigate the origin of X-ray weakness. We fit their UV SDSS spectra and analyse their CIV properties (e.g., equivalent width, EW; line peak velocity, $\upsilon_{\rm peak}$) as a function of the X-ray photon index and 2-10 keV flux. We confirm the trends of CIV $\upsilon_{\rm peak}$ and EW with UV luminosity at 2500 angstrom for both X-ray weak and X-ray normal quasars, as well as the correlation between X-ray weakness and CIV EW. In contrast to some recent work, we do not observe any clear relation between the 2-10 keV luminosity and $\upsilon_{\rm peak}$. We find a correlation between the hard X-ray flux and the integrated CIV flux for X-ray normal quasars, whilst X-ray weak quasars deviate from the main trend by more than 0.5 dex. We argue that X-ray weakness might be interpreted in a starved X-ray corona picture associated with an ongoing disc-wind phase. If the wind is ejected in the vicinity of the black hole, the extreme-UV radiation that reaches the corona will be depleted, depriving the corona of seeds photons and generating an X-ray weak quasar. Yet, at the largest UV luminosities (>10$^{47}$ erg s$^{-1}$), there will still be an ample reservoir of ionising photons that can explain the excess CIV emission observed in the X-ray weak quasars with respect to normal sources of similar X-ray luminosities., Comment: 22 pages, 15 figures (with 3 more figures in the Appendix), abstract abridged. Accepted for publication in A&A

Experiments of periodically sheared colloidal suspensions or soft amorphous solids display a transition from reversible to irreversible particle motion that, when analysed stroboscopically in time, is interpreted as an absorbing phase transition with infinitely many absorbing states. In these systems interactions mediated by hydrodynamics or elasticity are present, causing passive regions to be affected by nearby active ones. We show that mediated interactions induce a new universality class of absorbing phase transitions, distinct from Conserved Directed Percolation, and we obtain the corresponding critical exponents. We do so with large-scale numerical simulations of a minimal model for the stroboscopic dynamics of sheared soft materials and derive the minimal field theoretical description., Comment: 4 pages, 4 figures

Modern search engine ranking pipelines are commonly based on large machine-learned ensembles of regression trees. We propose LEAR, a novel - learned - technique aimed to reduce the average number of trees traversed by documents to accumulate the scores, thus reducing the overall query response time. LEAR exploits a classifier that predicts whether a document can early exit the ensemble because it is unlikely to be ranked among the final top-k results. The early exit decision occurs at a sentinel point, i.e., after having evaluated a limited number of trees, and the partial scores are exploited to filter out non-promising documents. We evaluate LEAR by deploying it in a production-like setting, adopting a state-of-the-art algorithm for ensembles traversal. We provide a comprehensive experimental evaluation on two public datasets. The experiments show that LEAR has a significant impact on the efficiency of the query processing without hindering its ranking quality. In detail, on a first dataset, LEAR is able to achieve a speedup of 3x without any loss in NDCG1@0, while on a second dataset the speedup is larger than 5x with a negligible NDCG@10 loss (< 0.05%)., Comment: 5 pages, 3 figures, ACM SIGIR Conference on Research and Development in Information Retrieval (SIGIR 21)

Since 2015 the gravitational-wave observations of LIGO and Virgo have transformed our understanding of compact-object binaries. In the years to come, ground-based gravitational-wave observatories such as LIGO, Virgo, and their successors will increase in sensitivity, discovering thousands of stellar-mass binaries. In the 2030s, the space-based LISA will provide gravitational-wave observations of massive black holes binaries. Between the $\sim 10$-$10^3~\mathrm{Hz}$ band of ground-based observatories and the $\sim10^{-4}$-$10^{-1}~\mathrm{Hz}$ band of LISA lies the uncharted decihertz gravitational-wave band. We propose a Decihertz Observatory to study this frequency range, and to complement observations made by other detectors. Decihertz observatories are well suited to observation of intermediate-mass ($\sim10^2$-$10^4 M_\odot$) black holes; they will be able to detect stellar-mass binaries days to years before they merge, providing early warning of nearby binary neutron star mergers and measurements of the eccentricity of binary black holes, and they will enable new tests of general relativity and the Standard Model of particle physics. Here we summarise how a Decihertz Observatory could provide unique insights into how black holes form and evolve across cosmic time, improve prospects for both multimessenger astronomy and multiband gravitational-wave astronomy, and enable new probes of gravity, particle physics and cosmology., Comment: 13 pages, 1 figure. Published in Experimental Astronomy. Summarising white paper arXiv:1908.11375

The classification of the optical spectra of active galactic nuclei (AGN) into different types is well founded on AGN physics, but it involves some degree of human oversight and cannot be reliably scaled to large data sets. Machine learning (ML) tackles such a classification problem in a fast and reproducible way, but is often perceived as a black box. However, ML interpretability and explainability are active research areas in computer science, increasingly providing us with tools to alleviate this issue. We applied ML interpretability tools to a classifier trained to predict AGN type from spectra, to demonstrate the use of such tools in this context. We trained a support-vector machine on 3346 high-quality, low redshift AGN spectra from SDSS DR15 with an existing reliable classification as type 1, type 2, or intermediate type. On a selection of test-set spectra, we computed the gradient of the predicted class probability and we built saliency maps. We also visualized the high-dimensional space of AGN spectra using t-distributed stochastic neighbor embedding (t-SNE), showing where the spectra for which we computed a saliency map are located. Regions that affect the predicted AGN type often coincide with physically relevant features, such as spectral lines. t-SNE visualization shows good separability of type 1 and type 2 spectra, while intermediate-type spectra either lie in-between as expected or appear mixed with type 2 spectra. Saliency maps show why a given AGN type was predicted by our classifier, resulting in a physical interpretation in terms of regions of the spectrum that affected its decision, making it no longer a black box. These regions coincide with those used by human experts such as relevant spectral lines, and are even used in a similar way, with the classifier e.g. effectively measuring the width of a line by weighing its center and its tails oppositely., Comment: Under review by A&A

