Skip to content

Changes

Page history
Update siqm post o3a plans authored by Muhammed Saleem's avatar Muhammed Saleem
Show whitespace changes
Inline Side-by-side
siqm-post-o3a-plans.md
View page @ 01efe7b2
......@@ -2,6 +2,28 @@
This page is for top-level management of all the ongoing activities. Avoid adding low-level details which can be added as additional pages linked from here. At present, we think of three follow-up studies to be concluded in the next three months. The broader idea and the milestones are given on this page. Milestones are marked :white_check_mark: once achieved. The plans are below.
# Stealth/systematic biases in the compact binary parameter estimation due to spin induced moments.
To assess how the presence of a non-BH compact object with spin-induced deformations can bias the inferences of binary parameters such as masses, spins and distances. Preliminary studies are done and found that the spin inference is most affected by ignoring the non-unity value of kappa parameter. Apart from reporting how different parameters get affected, it might be worth also exploring how the formation scenarios inferences (based on effective spin distribution) can also get biased. [Dedicated page](/nv.krishnendu/lalsuite/-/wikis/Stealth%20bias)
#### Detectability of non-BH binaries with BBH template banks
[This paper](https://arxiv.org/abs/2004.06729) argued that if $\\delta\\kappa > 20$, then BBH template banks might miss them. We need to revisit this with more realistic settings such as IMR waveforms and low spin events etc.
Knowing the edges of the parameter space of detectable binaries is extremely important in the inference of kappa. Right now we use a wide parameter space that spans to regions that will not have detectable SNR. This will add unwanted Occam's penalty into the Bayes factors, eventually leading to the BBH fraction estimates.
#### The systematics study should cover the following:
1. Systematics from the prior choices. Are we using prior wider than needed? Assess this from the analysis on what can be detected. Should we expand the template bank to make them catch BH mimickers?
2. Once we identify and start using the correct prior, investigate how the hidden SIQM moments can bias the BBH PE. What kind of system will stand out from BBH and what kind of system will mimic BBHs? A combination of spins, masses, and distance will be required to answer these. See [this presentation ](https://git.ligo.org/muhammed.saleem/stealthbias/-/blob/master/siqm-population/population_and_stealthbias.pdf)for the details of preliminary studies on this.
3. Population inference suffers systematics from accumulated Occam's factor that comes from the inappropriate prior. How can we correct this by marginalizing over the hyper-parametrized prior?
4. Is there an optimal choice of the parameter to be estimated? Is it indeed Ks with the Ka=0 assumption? Is this assumption leading to any systematics? Can we do SVD to find the optimal choice?
### Milestones
* Pankaj and Sajad: Read the population paper (method part and appendix) and schedule a meeting with Krishnendu/Saleem
*
# Deriving constraints that can be used to constrain the BH mimicker parameter space
Currently our bounds are powerful in terms of searching for non-BH signatures but are not useful in constraining the parameter space of BH mimickers due to the strong assumptions that has gone in. There are two scenarios where this test can yield bounds that are more useful for theoretical community. [Dedicated page](https://git.ligo.org/nv.krishnendu/lalsuite/-/wikis/constrain-BH-mimicker-parameter-space)
......@@ -13,7 +35,7 @@ Both can yield meaninful constraints on the BH mimicker parameter space, though
---
# Inference on the population of BBH or BH mimicker binaries :white_check_mark:
# Population Inference of the BH-mimicker binaries :white_check_mark:
Use the O1/O2 population to make predictions on our future capability to identify BBH and non-BBH populations. [Dedicated page](SIQM-population-inference)
......@@ -29,15 +51,7 @@ Use the O1/O2 population to make predictions on our future capability to identif
---
# Stealth/systematic biases in the compact binary parameter estimation due to spin induced moments.
To assess how the presence of a non-BH compact object with spin-induced deformations can bias the inferences of binary parameters such as masses, spins and distances. Preliminary studies are done and found that the spin inference is most affected by ignoring the non-unity value of kappa parameter. Apart from reporting how different parameters get affected, it might be worth also exploring how the formation scenarios inferences (based on effective spin distribution) can also get biased. [Dedicated page](Stealth%20bias)
### Milestones
* Launch all the required runs. The injection parameters are created. Stuck at some technical point in reproducing the correct SNR (the run identifies a different sky-location than the one injected). Krishnendu/Saleem
* Revisit the basic idea and take stock of the existing studies and results (Krishnendu).
* Drafting
#
---
......@@ -55,12 +69,6 @@ To explore how the measurement of kappaS is possible for signals with higher har
---
# Detectability of non-BH binaries with BBH template banks
A [recent paper](https://arxiv.org/abs/2004.06729) argued that if $\\delta\\kappa > 20$, then BBH template banks might miss them. We need to revisit this with more realistic settings such as IMR waveforms and low spin events etc.
---
# Action items (currently unfinished)
* Krishnendu: submit all the 01-O2 events for imr case: :white_check_mark: [details](https://git.ligo.org/nv.krishnendu/lalsuite/-/wikis/SIQM-population-inference/o1-o2-event-details)
......