Core collapse supernovae are extraordinary stellar explosions that mark the death throes of massive stars, with mass greater than eight to ten times the mass of the Sun. Such explosions are the dominant source of elements in the Universe between oxygen and iron and are believed to be responsible for half of the elements heavier than iron. They are a key link in the origin of life in the Universe. More massive stars develop cores with masses exceeding the Chandrasekhar mass and they must undergo gravitational collapse in their deaths.
Our understanding of weak interactions (especially concerning electron capture rates), the equation of state of the star, and new computer calculations which couple general relativity, hydrodynamics and neutrino transport, are the ingredients to simulate the physics of the gravitational collapse.
However, gravitational waves (GW) and neutrino emission are the only way of directly observing the core dynamics of a supernova.
The entire process seems to have a few distinguishing characteristics. The emission consists of a first prominent peak associated with star core bounce. Then, after bounce, we have the excitation of modes in the proto-neutron star: this part of the signal can last for several 100 ms.
These feature of the GW signal should appear in every core-collapse supernova and our proposal is to use techniques of pattern recognition to extract the signal in the GW data burden produced by the advanced GW detectors LIGO and Virgo. To have an efficient computational approach we plan to develop codes dedicated to computers equipped with the last generation of Graphical Processing Units (GPUs).
The search goal is to find GWs and neutrinos of common origin.
A neutrino carries information about the event that produced it. Through the study of solar neutrinos, scientists have deduced a great deal about the inner workings of our sun, including what kind of nuclear processes keep it burning. Supernova neutrinos hold the promise of carrying similar information about what goes on inside an exploding star. Those explosions are thought to produce all the heavy elements in our universe elements crucial for life as we know it, but major uncertainties surround this theory. Learning about the processes taking place inside a supernova might settle some of the debate. This information will most certainly increase scientists understanding of those intense nuclear reactions and feed into the study of nuclear physics, which has applications in energy research.
There are multiple scientific benefits of simultaneously observing GWs and high-energy neutrinos from a common source:
- the combined information from GW and high-energy neutrino observatories can greatly enhances our confidence in a joint detection
- GWs and high-energy neutrinos both carry information from the depth of their source that is, to a large extent, complementary to the information carried by electromagnetic radiation. While the GW signature of cosmic events is characteristic of the dynamics of their central engine, a high-energy neutrino flux is reflective of the presence of hadrons in the relativistic outflow generated and driven by the central engine.
One of the most interesting prospects of joint GW - high-energy neutrino searches are common sources that are dark in gamma rays. Prominent sources of this type are choked GRBs or low-luminosity GRBs. These sources are difficult to detect with electromagnetic observatories, and hence provide an exciting opportunity to joint GW+high-energy neutrinos searches that can discover them and/or constrain their population.
The data of the GW detectors, as LIGO and Virgo, are available and they can be analysed jointly with those of neutrinos identified by detectors as ANTARES and KM3NeT in Sicily, IceCube in the South Pole and other neutrino telescopes focused on low energy neutrinos as LVD, Kamiokande II and Borexino . With all these experiments carried on by international collaborations , there are already memorandum of understanding permitting the data exchange.
The multimessenger search will be developed to look for temporally and directionally coincident GW and neutrino signals.
The idea is to perform a full sky search and additionally use galaxy catalogues, as many of the target phenomena are expected to be occurring from within or near galaxies other than the Milky Way.
The joint analysis uses a test statistic for GW + neutrino event candidates, which combines the significance and directional distribution of GW and astrophysical neutrino event candidates.
As both GWs and high-energy neutrinos can arrive prior to the onset of electromagnetic emission from sources such as GRBs, joint GW + high-energy neutrino events may be used as trigger to search coincident data taken by the satellites looking for GRB electromagnetic signals.
IceCube, ANTARES and KM3NeT and the low energy neutrino detectors will provide search triggers. On the base of these triggers we will select the GW data stretches around this time with a typical time windows spanning from 500 o 1000 s. In parallel longer stretches of data will be selected adjacent to those windows to assess the statistical confidence of the detection. We will use data of the LIGO and Virgo interferometers, located in different sites on the Earth. This implies that the arrival time of GW signal in the various detectors has difference of several milliseconds depending on the source location in the sky. The high energy neutrinos produced from turbulent gaseous environments left over by supernova explosions and emitted far from the core, will provide this information.
The main concern to extract GW triggers related to supernovae explosions, is to develop efficient method for extracting this category of signal from the noisy data provided by the network of advanced GW detectors, as LIGO and Virgo.
We propose to solve the problem of searching these features in the time-frequency plot of the GW data by using pattern recognition techniques.
In this case the generation of the training set will be based on phenomenological templates of GW signals, emitted by supernovae. These are provided by the numerical relativity groups as that of the university of Valencia, member of the Virgo collaboration. First these signals will be embedded in Gaussian noise to debug the algorithm and then we will inject them in the real data to measure the efficiency.
As for all applications dealing with image analysis, the most efficient computational approach is based on the use of dedicated computers equipped with the last generation of Graphical Processing Units (GPUs).