In passive fluid-fluid phase separation, a single interfacial tension sets both the capillary fluctuations of the interface and the rate of Ostwald ripening. We show that these phenomena are governed by two different tensions in active systems, and compute the capillary tension $\sigma_{cw}$ which sets the relaxation rate of interfacial fluctuations in accordance with capillary wave theory. We discover that strong enough activity can cause negative $\sigma_{cw}$. In this regime, depending on the global composition, the system self-organizes, either into a microphase-separated state in which coalescence is highly inhibited, or into an `active foam' state. Our results are obtained for Active Model B+, a minimal continuum model which, although generic, admits significant analytical progress., Comment: as accepted for publication in Phys. Rev. Lett

We study the confinement/deconfinement phase transition of the radion field in a warped model with a polynomial bulk potential. The backreaction of the radion on the metric is taken into account by using the superpotential formalism, while the radion effective potential is obtained from a novel formulation which can incorporate the backreaction. The phase transition leads to a stochastic gravitational wave background that depends on the energy scale of the first Kaluza-Klein resonance, $m_{\textrm{KK}}$. This work completes previous studies in the following aspects: i) we detail the evaluation of the radion spectrum; ii) we report on the mismatches between the thick wall approximation and the numerical bounce solution; iii) we include a suppression factor in the spectrum of sound waves accounting for their finite lifetime; and, iv) we update the bound on $m_{\textrm{KK}}$ in view of the O3 LIGO and Virgo data. We find that the forthcoming gravitational wave interferometers can probe scenarios where $m_{\textrm{KK}} \lesssim 10^9$ TeV, while the O3-run bounds rule out warped models with $10^4 \textrm{TeV} \lesssim m_{\textrm{KK}} \lesssim 10^7$ TeV exhibiting an extremely strong confinement/deconfinement phase transition., Comment: 16 pages, 7 figures; v2 extended version: added references and Figs. 2, 3, 5 and 7 (lower panels), Figs. 6 and 7 (upper panels) updated, extended discussion in Secs. 3.3, 4, 5 and 6. Talk given by E.Megias at the 9th International Conference on New Frontiers in Physics (ICNFP 2020), 4 Sep - 2 Oct 2020, Kolymbari, Crete, Greece

Outflows driven by active galactic nuclei (AGN) are expected to have a significant impact on the host galaxy evolution, but it is still debated how they are accelerated and propagate on galaxy-wide scales. This work addresses these questions by studying the link between X-ray, nuclear ultra-fast outflows (UFOs) and extended ionised outflows, for the first time in two quasars close to the peak of AGN activity ($z\sim2$), where AGN feedback is expected to be more effective. As targets, we selected two multiple-lensed quasars at $z\sim1.5$, HS 0810+2554 and SDSS J1353+1138, known to host UFOs and observed with the near-IR integral field spectrometer SINFONI at the VLT. We performed a kinematical analysis of the [O III]$\lambda$5007 optical emission line, in order to trace the presence of ionised outflows. We detected spatially resolved ionised outflows in both galaxies, extended more than 8 kpc and moving up to $v>2000$ km/s. We derived mass outflow rates of $\sim$12 M$_{sun}$/yr and $\sim$2 M$_{sun}$/yr for HS 0810+2554 and SDSS J1353+1138. Comparing with the co-hosted UFO energetics, the ionised outflow energetics in HS 0810+2554 is broadly consistent with a momentum-driven regime of wind propagation, while in SDSS J1353+1138 it differs by a factor of $\sim$100 from theoretical predictions, requiring either a massive molecular outflow or a high variability of the AGN activity to account for such a discrepancy. By additionally considering our results with those from the small sample of well-studied objects (all local but one), with both UFO and extended (ionised/atomic/molecular) outflow detections, we found that in 10 out of 12 galaxies the large-scale outflow energetics is consistent with the theoretical predictions of either a momentum- or an energy-driven scenario. This suggests that such models explain relatively well the acceleration mechanism of AGN-driven winds on large scales., Comment: 19 pages, 9 figures. Accepted for publication in A&A

Angiogenesis is the process by which blood vessels form from pre-existing vessels. It plays a key role in many biological processes, including embryonic development and wound healing, and contributes to many diseases including cancer and rheumatoid arthritis. The structure of the resulting vessel networks determines their ability to deliver nutrients and remove waste products from biological tissues. Here we simulate the Anderson-Chaplain model of angiogenesis at different parameter values and quantify the vessel architectures of the resulting synthetic data. Specifically, we propose a topological data analysis (TDA) pipeline for systematic analysis of the model. TDA is a vibrant and relatively new field of computational mathematics for studying the shape of data. We compute topological and standard descriptors of model simulations generated by different parameter values. We show that TDA of model simulation data stratifies parameter space into regions with similar vessel morphology. The methodologies proposed here are widely applicable to other synthetic and experimental data including wound healing, development, and plant biology.

