Lecture Questions - Day 2 (2023)

Can you talk more about limitations of it

FAR is important to decide that the detection is due to the astrophysical source, not from the background noise. FAR estimates what is the chance that the detection is due to the background noise.


FAR can also be useful for follow up of the astrophysical event in electromagnetic spectrum data. Like the GW170817 was observed in Gamma ray band also. So to detect these EM events associated with GW events, you need to identify these events as of astrophysical events which FAR helps and then you can communicate with EM astronomers to look for these events using other Telescopes.


Bayeswave uses wavelets to model signal and noise transients, so how does it distinguish between the two?

What is ground motion noise?

What is the difference between category 2 and category 3 vetoes? I don’t understand the difference in wording of the definitions, specifically the difference between “known noise coupling to h(t)” (category 2) and “statistical noise coupling to h(t)” (category 3)

These are distinguished exploiting the coherence of the data among different detectors: roughly speaking, you fix some parameters of the wavelets in evaluating data of different detectors if you are modeling a signal, while in modeling noise you are evaluating totally uncorrelated wavelets

Hi @lua, with ground motion noise we mean every vibration or oscillation of the ground, e.g. earthquakes, a truck passing nearby on a bridge, a train or an elephant falling from the first floor. Here a reference Limiting the effects of earthquakes on gravitational-wave interferometers or LIGO Technology | LIGO Lab | Caltech .

In here
What does f_lower mean? Is it binary rotation frequency? Then how do we know its starting frequency?

f_lower means the frequency the gravitational wave radiation, which is twice that of the orbital frequency (check https://arxiv.org/pdf/1608.01940.pdf where all is explained in a very clear way). Actually the f_lower you set is the frequency you give to your model to start to generate the waveform, i.e. you consider your system when its gravitational wave radiation starts to have a frequency equal to f_lower. E.g. if you fix f_lower=20 Hz you have to consider indeed that the binary system is rotating well before in time and with lower frequencies, but the fact that gravitational wave emission is lower in this pre-stage jointly with our lower sensitivity at low frequencies implies that usually we detect the binary system only at its last stage