Newsletter English December/January 2022 Print E-mail


Logo ICRANet Logo ICRA Logo MAE

ICRANet Newsletter



ICRANet Newsletter
December/January 2022





1. Deep Learning in Searching the Spectroscopic Redshift of Quasars

The new work co-authored by ICRANet scientists is published by MNRAS on January 19, 2022


Fig. 1. The chosen architecture of 1-dimensional of FNet to learn higher-order features hidden in the input flux. It slides the flux via convolutional layers of kernel size = 500, 200 and 15, respectively to search for the "global" and "local" patterns in the flux of quasars. The fully connected layers output the redshift. Left: The structure of a residual block, the input x goes through two convolutional layers as H(x) then add itself as H(x)+x, batch normalization is applied after each convolutional layer, and the activation function ReLU acts on the first batch normalization layer. Right: The entire structure: The flux goes through 24 residual blocks, the first 21 blocks have channel size 32, followed by three blocks of channel size 64, 32 and 16 respectively. The output of blocks then is flattened and passes three fully convolutional layers and eventually outputs the redshift. The rectified linear unit (ReLU) is applied after each fully connected layer.

Quasi-stellar radio sources (Quasars) are high-luminosity active galactic nuclei (AGN). There is a broad consensus that quasars are powered by a gaseous accretion disk around a supermassive black hole (SMBH) of ~106 - 109 solar mass. Because of their high luminosity, quasars have been observed up to redshifts z~8 when the universe formed its first structures, i.e., in the reionization epoch. It makes quasars a powerful tool to study the cosmic history and structure formation in the early universe and, specifically, probe the physics governing the SMBHs and their surrounding accretion disk, as the rapid growth of SMBHs occurs at redshifts z = 5-10. Regarding their existence in a wide range of redshifts, quasars, as the cosmology’s standard candle, can also be utilized to put more constraints on the cosmological parameters.
On the one hand, the massive volume of data of astrophysical surveys makes the procedure for visual inspection of each spectrum to classify and determine the redshift highly time-consuming. For example, the Sloan Digital Sky Survey IV (SDSS-IV) quasar catalog from Data Release 16 (DR16) of the extended Baryon Oscillation Spectroscopic Survey (eBOSS), includes a "superset" of objects labelled as quasars containing 1440615 spectra.
On the other hand, the adapted automatic methods, that operate based on comparing each spectrum with a dataset of spectra, usually perform worse than visual inspection methods in classification tasks. Thus, implementing automated strategies with human–expert precision is of great importance.


Fig. 2. The accuracy of prediction vs. Δ𝜈. The red solid line and black dashed line represent the accuracy of the FNet and QuasarNE, respectively, for DR16Q sample. The blue solid line and grey dashed line represent the accuracy of the FNet and QuasarNET, respectively, for DR16Q sample when DR12Q is excluded. The green dashed lines shows the accuracy of FNet for 5,190 visually inspected sources in DR16Q when QuasarNET fails to estimate.