Neural Networks (NN), although successfully applied to several Artificial Intelligence tasks, are often unnecessarily over-parametrised. In edge/fog computing, this might make their training prohibitive on resource-constrained devices, contrasting with the current trend of decentralising intelligence from remote data centres to local constrained devices. Therefore, we investigate the problem of training effective NN models on constrained devices having a fixed, potentially small, memory budget. We target techniques that are both resource-efficient and performance effective while enabling significant network compression. Our Dynamic Hard Pruning (DynHP) technique incrementally prunes the network during training, identifying neurons that marginally contribute to the model accuracy. DynHP enables a tunable size reduction of the final neural network and reduces the NN memory occupancy during training. Freed memory is reused by a \emph{dynamic batch sizing} approach to counterbalance the accuracy degradation caused by the hard pruning strategy, improving its convergence and effectiveness. We assess the performance of DynHP through reproducible experiments on three public datasets, comparing them against reference competitors. Results show that DynHP compresses a NN up to $10$ times without significant performance drops (up to $3.5\%$ additional error w.r.t. the competitors), reducing up to $80\%$ the training memory occupancy.

Agent-based models provide a flexible framework that is frequently used for modelling many biological systems, including cell migration, molecular dynamics, ecology, and epidemiology. Analysis of the model dynamics can be challenging due to their inherent stochasticity and heavy computational requirements. Common approaches to the analysis of agent-based models include extensive Monte Carlo simulation of the model or the derivation of coarse-grained differential equation models to predict the expected or averaged output from the agent-based model. Both of these approaches have limitations, however, as extensive computation of complex agent-based models may be infeasible, and coarse-grained differential equation models can fail to accurately describe model dynamics in certain parameter regimes. We propose that methods from the equation learning field provide a promising, novel, and unifying approach for agent-based model analysis. Equation learning is a recent field of research from data science that aims to infer differential equation models directly from data. We use this tutorial to review how methods from equation learning can be used to learn differential equation models from agent-based model simulations. We demonstrate that this framework is easy to use, requires few model simulations, and accurately predicts model dynamics in parameter regions where coarse-grained differential equation models fail to do so. We highlight these advantages through several case studies involving two agent-based models that are broadly applicable to biological phenomena: a birth-death-migration model commonly used to explore cell biology experiments and a susceptible-infected-recovered model of infectious disease spread.

Outflows accelerated by AGN are commonly observed in the form of coherent, mildly collimated high-velocity gas directed along the AGN ionisation cones and kinetically powerful (>$10^{44-45}$ erg/s) jets. Recent works found that outflows can also be accelerated by low-power (<$10^{44}$ erg/s) jets, and the most recent cosmological simulations indicate that these are the dominant source of feedback on sub-kpc scales. We study the relation between radio jets and the distribution and kinematics of the ionised gas in IC 5063, NGC 5643, NGC 1068, and NGC 1386 as part of our MAGNUM survey of nearby Seyfert galaxies. All these objects host a small-scale (<1 kpc) low-power (<$10^{44}$ erg/s) radio jet that has small inclinations (<45\deg) with respect to the galaxy disc. We employed seeing-limited optical integral field spectroscopic observations from MUSE at VLT to obtain flux, kinematic, and excitation maps of the extended ionised gas, that we compared with archival radio images and Chandra X-ray observations. We detect a strong (up to >800-1000 km/s), extended (>1 kpc) and shock-excited emission-line velocity spread perpendicular to the AGN ionisation cones and jets in all four targets. These broad and symmetric line profiles are not associated with a single coherent velocity of the gas, differently from the 'classical' asymmetric-line outflow observed along the ionisation cones and jets. We interpret the observed phenomenon as due to the action of the jets perturbing the gas in the galaxy disc. These intense and extended velocity spreads perpendicular to AGN jets and cones are indeed currently only observed in galaxies hosting a low-power jet whose inclination is sufficiently low with respect to the galaxy disc to impact on and strongly affect its material. In line with cosmological simulations, our results demonstrate that low-power jets are indeed capable of affecting the host galaxy., Comment: 19 pages, 13 figures. Astronomy & Astrophysics, in press

New Swift monitoring observations of the variable, radio-quiet quasar, PDS 456, are presented. A bright X-ray flare was captured in September 2018, the flux increasing by a factor of 4 and with a doubling time-scale of 2 days. From the light crossing argument, the coronal size is inferred to be about 30 gravitational radii for a black hole mass of $10^{9} {\rm M}_{\odot}$ and the total flare energy exceeds $10^{51}$ erg. A hardening of the X-ray emission accompanied the flare, with the photon index decreasing from $\Gamma=2.2$ to $\Gamma=1.7$ and back again. The flare is produced in the X-ray corona, the lack of any optical or UV variability being consistent with a constant accretion rate. Simultaneous XMM-Newton and NuSTAR observations were performed, $1-3$ days after the flare peak and during the decline phase. These caught PDS 456 in a bright, bare state, where no disc wind absorption features are apparent. The hard X-ray spectrum shows a high energy roll-over, with an e-folding energy of $E_{\rm fold}=51^{+11}_{-8}$ keV. The deduced coronal temperature, of $kT=13$ keV, is one of the coolest measured in any AGN and PDS 456 lies well below the predicted pair annihilation line in X-ray corona. The spectral variability, becoming softer when fainter following the flare, is consistent with models of cooling X-ray coronae. Alternatively, an increase in a non-thermal component could contribute towards the hard X-ray flare spectrum., Comment: 18 pages, 16 figures, accepted for publication in MNRAS

Many computer science disciplines (e.g., combinatorial optimization, natural language processing, and information retrieval) use standard or established test suites for evaluating algorithms. In visualization, similar approaches have been adopted in some areas (e.g., volume visualization), while user testimonies and empirical studies have been the dominant means of evaluation in most other areas, such as designing colormaps. In this paper, we propose to establish a test suite for evaluating the design of colormaps. With such a suite, the users can observe the effects when different continuous colormaps are applied to planar scalar fields that may exhibit various characteristic features, such as jumps, local extrema, ridge or valley lines, different distributions of scalar values, different gradients, different signal frequencies, different levels of noise, and so on. The suite also includes an expansible collection of real-world data sets including the most popular data for colormap testing in the visualization literature. The test suite has been integrated into a web-based application for creating continuous colormaps (https://ccctool.com/), facilitating close inter-operation between design and evaluation processes. This new facility complements traditional evaluation methods such as user testimonies and empirical studies.

