The Virgo Collaboration

Mario Martínez

The detection of Gravitational Waves (GWs) from a black hole (BH) binary merger by LIGO in 2015 is a milestone that represents the beginning of a new era in the exploration of the universe. Shortly after the addition of the VIRGO antenna into the network lead to the detection of the neutron star (NS) binary merger that could be followed in electromagnetic signals, and thus represents the beginning of multi-messenger astronomy. These events have radically changed the stage for several areas of physics, from astrophysics to particle physics.


During 2020, the IFAE group in Virgo has made great progress in the construction of the new instrumented baffle for the upgrade of the experiment and remained deeply involved in the analysis of the LIGO/Virgo data. In addition, the group played a central role in promoting the new Einstein Telescope project.

New Instrumentation for Virgo being integrated at EGO

The new baffle for Virgo is instrumented with 76 novel Si-based photosensors developed by Hamamatsu, with modifications in the packaging that render them compatible with ultra-high vacuum conditions. The sensors are mounted on two large PCBs, which are not exposed to the light and include eight temperature sensors distributed across their surface. The infrared light from the IMC cavity penetrates the holed mirror-polished stainless steel baffle reaching the photo-sensors behind them. The baffle, as a whole, is designed to preserve as much as possible the optical properties of the existing one in terms of reflectivity and total scattering to maintain the performance of the optical cavity. In particular, both baffle and sensor surfaces are anti-reflecting coated and all the components have been validated for ultra-high vacuum compatibility. Furthermore, the inner edge of the baffle and the edges of the holes are produced with a small radius of curvature to prevent scattering. Finally, the weight and centre-of-gravity of the new baffle will be very similar to those of the current one, to facilitate the installation in the existing suspension system.
Figure 1: Pictures from the integration test of the new instrumented baffle in the EGO clean room. A laser was used to demonstrate a perfect weight balance and alignment between payload and baffle after the whole system is suspended by the two hires. A dummy mirror made of metal with identical dimensions and weight as the real one was employed in the test.

2020 activities

In 2020, the IFAE group in Virgo carried out a frenetic activity in different fronts related to the construction of the new detector, the understanding of the performance of the interferometer in the O3 observation run, and the physics analysis of the data. In addition, the group increased their international visibility inside and outside the experiment, playing also a central role in promoting the Einstein Telescope project. In the following, the activities carried out by the group are described, separated in the different aspects.

Contributions to the operations of Virgo

In 2020, the IFAE team maintained a deep involvement in the activities related to the understanding of the performance of the interferometer. The work led to a short author list publication: I. Fiori et al., “The hunt for environmental noise in Virgo during the third observing run”, Galaxies 2020, 8(4), 82;

Contributions to Governance

During 2020, IFAE maintained its responsibilities in the control of the stray light in the interferometer, with Ll. M. Mir acting as coordinator of the SLC working group in Virgo. M. Martínez was member of the Virgo Steering Committee, and was appointed member of the Virgo Organization Committee (VOC), with the charge of preparing new bylaws in the scenario of a fast expansion of the collaboration, thus bringing to the discussion the experience from the organization of very large HEP collaborations. Finally, M. Martinez maintained its involvement in the Einstein Telescope, as member of the Steering Committee.

Contributions to the upgrade for Advanced Virgo (AdV+ phase I)

During 2020, the IFAE team carried out an intense activity with the mission to construct a first instrumented baffle for AdV+. It would be installed in the IMC suspended mirror acting as a demonstrator of the technology and providing unique monitoring of the stray light in the IMC cavity. The basic requirements of the instrumented baffles are rather challenging. The total reflectivity of the surface at 1064 nm should be maintained at the level of 0.5% and the total scattering (TIS) within 300 - 500 ppm. As suspended by a couple of wires, no active cooling is possible and the readout cables should be minimized to avoid affecting the performance of the payload and seismic attenuation systems. Most importantly, the ultra-high vacuum (UHV) conditions in the towers (with a vacuum level of $10{^-9}$ mbars) impose stringent constraints to all the materials to avoid contaminants affecting the mirrors, and force a rather novel PCB design with low power consumption and efficient heat dissipation. In addition, the baffle DAQ system should not interfere with the rest of the electronics in the tower, governing the feedback loop systems that control the position of the mirror, leading to the exploration of wireless read-out solutions. Finally, the total weight of the new device should meet within 100 g the weight of the non-instrumented version to facilitate the integration in the suspension system.