Over the last few years, machine learning (ML) and deep learning (DL) algorithms have become increasingly popular in astronomy and astrophysics. They are using various recognition patterns and are able to identify the spectral features of astrophysical objects such as emission/absorption lines, spectral breaks, and flux correlations, and perform the classification and redshift determination as accurate as the visual inspection.
With the goal to estimate the redshift of quasars in Sloan Digital Sky Survey IV (SDSS-IV) catalog from DR16 quasar-only (DR16Q) of eBOSS on a broad range of signal-to-noise ratios, the group of researchers which includes ICRANet scientists has developed a new tool, the FNet. FNet is a 1–dimensional convolutional neural network (CNN) with a residual neural network (ResNet) structure. This network has 24 convolutional layers and the ResNet structure with different kernel sizes of 500, 200, and 15, which makes it to discover the "local" and "global" patterns, as well as correlations of fluxes in a different wavelength, in the whole sample of spectra by a self-learning procedure; see Fig. 1. It reaches the accuracy of 97.0 % for the velocity difference for redshift, | Δ𝜈 |< 6000 km/s and 98.0% for | Δ𝜈 | <12000 km/s. Here Δ𝜈 = c (Z-Z_VI)/(1+ZVI) is the velocity difference from redshift, c is the speed of light, Z_VI is the redshift from the visually inspected quasar sample, and Z is the predicted redshift.
While QuasarNET, which is a standard CNN adopted in the SDSS routine and is constructed by 4 convolutional layers (no ResNet structure), with kernel sizes of 10, to measure the redshift via identifying seven emission lines (local patterns), fails in estimating redshift of ∼ 1.3% of visually inspected quasars in DR16Q catalog, and it gives 97.8% for | Δ𝜈 |< 6000 km/s and 97.9 % for | Δ𝜈 |< 12000 km/s; see Fig 2. FNet is applicable for a wider range of SDSS spectra, especially for those missing the clear emission lines, which is necessary for other standard methods to work. These properties of FNet, together with the fast predictive power of machine learning, allow it to be a more accurate alternative for the pipeline redshift estimator and can make it practical in the upcoming catalogs to reduce the number of spectra to inspect visually.
The paper by F. Rastegar Nia, M. T. Mirtorabi, R. Moradi, A. Vafaei. Sadr, Y. Wang «Deep learning in searching the spectroscopic redshift of quasars» is published in Monthly Notices of the Royal Astronomical Society on 19 January 2022.
Link: https://doi.org/10.1093/mnras/stac076



2. LeCosPA 4th International Symposium, online, November 29 – December 3, 2021

The "LeCosPA 4th International Symposium, Unity of Physics – From Plasma Wakefields to Black Holes" has been held at the National Taiwan University (Taiwan) and online from November 29 to December 3, 2021. During the meeting, a lot of interesting topics have been discussed, ranging from Extreme Light, High Field Science, and Plasma Wakefields to Early and Late Universe, Particle Astrophysics, Black Holes, Gravitational Waves and Testing of Modified Gravity, Information Loss Paradox and Analog Black Holes. On December 2, 2021 Prof. Ruffini, Director of ICRANet, presented a lecture titled "What is the role of the rotational energy extraction from Black Holes", here is the abstract:
We have demonstrated that the inner engine of GRBs and AGN produces high-energy emission by synchrotron radiation of electrons/protons that are accelerated in the rotating BH vicinity. The angular momentum of the Kerr BH and the surrounding magnetic field determine the energetics and characteristic radiation frequency, and their relative direction determines whether the motion of the electrons around the magnetic field lines follows a right-handed or a left-handed helix, and likewise the angular momentum inherited by the radiation.
Link to Prof. Ruffini presentation on YouTube:
https://www.youtube.com/watch?v=NaUJ7NqW3LQ


Fig. 3: Prof. Ruffini delivering his lecture on the occasion of the LeCosPA 4th International Symposium, Unity of Physics, on December 2, 2021.



3. Antarctic eclipse and meridian transits, online meeting, December 4, 2021

The meeting "Antarctic eclipse and meridian transits. Experiments of Celestial mechanics and Astrophysics" has been held virtually on December 4, 2021. Prof. Costantino Sigismondi, ICRANet collaborator and chair of the event, thanks also to the support of ICRANet and many other scientists from all over the world, organized this virtual meeting as well as a podcast meeting in order to create a nice occasion for discussion among students and researchers.
The total eclipse of the Sun gave the occasion to reconsider the solar diameter as well as its measures by comparison with the ephemerides with a standard diameter. The virtual meeting started at 10 AM on Saturday December 4, with the opening remarks made by Prof. Sigismondi and went on with some important contributions by other relevant scientists on the field on "The consultation of ephemerides", "Eclipse and timing: the measurement uncertainty and its role for the eclipse", "Meridian transits and timing", "The first results from the obelisk – meridian of Saint Peter on the transit and the solar diameter", "The entry of the Sun in Scorpio and in Sagittarius 2021 at Saint Peter, with the Sun and the stars (preliminary results and calibration of the instrument)" as well on "The eclipse in Antarctica and the double eclipse of the year 810 in the letter by Dungal to Charlemagne".