We present a data analysis methodology for a model-independent reconstruction of the spectral shape of a stochastic gravitational wave background with LISA. We improve a previously proposed reconstruction algorithm that relied on a single Time-Delay-Interferometry (TDI) channel by including a complete set of TDI channels. As in the earlier work, we assume an idealized equilateral configuration. We test the improved algorithm with a number of case studies, including reconstruction in the presence of two different astrophysical foreground signals. We find that including additional channels helps in different ways: it reduces the uncertainties on the reconstruction; it makes the global likelihood maximization less prone to falling into local extrema; and it efficiently breaks degeneracies between the signal and the instrumental noise., Comment: 35 pages, 12 figures

We use MUSE adaptive optics (AO) data in Narrow Field Mode to study the properties of the ionised gas in MR 2251-178 and PG 1126-041, two nearby (z~0.06) bright quasars hosting sub-pc scale Ultra Fast Outflows (UFOs) detected in the X-ray band. We decompose the optical emission from diffuse gas into a low- and a high-velocity components. The former is characterised by a clean, regular velocity field and a low (~80 km/s) velocity dispersion. It traces regularly rotating gas in PG 1126-041, while in MR 2251-178 it is possibly associated to tidal debris from a recent merger or flyby. The other component is found to be extended up to a few kpc from the nuclei, and shows a high (~800 km/s) velocity dispersion and a blue-shifted mean velocity, as expected from AGN-driven outflows. We estimate mass outflow rates up to a few Mo/yr and kinetic efficiencies between 0.1-0.4 per cent, in line with those of galaxies hosting AGNs of similar luminosity. The momentum rates of these ionised outflows are comparable to those measured for the UFOs at sub-pc scales, consistent with a momentum-driven wind propagation. Pure energy-driven winds are excluded unless about 100x additional momentum is locked in massive molecular winds. By comparing the outflow properties of our sources with those of a small sample of well-studied QSOs hosting UFOs from the literature, we find that winds seem to systematically lie either in a momentum-driven or in an energy-driven regime, indicating that these two theoretical models bracket very well the physics of AGN-driven winds., Comment: 18 pages, 10 figures. Key figures are 4 and 9. Accepted for publication in A&A

We present new broadband X-ray observations of the type-I Seyfert galaxy IRAS 09149-6206, taken in 2018 with $XMM$-$Newton$, $NuSTAR$ and $Swift$. The source is highly complex, showing a classic 'warm' X-ray absorber, additional absorption from highly ionised iron, strong relativistic reflection from the innermost accretion disc and further reprocessing by more distant material. By combining X-ray timing and spectroscopy, we have been able to fully characterise the supermassive black hole in this system, constraining both its mass and - for the first time - its spin. The mass is primarily determined by X-ray timing constraints on the break frequency seen in the power spectrum, and is found to be $\log[M_{\rm{BH}}/M_{\odot}] = 8.0 \pm 0.6$ (1$\sigma$ uncertainties). This is in good agreement with previous estimates based on the H$\alpha$ and H$\beta$ line widths, and implies that IRAS 09149-6206 is radiating at close to (but still below) its Eddington luminosity. The spin is constrained via detailed modelling of the relativistic reflection, and is found to be $a^* = 0.94^{+0.02}_{-0.07}$ (90% confidence), adding IRAS 09149-6206 to the growing list of radio-quiet AGN that host rapidly rotating black holes. The outflow velocities of the various absorption components are all relatively modest ($v_{\rm{out}} \lesssim 0.03c$), implying these are unlikely to drive significant galaxy-scale AGN feedback., Comment: 19 pages, 15 figures, accepted for publication in MNRAS

We present a reanalysis of the cumulative ACIS S Chandra data set pointed at the double AGNs of the NGC 6240 merging galaxy, focusing on the hard energy bands containing the hard spectral continuum (5.5-5.9 keV), the redshifted Fe I K alpha line (6.0-6.4 keV), and the redshifted Fe XXV line (6.4-6.7 keV). We have used to the full the Chandra telescope angular resolution, and we have modeled the Chandra PSF by comparing pre-flight calibration model to the data for the two bright AGNs. With two complementary approaches: (1) studying the residuals after PSF subtraction, and (2) producing reconstructed Expectation through Markov Chain Monte Carlo (EMC2) images, we are able to resolve structures extending from 1 kpc to <200 pc in the S AGN. The latter are within the sphere of influence of this BH. We find significant extended emission in both continuum and Fe lines in the 2'' (1 kpc) region surrounding the nuclei, in the region between the N and S AGN, and in a sector of PA 120-210 deg. extending to the SE from the centroid of the S AGN surface brightness. The extended Fe I K alpha emission is likely to originate from fluorescence of X-ray photons interacting with dense molecular clouds, providing a complementary view to recent high-resolution ALMA studies. The non-thermal emission is more prevalent in the region in between the two active X-ray nuclei, and in the N AGN. We do not find strong evidence of X-ray emission associated with the 3rd nucleus recently proposed for NGC 6240., Comment: in press on The Astrophysical Journal