In October 2020, a first integration of the instrumented baffle in the new IMC payload (already mirror polished but before applying final AR-coating) was carried out successfully in an EGO clean room (see Figure 1). The final integration of the baffle in the IMC tower is scheduled for Spring 2021 after an UHV recertification at EGO takes place.

Figure 1: Pictures from the integration test of the new instrumented baffle in the EGO clean room. A laser was used to demonstrate a perfect weight balance and alignment between payload and baffle after the whole system is suspended by the two hires. A dummy mirror made of metal with identical dimensions and weight as the real one was employed in the test.
Figure 1:

As part of the SLC responsibility, members of the IFAE team, in collaboration with LIGO-Caltech and Virgo-EGO scientists, continued the development of simulations of the light propagation in the interferometer. Special emphasis was put in the understanding of the light distribution surrounding the main mirrors in the suspended areas, as an important input for the design of the instrumented baffles. The results led to several presentations and was accepted for publication in Classical and Quantum Gravity Journal, with IFAE people as corresponding authors [A. Romero et al., Class. Quantum Grav. 38 045002 [arXiv:2008.13740] (2020). In March 2020, one of the IFAE students was invited to Caltech (USA) with a LIGO visiting fellowship that unfortunately had to be cancelled due to Covid-19 restrictions. In addition, the IFAE team built the skills for the use of 2-D and 3-D optical simulation packages for tracing optical paths and determining how stray light ghost beams originate due to defects in optical elements. This served to determine the optimal installation strategy for the new signal recycling mirror at Virgo (before actual installation took place) and its parking position during the global Virgo alignment, as well as to identify potential sources of stray light coming from the signal recycling mirror. IFAE also contributed to the reduction of stray light contamination by complementing the interferometer with additional baffles in strategic locations. Virgo is installing the new frequency dependent squeezing system leading to new towers and vacuum pipes linking the new system to the detection tower. Simulations have indicated that the stray light contamination in the links needs to be suppressed in order to preserve Virgo sensitivity. This translated into the construction at IFAE of twelve small non-instrumented baffles to be integrated into the new vacuum pipes.

Physics Exploitation and Computing

In 2020, IFAE carried out part of its physics program, as designed according to the expected number of binary events in O3 (2019 – 2020), O4 (2022-2023) and O5 (2025 - 2026) observation periods. It contains three main pillars: the search and study of compact binary coalescence events (CBC) with emphasis on fundamental physics related to the test of General Relativity and dark matter searches; the use of Gravitational Waves (GWs) for cosmological tests; and the search for stochastic GW signals as probes for early Universe phenomena.

IFAE continued the search for GW signals from binary mergers adopting a deep learning approach in the analysis of the data. Different convoluted neural networks (CNNs) for low and high BH mass ranges were trained using O1+O2 data and a full bank of templates for different masses and luminosity distances of the BH-BH system. The results indicate that the CNN presents a performance comparable to that of dedicated analysis pipelines using matched-filtering techniques. Given the fast response of the CNN compared to that of the CPU-intensive matched-filtering pipelines, the online implementation of the CNN in low latency analyses is extremely promising. The results were translated into a short author list publication (A. Menéndez-Vázquez et al., arXv: 2012.10702) recently accepted in Phys. Rev. D. At the moment, LIGO-Virgo O3 data are being analyzed using CNNs. The study is being performed in collaboration with other EGO scientists (E. Cuoco et al. ) and within the framework of the EU COST17137 EU initiative.

A member of IFAE (C. Karathanasis) has played a central role in the finalization of the analysis for the determination of the Hubble constant using the O2 catalog data [LIGO-DCC-P1900015,, accepted for publication in ApJ] (see Figure 2). The IFAE group contributed to the development of the central code (Bilby, a Bayesian inference library) performing the parameter estimation of the masses and luminosity distance to the GW source, and the correlation with galaxy catalogs, using black holes with/without EM counterparts as standard sirens [LIGO-DCC-P2000168]. The results were published in a short author list paper [I M Romero-Shaw et al., Monthly Notices of the Royal Astronomical Society, Volume 499, Issue 3, 2020]. IFAE will play a leading role in the O3 version of the analysis. C. Karathanasis was already appointed as one of the corresponding editors of the future publication. In addition, C. Karathanasis participated in two short author list publications using two GW candidates (GW170817 and GW190521) with known EM counterpart candidates to constrain the Hubble constant and cosmological parameters, as well as the predictions from modified gravity theories beyond GR [LIGO-DCC-P2000236, LIGO-DCC-P2000390]. In particular, models with a time-varying Planck mass, large extra-dimensions, and a phenomenological parametrization covering several beyond-GR theories were considered. In all three cases, this introduces a friction term into the GW propagation equation, effectively modifying the GW luminosity distance. The results have been accepted for publication in JCAP [S. Mastrogiovanni, et al., arXiv:2010.04047 [gr-qc]].