Fig. 4: The Sun projected to the meridian of Saint Peter on November 24, 2021. Fig. 5: Engraving of the Cesarean (1521) of Vitruvius, De Architectura, Book IX - chapter VII, with the sentence on the legend of Gerbertus and the reassure of Augustus Fig. 6: the Sun at the meridian of Saint Peter, just before the meridian eclipse on November 26, 2021.

This theoretical section was also integrated with the podcast materials prepared by Prof. Sigismondi. The program of the event and all the relevant podcast materials, can be found at the following link:
http://www.icranet.org/index.php?option=com_content&task=view&id=1399
A recording of the meeting can also be found on ICRANet YouTube channel: https://www.youtube.com/watch?v=fLfBehXC-H4&t=696s



4. Amati Fest meeting, December 6-7, 2021

On the occasion of the celebration of the 50th anniversary of "Introducing the Black Hole" by Prof. Remo Ruffini and Prof. John A. Wheeler, ICRANet organized an online meeting, in order to better discuss about the most recent results on the understanding of Gamma-Ray Bursts (GRBs) and their "inner-engines". In this framework, a particular attention has been given to the understanding of the Amati relation.


Fig. 7: The afterglow of GRB 140114C as observed from Radio to VHE bands.

This meeting took place at ICRANet Hq in Pescara (Italy) and online from December 6 to 7, 2021 and it has also been broadcasted worldwide on ICRANet YouTube channel. The meeting started with the opening remarks made by Prof. Remo Ruffini, Director of ICRANet, and by Prof. Lorenzo Amati. During the event, several invited speaker took the floor in order to present their recent results, namely Prof. Marco Tavani (President of INAF), Prof. Narek Sahakyan (Director of ICRANet Seat in Armenia), Prof. Michael Kramer (Director - Max-Planck-Institut für Radioastronomie), Prof. Jorge Armando Rueda Hernandez (ICRANet, University of Ferrara), Prof. Carlo Luciano Bianco (ICRA, ICRANet), Prof. Gregory Vereshchagin (ICRANet), Prof. She Sheng Xue (ICRANet), Prof. Carlos Raul Arguelles (ICRANet, CONICET, Universidad Nacional de La Plata), Prof. Soroush Shakeri (ICRANet, Isfahan University of Technology), Prof. Piero Rosati (University of Ferrara), Prof. Razmik Mirzoyan (Max-Planck-Institute for Physics), Prof. Cristiano Guidorzi (University of Ferrara), Prof. Massimo Della Valle (ICRANet, INAF Osservatorio astronomico di Capodimonte), Dr Luca Izzo (Osservatorio astronomico di Capodimonte), Prof. Yifu Cai (University of Sciences and Technology of China), Prof. Yefei Yuan (University of Sciences and Technology of China), Prof. Mimoza Hafizi (University of Tirana), Prof. Claus Lämmerzahl (ZARM University of Bremen), Prof. Stefano Scopel (CQUeST, Sogang University), Prof. Simonetta Filippi (ICRA, University Campus Bio-medico of Rome), Prof. Christian Cherubini (ICRA, University Campus Bio-medico of Rome), Prof. Stefano Ansoldi (University of Udine), Prof. Aldo Treves (University of Insubria), Prof. Francesco Haardt (University of Insubria), Dr Ana Penacchioni (CONICET, Universidad Nacional de La Plata), Dr Laura Marcela Becerra Bayona (ICRANet, Universidad Católica de Chile), Prof. Wang Yu (ICRANet), Prof. Liang Li (ICRANet), Prof. Rahim Moradi (ICRANet), Dr Yerlan Aimuratov (ICRANet, Fesenkov Astrophysical Institute), Dr Yunlong Zheng (ICRANet, University of Sciences and Technology of China), Eduar Antonio Becerra Vergara (ICRANet), Dr Fatemeh Rastegar Nia (ICRANet, Alzahra University) and Dr Sareh Eslamzadeh Askestani (University of Mazandaran).
After those intervention, there has been a long discussion on the data and results presented among all the scientists, who had the opportunity to have fruitful exchanges on these important topics. The meeting ended on Tuesday, December 7 with the concluding remarks made by prof. Remo Ruffini and Prof. Lorenzo Amati.