We review and extend recent developments on the statistical properties of Active Ornstein Uhlenbeck particles (AOUPs). In this simplest of models, the Gaussian white noise of overdamped Brownian colloids is replaced by a Gaussian colored noise. This suffice to grant this system the hallmark properties of active matter, while still allowing for analytical progress. We first detail the perturbative derivation of the steady state of AOUPs in the small persistence time limit. We show the existence of an effective equilibrium regime in which detailed-balance is obeyed with respect to a non-Boltzmann distribution and detail the corresponding fluctuation-dissipation theorem. We then characterize the departure from equilibrium by computing several relevant observables (entropy production, ratchet current). At the collective level, we show AOUPs to experience motility-induced phase separation both in the presence of pairwise forces or due to quorum-sensing interactions. The latter can be accounted for by considering the steady-state of AOUPs with spatially varying propulsion speed or persistence time. Finally, we discuss how the emerging properties of AOUPs can be characterized from the dynamics of their collective modes, which we construct explicitly.

The properties of large-scale galactic outflows, such as their kinetic energy and momentum rates, correlate with the luminosity of the active galactic nucleus (AGN). This is well explained by the wind-driven outflow model, where a fraction of the AGN luminosity drives the outflow. However, significant departures from these correlations have been observed in a number of galaxies. This may happen because AGN luminosity varies on a much shorter timescale ($\sim 10^4-10^5$~yr) than outflow properties do ($\sim 10^6$~yr). We investigate the effect of AGN luminosity variations on outflow properties using 1D numerical simulations. This effect can explain the very weak outflow in PDS 456: if its nucleus is currently much brighter than the long-term average luminosity, the outflow has not had time to react to this luminosity change. Conversely, the outflow in Mrk 231 is consistent with being driven by an almost continuous AGN, while IRAS F11119+3257 represents an intermediate case between the two. Considering a population of AGN, we find that very low momentum loading factors $\dot{p}_{\rm out} < L_{\rm AGN}/c$ should be seen in a significant fraction of objects - up to $15\%$ depending on the properties of AGN variability and galaxy gas fraction. The predicted distribution of loading factors is consistent with the available observational data. We discuss how this model might help constrain the duty cycles of AGN during the period of outflow inflation, implications for multiphase and spatially distinct outflows, and suggest ways of improving AGN prescriptions in numerical simulations., Comment: 15 pages, 7 figures; accepted for publication in MNRAS

We adopt the Pearson cross-correlation measure to analyze the LIGO Hanford and LIGO Livingston detector data streams around the events GW150914, GW151012,GW151226 and GW170104. We find that the Pearson cross-correlation method is sensitive to these signals, with correlations peaking when the black hole binaries reconstructed by the LIGO Scientific and Virgo Collaborations, are merging. We compare the obtained cross-correlations with the statistical correlation fluctuations arising in simulated Gaussian noise data and in LIGO data at times when no event is claimed. Our results for the significance of the observed cross-correlations are broadly consistent with those announced by the LIGO Scientific and Virgo Collaborations based on matched-filter analysis. In the same data, if we subtract the maximum likelihood waveforms corresponding to the announced signals, no residual cross-correlations persists at a statistically significant level., Comment: 17 pages, 6 figures; matches published version

We present a new catalogue of ~2,400 optically selected quasars with spectroscopic redshifts and X-ray observations from either Chandra or XMM-Newton. The sample can be used to investigate the non-linear relation between the UV and X-ray luminosity of quasars, and to build a Hubble diagram up to redshift z~7.5. We selected sources that are neither reddened by dust in the optical/UV nor obscured by gas in the X-rays, and whose X-ray fluxes are free from flux-limit related biases. After checking for any possible systematics, we confirm, in agreement with our previous works, that (i) the X-ray to UV relation provides distance estimates matching those from supernovae up to z~1.5, and (ii) its slope shows no redshift evolution up to z~5. We provide a full description of the methodology for testing cosmological models, further supporting a trend whereby the Hubble diagram of quasars is well reproduced by the standard flat $\Lambda$CDM model up to z~1.5-2, but strong deviations emerge at higher redshifts. Since we have minimized all non-negligible systematic effects, and proven the stability of the $L_{\rm X}-L_{\rm UV}$ relation at high redshifts, we conclude that an evolution of the expansion rate of the Universe should be considered as a possible explanation for the observed deviation, rather than some systematic (redshift-dependent) effect associated with high-redshift quasars., Comment: 24 pages, 17 figures, accepted for publication in A&A

When a black hole accretes close to the Eddington limit, the astrophysical jet is often accompanied by radiatively driven, wide-aperture and mildly relativistic winds. Powerful winds can produce significant non-thermal radio emission via shocks. Among the nearby critical accretion quasars, PDS 456 has a very massive black hole (about one billion solar masses), shows a significant star-forming activity (about seventy solar masses per year) and hosts exceptionally energetic X-ray winds (power up to twenty per cent of the Eddington luminosity). To probe the radio activity in this extreme accretion and feedback system, we performed very-long-baseline interferometric (VLBI) observations of PDS 456 at 1.66 GHz with the European VLBI Network (EVN) and the enhanced Multi-Element Remotely Linked Interferometry Network (e-MERLIN). We find a rarely-seen complex radio-emitting nucleus consisting of a collimated jet and an extended non-thermal radio emission region. The diffuse emission region has a size of about 360 pc and a radio luminosity about three times higher than the nearby extreme starburst galaxy Arp 220. The powerful nuclear radio activity could result from either a relic jet with a peculiar geometry (nearly along the line of sight) or more likely from diffuse shocks formed naturally by the existing high-speed winds impacting on high-density star-forming regions., Comment: 8 pages, 3 figures, accepted for publication in MNRAS

We study the role of noise on the nature of the transition to collective motion in dry active matter. Starting from field theories that predict a continuous transition at the deterministic level, we show that fluctuations induce a density-dependent shift of the onset of order, which in turns changes the nature of the transition into a phase-separation scenario. Our results apply to a range of systems, including the topological models in which particles interact with a fixed number of nearest neighbors, which were believed so far to exhibit a continuous onset of order. Our analytical predictions are confirmed by numerical simulations of fluctuating hydrodynamics and microscopic models.

