Projects | Ronaldas Macas

Fast glitch modeling

Mon, 01 Jan 0001 00:00:00 +0000

Gravitational-wave data contains excess noise that affects the estimation of astrophysical source parameters. In some cases, the excess noise looks nearly identical to the noise seen previously, enabling us to categorize this noise into different classes (so-called glitch classes).

In my previous work (see antiglitch ), I showed that it is possible to model short-duration glitches using a quasi-physical model. Using JAX, the glitches can be fitted to the quasi-physical model which then allows us to remove these glitches.

In this project, I show that short-duration glitches like blips and tomtes can also be modeled using autoencoders. With an autoencoder, I am able to fit a glitch model to the data orders of magnitude faster than the previous JAX-based fitting method, while keeping the same precision.

View the code on GitHub .

Antiglitch

Mon, 01 Jan 0001 00:00:00 +0000

The excess noise in gravitational-wave data can be categorized into different classes based on the noise morphology. Some excess noise is very short and well-localized in time, e.g. so-called blips, while other noise sometimes lasts minutes, for example, light scattering noise.

In this project, my collaborators and I made a quasi-physical model for three classes of noise: blips, tomte and koi fish. The model has five parameters: amplitude, time, phase, frequency and bandwidth. Using JAX, we fit these five parameters to the gravitational-wave data which allows us to precisely model the excess noise. Once the excess noise is modeled, it can be removed from the data thus allowing for more accurate estimation of astrophysical source parameters.

Read the paper in Physical Review D or the free version on arXiv .

Probabilistic noise estimation

Mon, 01 Jan 0001 00:00:00 +0000

Gravitational-wave interferometers are among the most sensitive instruments in the world. They can measure gravitational-wave strain as small as $10^{-20}$. However, such a sensitivity comes at a cost: even ocean waves hundreds of kilometers away can introduce noise in the gravitational-wave data.

This excess noise needs to be removed, especially if it overlaps an astrophysical signal in the gravitational-wave data. For example, I created a machine-learning algorithm to remove the radio frequency noise around GW200129 (see the nlsub project ).

To determine whether this algorithm removes noise better than other techniques, I had to develop a sensitive method to measure the amount of non-Gaussian noise in the data. Using mixture models, I showed that my machine-learning algorithm indeed removes more noise than the conventional methods used at the time.

Read the paper in Physical Review D or the free version on arXiv .

Sky localization

Mon, 01 Jan 0001 00:00:00 +0000

Gravitational-wave data often contains noise artifacts (also called glitches). In fact, about 24% of gravitational-wave candidates during the third LIGO-Virgo-KAGRA observing run had a glitch nearby. A glitch can be confused with a gravitational-wave signal causing incorrect estimation of source parameters.

One example of a problem arising from confusing glitches with astrophysical signals is the incorrect sky localization of the signal. If the sky localization of an astrophysical signal is incorrectly determined, an electromagnetic counterpart is less likely to be detected. This would result in losing loads of information about the source of gravitational waves.

I tested how different glitches affect the sky localization of various astrophysical signals (binary black holes, binary neutron stars, and neutron star-black hole binary). I found that short signals, such as binary black holes, are not usually affected by glitches unless they exactly overlap (see the image above). However, longer-duration signals like a neutron star-black hole binary are affected by glitches.

Read the paper on arXiv , published in Physical Review D.

Non-linear noise subtraction

Mon, 01 Jan 0001 00:00:00 +0000

Gravitational-wave noise can sometimes overlap an astrophysical signal in the gravitational-wave data. A perfect example is a binary black hole merger GW200129 which overlapped with radio-frequency noise. This noise introduced additional uncertainty when estimating binary black hole parameters: by some estimates, the binary was not highly precessing, while other estimates indicated that this was the most strongly precessing binary black hole ever observed.

Luckily, the radio frequency noise was also recorded by witness channels, i.e. the channels that record only this particular noise but not the astrophysical signal. Our neural network reduced the noise more effectively than the method used by the LIGO–Virgo–KAGRA collaboration. Reanalyzing the cleaned data showed that evidence for precession remained.

Read the paper on arXiv .

View the code on GitHub .

Gamma-ray bursts

Mon, 01 Jan 0001 00:00:00 +0000

Gamma-ray bursts are extremely energetic explosions, first observed in the 1960s with satellites built to detect nuclear explosions on Earth. There are two scenarios of how such energetic bursts can be formed naturally: by a collision of two stars (e.g., neutron stars) or when a massive star reaches the end of its life and implodes. In both cases, we also expect gravitational waves to be produced due to the large quadrupole moment.

Advances in satellites and observing techniques now allow about one gamma-ray burst to be observed daily. However, many of these bursts are so distant that gravitational-wave detectors cannot find the counterparts for these bursts using usual gravitational-wave searches. As a result, we perform a subthreshold coincident search which is more efficient because the time and location of a gamma-ray burst is known.

I analyzed the data around tens of gamma-ray bursts throughout the second and third LIGO-Virgo-KAGRA observing runs, including the famous GW170817–GRB 170817A. In addition, I led the analysis group that searched for unmodeled gravitational waves in coincidence with gamma-ray bursts. As a group leader, I presented the analyses and results at various conferences and seminars.

For more details, see the collaboration papers published in ApJ: O2 search results , O3a search results , O3b search results , GW170817 and GRB 170817A .

GLADE

Mon, 01 Jan 0001 00:00:00 +0000

Searching for gravitational-wave counterparts with electromagnetic waves requires galaxy catalogs. Because we expect that the majority of gravitational-wave events should happen in the Universe where stuff exists, i.e., galaxies, having a galaxy catalog significantly increases the chances of finding a counterpart. Finding an electromagnetic counterpart reveals information that gravitational waves alone cannot provide. For example, GW170817 was observed in both gravitational waves and light, linking a neutron star merger to a short gamma-ray burst and providing evidence that these mergers produce heavy elements.

My collaborators and I produced GLADE, a galaxy catalog for multimessenger searches in the advanced gravitational-wave detector era. I contributed to this work by developing a statistical method to merge different smaller galaxy catalogs into a single, bigger catalog (GLADE). It is not straightforward to merge multiple galaxy catalogs into one because some of the entries in the catalogs may actually be referencing the same source. Hence a method that avoids duplicate entries is desirable.

The method I developed to merge different catalogs into one while avoiding duplicate entries relies on the fact that there are duplicate entries in galaxy catalogs. For example, we would expect different all-sky catalogs to observe M31 (one of the biggest and closest galaxies to us) and some other bright galaxies. By identifying objects with the same name between different galaxy catalogs, we can estimate the variance of various parameters between different galaxy catalogs. This allows us to select a contamination threshold. For example, a threshold of 0.99 corresponds to a missed-duplicate rate of 0.01 (1 − 0.99): 1% of duplicate entries are incorrectly treated as separate galaxies.

Read the paper in Monthly Notices of the Royal Astronomical Society .