Fig. 8: Prof. Lorenzo Amati and Prof. Remo Ruffini during the Amati Fest meeting. Fig. 9: From the left to the right: Prof. Luca Izzo, Prof. Jorge Rueda, Prof. Lorenzo Amati, Prof. Remo Ruffini, Prof. Rahim Moradi, Prof. Liang Lia and Prof Wang Yu, discussing at ICRANet Hq on the occasion of the Amati Fest meeting.



5. Scientific visits to ICRANet 

  • Prof. Lorenzo Amati
    INAF - Osservatorio di Astrofisica e Scienza dello Spazio
    December 6-7, 2021

  • Prof. Jorge Armando Rueda Hernandez
    ICRA, ICRANet and University of Ferrara
    December 6-7, 2021

  • Prof. Luca Izzo
    Observatory of Capodimonte – Italy
    December 6-7, 2021

Prof. Lorenzo Amati Prof. Jorge Armando Rueda Hernandez Prof. Luca Izzo

During their visit, those scientists had an opportunity to discuss their scientific research and to have fruitful exchange of ideas with other researchers from ICRANet and from different parts of the world. They also took part in the Amati fest meeting, which was held at ICRANet Hq in Pescara (Italy) and online from December 6 to 7, 2021.



6. Recent publications

J. A. Rueda and R. Ruffini, The quantum emission of an alive black hole, published in International Journal of Modern Physics D VOL. 30, NO. 14.
A long march of 50 years of successive theoretical progress and new physics discovered using observations of gamma-ray bursts has finally led to the formulation of an efficient mechanism able to extract the rotational energy of a Kerr black hole to power these most energetic astrophysical sources and active galactic nuclei. We here present the salient features of this long-sought mechanism, based on gravito-electrodynamics, and which represents an authentic shift of paradigm of black holes as forever "alive" astrophysical objects.
This essay is awarded third prize in the 2021 Essay Competition of the Gravity Research Foundation.
DOI: https://doi.org/10.1142/S0218271821410030


F. Rastegar Nia, M. T. Mirtorabi, R. Moradi, A. Vafaei Sadr, Y. Wang, Deep learning in searching the spectroscopic redshift of quasars, published in Monthly Notices of the Royal Astronomical Society on January 19, 2022.
Studying the cosmological sources at their cosmological rest-frames is crucial to track the cosmic history and properties of compact objects. In view of the increasing data volume of existing and upcoming telescopes/detectors, we here construct a 1–dimensional convolutional neural network (CNN) with a residual neural network (ResNet) structure to estimate the redshift of quasars in Sloan Digital Sky Survey IV (SDSS-IV) catalog from DR16 quasar-only (DR16Q) of eBOSS on a broad range of signal-to-noise ratios, named FNet. Owing to its 24 convolutional layers and the ResNet structure with different kernel sizes of 500, 200 and 15, FNet is able to discover the ‘local’ and ‘global’ patterns in the whole sample of spectra by a self-learning procedure. It reaches the accuracy of 97.0 per cent for the velocity difference for redshift, |Δν|<6000 km s−1 and 98.0 per cent for |Δν|<12000 km s−1. While QuasarNET, which is a standard CNN adopted in the SDSS routine and is constructed by 4 convolutional layers (no ResNet structure), with kernel sizes of 10, to measure the redshift via identifying seven emission lines (local patterns), fails in estimating redshift of ∼1.3 per cent of visually inspected quasars in DR16Q catalog, and it gives 97.8 per cent for |Δν|<6000 km s−1 and 97.9 per cent for |Δν|<12000 km s−1. Hence, FNet provides similar accuracy to QuasarNET, but it is applicable for a wider range of SDSS spectra, especially for those missing the clear emission lines exploited by QuasarNET. These properties of FNet, together with the fast predictive power of machine learning, allow FNet to be a more accurate alternative for the pipeline redshift estimator and can make it practical in the upcoming catalogs to reduce the number of spectra to visually inspect.
LINK: https://doi.org/10.1093/mnras/stac076