We present joint \textit{XMM-Newton} and \textit{NuSTAR} observations of the `bare' narrow line Seyfert 1 Ton S180 ($z=0.062$), carried out in 2016 and providing the first hard X-ray view of this luminous galaxy. We find that the 0.4--30 keV band cannot be self-consistently reproduced by relativistic reflection models, which fail to account simultaneously for the soft and hard X-ray emission. The smooth soft excess prefers extreme blurring parameters, confirmed by the nearly featureless nature of the RGS spectrum, while the moderately broad Fe K line and the modest hard excess above 10 keV appear to arise in a milder gravity regime. By allowing a different origin of the soft excess, the broadband X-ray spectrum and overall spectral energy distribution (SED) are well explained by a combination of: (a) direct thermal emission from the accretion disc, dominating from the optical to the far/extreme UV; (b) Comptonization of seed disc photons by a warm ($kT_{\rm e}\sim0.3$ keV) and optically thick ($\tau\sim10$) corona, mostly contributing to the soft X-rays; (c) Comptonization by a standard hot ($kT_{\rm e} \gtrsim 100$ keV) and optically thin ($\tau<0.5$) corona, responsible for the primary X-ray continuum; and (d) reflection from the mid/outer part of the disc. The two coronae are suggested to be rather compact, with $R_{\rm hot} \lesssim R_{\rm warm} \lesssim 10$ R$_{\rm g}$. Our SED analysis implies that Ton S180 accretes at super-Eddington rates. This is a key condition for the launch of a wind, marginal (i.e., 3.1$\sigma$ significance) evidence of which is indeed found in the RGS spectrum., Comment: 20 pages, 12 figures. Accepted for publication MNRAS

We revisit motility-induced phase separation in two models of active particles interacting by pairwise repulsion. We show that the resulting dense phase contains gas bubbles distributed algebraically up to a typically large cutoff scale. At large enough system size and/or global density, all the gas may be contained inside the bubbles, at which point the system is microphase-separated with a finite cut-off bubble scale. We observe that the ordering is anomalous, with different dynamics for the coarsening of the dense phase and of the gas bubbles. This phenomenology is reproduced by a "reduced bubble model" that implements the basic idea of reverse Ostwald ripening put forward in Tjhung et al. [Phys. Rev. X 8, 031080 (2018)].

We present $NuSTAR$ observations of four active galactic nuclei (AGN) located within 15 Mpc. These AGN, namely ESO 121-G6, NGC 660, NGC 3486 and NGC 5195, have observed X-ray luminosities of $L_{\rm 2-10\ keV, obs} \lesssim$ 10$^{39}$ erg s$^{-1}$, classifying them as low luminosity AGN (LLAGN). We perform broadband X-ray spectral analysis for the AGN by combining our $NuSTAR$ data with $Chandra$ or $XMM-Newton$ observations to directly measure their column densities ($N_{\rm H}$) and infer their intrinsic power. We complement our X-ray data with archival and new high angular resolution mid-infrared (mid-IR) data for all objects, except NGC 5195. Based on our X-ray spectral analysis, we found that both ESO 121-G6 and NGC 660 are heavily obscured ($N_{\rm H}$ > 10$^{23}$ cm$^{-2}$; $L_{\rm 2-10\ keV,\ int} \sim$ 10$^{41}$ erg s$^{-1}$), and NGC 660 may be Compton-thick. We also note that the X-ray flux and spectral slope for ESO 121-G6 have significantly changed over the last decade, indicating significant changes in the obscuration and potentially accretion rate. On the other hand, NGC 3486 and NGC 5195 appear to be unobscured and just mildly obscured, respectively, with $L_{\rm 2-10\ keV,\ int} <$ 10$^{39}$ erg s$^{-1}$; i.e., genuine LLAGN. Both of the heavily obscured AGN have $L_{\rm bol} >$ 10$^{41}$ erg s$^{-1}$ and $\lambda_{\rm Edd} \gtrsim$ 10$^{-3}$, and are detected in high angular resolution mid-IR imaging, indicating the presence of obscuring dust on nuclear scale. NGC 3486 however, is undetected in high-resolution mid-IR imaging, and the current data do not provide stringent constraints on the presence or absence of obscuring nuclear dust in the AGN., Comment: 18 pages, 11 figures, 3 tables, accepted for publication in MNRAS