Figure 2: The gravitational-wave measurement of $H_0$ (dark blue) from the detections in the first two observing runs of Advanced LIGO and Virgo. The GW170817 estimate (orange) comes from the identification of its host galaxy NGC4993 (taken from arXiv: 1908.06060 (2020)).

IFAE has also made major contributions to the LIGO/Virgo analysis searching for isotropic stochastic gravitational waves signals using O3 data. One of the members of IFAE (A. Romero) is at the core of the reduced team of analyzers producing the final results. For the first time, the Virgo data were included to cross-correlate the three interferometers. A new approach to gate the data was introduced to eliminate momentary very loud glitches in the data without a significant loss of data statistics. A new Bayesian analysis excludes the presence of correlated magnetic noise from Schumann resonances in favor of the data being explained only by Gaussian stationary noise. The final results are expressed in terms of 95% CL upper limits on the normalized energy density in GW from unresolved CBC sources in the range between 4E-9 and 6E-9, improving by a factor of five previous bounds. The paper [LIGO-DCC-P2000314] was recently submitted for publication in Phys. Rev. D (see Figure 3).

In parallel, members of IFAE (M. Martínez, A. Romero) in collaboration with IFAE theorists (O. Pujolàs et al., ) took the lead to promote inside LIGO/Virgo a reinterpretation of the results in terms of 95% CL limits on stochastic signals from cosmological first-order phase transitions from the early Universe at large temperature, relevant for QCD-axion models. The work has been carried out in close collaboration with other members of LIGO/Virgo from the USA and UK (M. Sakellariadou et al.). A short author list publication, with A. Romero as first author, [LIGO-DCC-2000518] was submitted to Phys. Rev. Letters [arXiv:2102.01714] (2020). It shows sensitivity to exclude model scenarios at very large temperatures (see Figure 3).

Figure 3: Posterior distributions for the CBC plus First-order phase transition search in the case of a phenomenological model with dominant bubble collision contributions as a function $\log \Omega_{\rm ref}$ (referring to CBC background) and the different parameters of the model. The 68$%$ and $95%$ CL exclusion contours are shown. The horizontal dashed line in the posteriors indicate the flat priors used in the analysis (taken from arXiv:2102.01714).

Contributions to 3rd generation projects

Although the current activities focus on the exploitation of LIGO/Virgo/KAGRA physics, the team plans for an early involvement in the initial discussion on 3rd generation interferometers, like the Einstein Telescope (ET), as long-term strategy in a moment when it is being considered as a possible ESFRI infrastructure in Europe, and given the building links with CERN and other HEP institutes in Europe. As a member of the ET Steering Committee, M. Martinez drove in 2020 the Spanish efforts to aggregate interest on the ET project, leading to the explicit expression of interest by 23 institutions in Spain including 4 ICTS, which in turn translated into the formal political support from the Spanish funding agency to the ET ESFRI candidature. M. Martínez contributed to the update of the CDR document and he has been recently appointed as co-coordinator of the stray light control working group in ET’s Instrument Science Board.

The IFAE team in Virgo has invaluable support from the PIC computing center. In 2019, PIC was fully integrated in the LIGO-Virgo computing grid, providing opportunistic resources to the experiments. In 2019-2020, PIC contributed 7% to the total LIGO-Virgo CPU accounting and about 4% of the GPU accounting.


The IFAE group has been also very active in organizing different workshops related to GW physics. IFAE organized the Virgo Collaboration meeting in April 2020 which, due to Covid19 was finally cancelled and postponed to a future date; co-organized the 2020 Spanish Meeting on ET ESFRI candidature and co-organized the 2020 Iberian Meeting on Gravitational Waves.