Davood Rafiei Karkevandi, Soroush Shakeri, Violetta Sagun, and Oleksii Ivanytskyi, Bosonic dark matter in neutron stars and its effect on gravitational wave signal, published in Phys. Rev. D 105, 023001 on January 3, 2022.
We study an impact of self-interacting bosonic dark matter (DM) on various observable properties of neutron stars (NSs). The analysis is performed for asymmetric DM with masses from few MeV to GeV, the self-coupling constant of order and various DM fractions. Allowing a mixture between DM and baryonic matter, the formation of a dense DM core or an extended dark halo has been explored. We find that both distribution regimes crucially depend on the mass and fraction of DM for sub-GeV boson masses in the strong coupling regime. From the combined analysis of the mass-radius relation and the tidal deformability of compact stars including bosonic DM, we set a stringent constraint on DM fraction. We conclude that observations of 2 M⊙ NSs together with Λ 1.4 ≤ 580 constraint, set by LIGO/Virgo Collaboration, favor sub-GeV DM particles with low fractions below ∼5%.
LINK: https://doi.org/10.1103/PhysRevD.105.023001


Sareh Eslamzadeh, Javad T. Firouzjaee, Kourosh Nozari, Radiation from Einstein-Gauss-Bonnet de Sitter black hole via tunneling process, accepted for publication on EPJC on January 4, 2022.
In this paper, we probe in novel 4D Einstein-Gauss-Bonnet (EGB) black hole and its thermodynamics. We illustrate the three asymptotically 4D EGB spacetime as an asymptotically flat, de Sitter, and Anti-de Sitter. Also, we apply the tunneling of the massless particles from the event horizon of 4D EGB gravity and we investigate the correlation between the emission modes and temperature of the event horizon. In asymptotically flat spacetime, the existence of the coupling constant alone constructs the regular spacetime, the radiation deviates from the pure thermal, and the temperature of the black hole horizon would be finite in the final stage of the black hole evaporation. If we consider the 4D ds-EGB structure, then we will have three horizons in the specific mass range of the black hole. By carefully examining the temperature of the black hole and cosmological horizons with the tunneling of the massless particles from these horizons. As a result, the evolution of these temperatures is in direction of the remaining rest mass with the probably same temperature for black hole and cosmological horizon. In addition, the exciting result is that temperature behaviors exactly match with the temperature behaviors of a regular black hole in Lovelock gravity in a higher dimension. This confirms the bypassing of EGB in four dimensions of spacetime from the Lovelock gravity in higher dimensions.
LINK: https://doi.org/10.1140/epjc/s10052-022-09992-6


Yunis, Rafael; Argüelles, Carlos R.; Scóccola, Claudia G.; López Nacir, Diana; Giordano, Gastón, Self-Interacting Dark Matter in Cosmology: accurate numerical implementation and observational constraints, accepted for publication on JCAP (24 January 2022).
This paper presents a systematic and accurate treatment of the evolution of cosmological perturbations in self-interacting dark matter models, for particles which decoupled from the primordial plasma while relativistic. We provide a numerical implementation of the Boltzmann hierarchies developed in a previous paper [JCAP, 09 (2020) 041] in a publicly available Boltzmann code and show how it can be applied to realistic DM candidates such as sterile neutrinos either under resonant or non-resonant production mechanisms, and for different field mediators. At difference with traditional fluid approximations - also known as a c_eff − c_vis parametrizations- our approach follows the evolution of phase-space perturbations under elastic DM interactions for a wide range of interaction models, including the effects of late kinetic decoupling. Finally, we analyze the imprints left by different self interacting models on linear structure formation, which can be constrained using Lyman-α forest and satellite counts. We find new lower bounds on the particle mass that are less restrictive than previous constraints.
ArXiv: https://arxiv.org/abs/2108.02657