We report on one of the brightest flux levels of the Seyfert 2 galaxy NGC 2992 ever observed in X-rays, on May 2019. The source has been monitored every few days from March 26, 2019 to December 14, 2019 by Swift-XRT, and simultaneous XMM-Newton (250 ks) and NuSTAR (120 ks) observations were triggered on May 6, 2019. The high count rate of the source (its 2-10 keV flux ranged between 0.7 and $1.0\times10^{-10}$ erg cm$^{-2}$ s$^{-1}$) allows us to perform a time-resolved spectroscopy, probing spatial scales of tens of gravitational radii from the central black hole. By constructing a map of the excess emission over the primary continuum, we find several emission structures in the 5.0-7.2 keV energy band. From fitting the 50 EPIC pn spectral slices of $\sim$5 ks duration, we interpret them as a constant narrow iron K$\alpha$ line and three variable components in the iron K complex. When a self-consistent model accounting for the accretion disk emission is considered (KYNrline), two of these features (in the 5.0-5.8 keV and 6.8-7.2 keV bands) can be ascribed to a flaring region of the accretion disk located at ${r_{in}}\simeq15$-40 r$_{g\rm }$ from the black hole. The third one (6.5-6.8 keV) is likely produced at much larger radii ($r_{in}>50$ r$_{g\rm }$). The inner radius and the azimuthal extension retrieved from the coadded spectra of the flaring states are ${ r_{in}}=15\pm3$ r$_{g\rm }$ and $\phi=165^{\circ}-330^{\circ}$, suggesting that the emitting region responsible for the broad iron K component is a relatively compact annular sector within the disk. Our findings support a physical scenario in which the accretion disk in NGC 2992 becomes more active at high accretion rates ($L_{\rm bol}/L_{\rm Edd}\geq4\%$)., Comment: 12 pages, 12 figures, 1 table. Accepted for publication in MNRAS

Biologically-informed neural networks (BINNs), an extension of physics-informed neural networks [1], are introduced and used to discover the underlying dynamics of biological systems from sparse experimental data. In the present work, BINNs are trained in a supervised learning framework to approximate in vitro cell biology assay experiments while respecting a generalized form of the governing reaction-diffusion partial differential equation (PDE). By allowing the diffusion and reaction terms to be multilayer perceptrons (MLPs), the nonlinear forms of these terms can be learned while simultaneously converging to the solution of the governing PDE. Further, the trained MLPs are used to guide the selection of biologically interpretable mechanistic forms of the PDE terms which provides new insights into the biological and physical mechanisms that govern the dynamics of the observed system. The method is evaluated on sparse real-world data from wound healing assays with varying initial cell densities [2].

Equation learning methods present a promising tool to aid scientists in the modeling process for biological data. Previous equation learning studies have demonstrated that these methods can infer models from rich datasets, however, the performance of these methods in the presence of common challenges from biological data has not been thoroughly explored. We present an equation learning methodology comprised of data denoising, equation learning, model selection and post-processing steps that infers a dynamical systems model from noisy spatiotemporal data. The performance of this methodology is thoroughly investigated in the face of several common challenges presented by biological data, namely, sparse data sampling, large noise levels, and heterogeneity between datasets. We find that this methodology can accurately infer the correct underlying equation and predict unobserved system dynamics from a small number of time samples when the data is sampled over a time interval exhibiting both linear and nonlinear dynamics. Our findings suggest that equation learning methods can be used for model discovery and selection in many areas of biology when an informative dataset is used. We focus on glioblastoma multiforme modeling as a case study in this work to highlight how these results are informative for data-driven modeling-based tumor invasion predictions.

We systematically study the holographic phase transition of the radion field in a five-dimensional warped model which includes a scalar potential with a power-like behavior. We consider Kaluza-Klein (KK) resonances with masses $m_{\rm KK}$ at the TeV scale or beyond. The backreaction of the radion field on the gravitational metric is taken into account by using the superpotential formalism. The confinement/deconfinement first order phase transition leads to a gravitational wave stochastic background which mainly depends on the scale $m_{\rm KK}$ and the number of colors, $N$, in the dual theory. Its power spectrum peaks at a frequency that depends on the amount of tuning required in the electroweak sector. It turns out that the present and forthcoming gravitational wave observatories can probe scenarios where the KK resonances are very heavy. Current aLIGO data already rule out vector boson KK resonances with masses in the interval $m_{\rm KK}\sim(1 - 10) \times 10^5$ TeV. Future gravitational experiments will be sensitive to resonances with masses $m_{\rm KK}\lesssim 10^5$ TeV (LISA), $10^8$ TeV (aLIGO Design) and $10^9$ TeV (ET). Finally, we also find that the Big Bang Nucleosynthesis bound in the frequency spectrum turns into a lower bound for the nucleation temperature as $T_n \gtrsim 10^{-4}\sqrt{N} \,m_{\rm KK}$., Comment: 11 pages, 6 figures; v2 extended version: added references and Figs. 1 (right), 2, 3, 4 (lower panels) and 5, Sec. IV, and extended discussion in Secs. V, VI and VII; v3 added references, extended discussion in Sec. VI. It matches the version published in Physical Review D

Search engine ranking pipelines are commonly based on large ensembles of machine-learned decision trees. The tight constraints on query response time recently motivated researchers to investigate algorithms to make faster the traversal of the additive ensemble or to early terminate the evaluation of documents that are unlikely to be ranked among the top-k. In this paper, we investigate the novel problem of \textit{query-level early exiting}, aimed at deciding the profitability of early stopping the traversal of the ranking ensemble for all the candidate documents to be scored for a query, by simply returning a ranking based on the additive scores computed by a limited portion of the ensemble. Besides the obvious advantage on query latency and throughput, we address the possible positive impact of query-level early exiting on ranking effectiveness. To this end, we study the actual contribution of incremental portions of the tree ensemble to the ranking of the top-k documents scored for a given query. Our main finding is that queries exhibit different behaviors as scores are accumulated during the traversal of the ensemble and that query-level early stopping can remarkably improve ranking quality. We present a reproducible and comprehensive experimental evaluation, conducted on two public datasets, showing that query-level early exiting achieves an overall gain of up to 7.5% in terms of NDCG@10 with a speedup of the scoring process of up to 2.2x., Comment: Accepted at SIGIR 2020 (short paper)

In precision-oriented tasks like answer ranking, it is more important to rank many relevant answers highly than to retrieve all relevant answers. It follows that a good ranking strategy would be to learn how to identify the easiest correct answers first (i.e., assign a high ranking score to answers that have characteristics that usually indicate relevance, and a low ranking score to those with characteristics that do not), before incorporating more complex logic to handle difficult cases (e.g., semantic matching or reasoning). In this work, we apply this idea to the training of neural answer rankers using curriculum learning. We propose several heuristics to estimate the difficulty of a given training sample. We show that the proposed heuristics can be used to build a training curriculum that down-weights difficult samples early in the training process. As the training process progresses, our approach gradually shifts to weighting all samples equally, regardless of difficulty. We present a comprehensive evaluation of our proposed idea on three answer ranking datasets. Results show that our approach leads to superior performance of two leading neural ranking architectures, namely BERT and ConvKNRM, using both pointwise and pairwise losses. When applied to a BERT-based ranker, our method yields up to a 4% improvement in MRR and a 9% improvement in P@1 (compared to the model trained without a curriculum). This results in models that can achieve comparable performance to more expensive state-of-the-art techniques., Comment: Accepted at SIGIR 2020 (long)

The identification of relevance with little textual context is a primary challenge in passage retrieval. We address this problem with a representation-based ranking approach that: (1) explicitly models the importance of each term using a contextualized language model; (2) performs passage expansion by propagating the importance to similar terms; and (3) grounds the representations in the lexicon, making them interpretable. Passage representations can be pre-computed at index time to reduce query-time latency. We call our approach EPIC (Expansion via Prediction of Importance with Contextualization). We show that EPIC significantly outperforms prior importance-modeling and document expansion approaches. We also observe that the performance is additive with the current leading first-stage retrieval methods, further narrowing the gap between inexpensive and cost-prohibitive passage ranking approaches. Specifically, EPIC achieves a MRR@10 of 0.304 on the MS-MARCO passage ranking dataset with 78ms average query latency on commodity hardware. We also find that the latency is further reduced to 68ms by pruning document representations, with virtually no difference in effectiveness., Comment: Accepted at SIGIR 2020 (short)

In a conversational context, a user expresses her multi-faceted information need as a sequence of natural-language questions, i.e., utterances. Starting from a given topic, the conversation evolves through user utterances and system replies. The retrieval of documents relevant to a given utterance in a conversation is challenging due to ambiguity of natural language and to the difficulty of detecting possible topic shifts and semantic relationships among utterances. We adopt the 2019 TREC Conversational Assistant Track (CAsT) framework to experiment with a modular architecture performing: (i) topic-aware utterance rewriting, (ii) retrieval of candidate passages for the rewritten utterances, and (iii) neural-based re-ranking of candidate passages. We present a comprehensive experimental evaluation of the architecture assessed in terms of traditional IR metrics at small cutoffs. Experimental results show the effectiveness of our techniques that achieve an improvement up to 0.28 (+93%) for P@1 and 0.19 (+89.9%) for nDCG@3 w.r.t. the CAsT baseline., Comment: 5 pages

Deep pretrained transformer networks are effective at various ranking tasks, such as question answering and ad-hoc document ranking. However, their computational expenses deem them cost-prohibitive in practice. Our proposed approach, called PreTTR (Precomputing Transformer Term Representations), considerably reduces the query-time latency of deep transformer networks (up to a 42x speedup on web document ranking) making these networks more practical to use in a real-time ranking scenario. Specifically, we precompute part of the document term representations at indexing time (without a query), and merge them with the query representation at query time to compute the final ranking score. Due to the large size of the token representations, we also propose an effective approach to reduce the storage requirement by training a compression layer to match attention scores. Our compression technique reduces the storage required up to 95% and it can be applied without a substantial degradation in ranking performance., Comment: Accepted at SIGIR 2020 (long)

Past X-ray observations of the nearby luminous quasar PDS 456 (at $z=0.184$) have revealed a wide angle accretion disk wind (Nardini et al. 2015), with an outflow velocity of $\sim-0.25c$, as observed through observations of its blue-shifted iron K-shell absorption line profile. Here we present three new XMM-Newton observations of PDS 456; one in September 2018 where the quasar was bright and featureless, and two in September 2019, 22 days apart, occurring when the quasar was five times fainter and where strong blue-shifted lines from the wind were present. During the second September 2019 observation, three broad ($\sigma=3000$ km s$^{-1}$) absorption lines were resolved in the high resolution RGS spectrum, which are identified with blue-shifted OVIII Ly$\alpha$, NeIX He$\alpha$ and NeX Ly$\alpha$. The outflow velocity of this soft X-ray absorber was found to be $v/c=-0.258\pm0.003$, fully consistent with iron K absorber with $v/c=-0.261\pm0.007$. The ionization parameter and column density of the soft X-ray component ($\log\xi=3.4$, $N_{\rm H}=2\times10^{21}$ cm$^{-2}$) outflow was lower by about two orders of magnitude, when compared to the high ionization wind at iron K ($\log\xi=5$, $N_{\rm H}=7\times10^{23}$ cm$^{-2}$). Substantial variability was seen in the soft X-ray absorber between the 2019 observations, declining from $N_{\rm H}=10^{23}$ cm$^{-2}$ to $N_{\rm H}=10^{21}$ cm$^{-2}$ over 20 days, while the iron K component was remarkably stable. We conclude that the soft X-ray wind may originate from an inhomogeneous wind streamline passing across the line of sight and which due to its lower ionization, is located further from the black hole, on parsec scales, than the innermost disk wind., Comment: 13 pages, accepted for publication in the Astrophysical Journal