MAGIC collaboration, Search for Very High-energy Emission from the Millisecond Pulsar PSR J0218+4232, published in ApJ, Volume 922, Number 2 on December 3, 2021.
PSR J0218+4232 is one of the most energetic millisecond pulsars known and has long been considered as one of the best candidates for very high-energy (VHE; >100 GeV) γ-ray emission. Using 11.5 yr of Fermi Large Area Telescope (LAT) data between 100 MeV and 870 GeV, and ~90 hr of Major Atmospheric Gamma Imaging Cherenkov (MAGIC) observations in the 20 GeV to 20 TeV range, we searched for the highest energy γ-ray emission from PSR J0218+4232. Based on the analysis of the LAT data, we find evidence for pulsed emission above 25 GeV, but see no evidence for emission above 100 GeV (VHE) with MAGIC. We present the results of searches for γ-ray emission, along with theoretical modeling, to interpret the lack of VHE emission. We conclude that, based on the experimental observations and theoretical modeling, it will remain extremely challenging to detect VHE emission from PSR J0218+4232 with the current generation of Imaging Atmospheric Cherenkov Telescopes, and maybe even with future ones, such as the Cherenkov Telescope Array.
LINK: https://doi.org/10.3847/1538-4357/ac20d7


MAGIC collaboration, Observation of the Gamma-Ray Binary HESS J0632+057 with the H.E.S.S., MAGIC, and VERITAS Telescopes, published in ApJ, Volume 923, Number 2 on December 24, 2021.
The results of gamma-ray observations of the binary system HESS J0632 + 057 collected during 450 hr over 15 yr, between 2004 and 2019, are presented. Data taken with the atmospheric Cherenkov telescopes H.E.S.S., MAGIC, and VERITAS at energies above 350 GeV were used together with observations at X-ray energies obtained with Swift-XRT, Chandra, XMM-Newton, NuSTAR, and Suzaku. Some of these observations were accompanied by measurements of the Hα emission line. A significant detection of the modulation of the very high-energy gamma-ray fluxes with a period of 316.7 ± 4.4 days is reported, consistent with the period of 317.3 ± 0.7 days obtained with a refined analysis of X-ray data. The analysis of data from four orbital cycles with dense observational coverage reveals short-timescale variability, with flux-decay timescales of less than 20 days at very high energies. Flux variations observed over a timescale of several years indicate orbit-to-orbit variability. The analysis confirms the previously reported correlation of X-ray and gamma-ray emission from the system at very high significance, but cannot find any correlation of optical Hα parameters with fluxes at X-ray or gamma-ray energies in simultaneous observations. The key finding is that the emission of HESS J0632 + 057 in the X-ray and gamma-ray energy bands is highly variable on different timescales. The ratio of gamma-ray to X-ray flux shows the equality or even dominance of the gamma-ray energy range. This wealth of new data is interpreted taking into account the insufficient knowledge of the ephemeris of the system, and discussed in the context of results reported on other gamma-ray binary systems.
LINK: https://doi.org/10.3847/1538-4357/ac29b7
 
Organization
ICRANet Seats
Director and Staff
Personal Pages
Official Documents
Scientific Agreements
Annual reports
Meetings
Upcoming Meetings
Marcel Grossmann
Galileo - Xu Guangqi
Italian-Korean
C. Lattes Meeting
Bego Scientific Rencontre
Zeldovich Meetings
Meetings in Armenia
Past meeting series
Single meetings
ICRANet Workshops
Other Meetings
IRAP Ph.D. Schools
Weekly Seminars
Research
Research Groups
IRAP Ph.D.
Objectives
Consortium
Faculty
Courses
Students
Schools
Thesis
Publications
Books
Articles
Proceedings
Outreach
Newsletter
Press releases
Recorded talks
Public Events
Visitors: 163579595