Source Modeling, Numerical Simulations, and Data Analysis Joan - - PowerPoint PPT Presentation

Source Modeling, Numerical Simulations, and Data Analysis

Joan Centrella

Laboratory for High Energy Astrophysics NASA/Goddard Space Flight Center

1

Source Modeling

• guide data analysis strategies
• provide templates
• identify sources

Data Analysis

• identify signals and sources
• feedback to astrophysics
• input to scenario building

2

Different scientific communities have different styles, approaches, and goals….

• Experimental Physics: traditional GW experimentalists

+ high energy physicists….

– build instruments, make measurements… – unambiguous detection of gravitational waves

• Fundamental Physics: relativists….

– test GR and constrain theories of gravity – explore strong-field regime of gravity – unambiguous detection of black holes

• Astrophysics: observers, theorists, modelers…

– GW are a new window through which to observe the universe – build physical models of the sources – build scenarios for BH formation, stellar evolution, structure formation, cosmology….

3

Data analysis: a simple view...

• Analyze the data from the detectors
• Identify the signals

– separate out the instrumental noise…. – remove environmental, androgenic, etc. effects – obtain the physical signal coming from the astrophysical source: determine which quantities are measured, and their values…

• Identify the sources: type, location.…

4

Astrophysical source modeling….

• Astrophysical scenarios

– types of sources, rates, gross characteristics… – provides input to detailed modeling…

• Models of specific sources

– formation and evolution of sources – produce waveforms, spectra…

• Two approaches to source modeling:

– analytic: derive expressions for key features of model analytically or by the solution of ODEs – numerical simulations: large-scale computational models obtained by solving sets of coupled PDEs in 3-D

• numerical relativity
• GR hydrodynamics

5

Numerical simulations of astrophysical sources…

• needed for many of the most interesting, energetic sources
• pose challenges for data analysis
• focus on 2 broad categories of sources:

– Collapses

• stellar collapse: Type II supernovae, AIC…
• supermassive stars
• Pop III stars

– Mergers

• NS/NS
• NS/BH
• BH/BH

– stellar BHs – intermediate mass BHs – massive BHs (MBHs)

6

Collapses….

• Spherical: no GW
• Axisymmetric:

– Type II supernovae are the best-studied – many parameters: input physics, rotation laws…. – Zwerger-Mueller: most detailed study, catalog of 78 waveforms short duration bursts: free-fall and bounce(s) with various waveform shapes

7

Rotational instabilities….

• Non-axisymmetric:

rapidly rotating stars can undergo global rotational instability bar formation

• equilibrium cores: simulations

show long-lived bar (New, Centrella, Tohline)

• friction with envelope could

disspiate bar

• may be difficult to form bar

during collapse (Brown)

8

Coalescing Binaries….

• Inspiral: widely separated, use PN approximation for

point masses

– early stages treated by analytic means templates

• Intermediate regime: PN approximation breaks down

– “effective one body” techniques…. – quasi-equililbrium models….

• Plunge and merger: components depart from quasi-

circular orbits and merge on dynamical timescales

• Ringdown: merged object radiates GW and settles

down into an equilibrium state

– black hole: quasinormal ringing well-understood – neutron star: more work needed on modes…

9

Mergers…burst sources.

• Need numerical simulations of Einstein eqns + GR hydro (for NS)
• Plunge begins at or near the ISCO…
• For NS/NS and NS/BH, signature of ISCO in data important to

extract physical information (e.g. EOS…)

• Sensitivity to system parameters:

– SPH simulations of NS/NS merger by Faber and Rasio

10

Full GR codes currently not able to evolve > 1 orbit near ISCO due to instabilities

Efforts for using different techniques to evolve near ISCO, through merger, and in ringdown…

• BH/BH mergers: Lazarus approach

(Baker, Campanelli, Lousto and Takahashi)

– Full GR Cactus code for merger – perturbation equations for ringdown

• NS/NS mergers: QE models near ISCO

(Duez, Baumgarte, Shapiro, Shibata, & Uryu)

– quasi-equilib models – match to full 3-D numerical relativity simulation

11

Challenges for simulations…

• Stability of numerical relativity codes…
• Numerical dissipation:can cause spurious inspiral (New & Tohline)

e.g. co-rotating equilibrium binary

– stable when evolved in rotating frame – spirals in when evolved in inertial frame – lower order advection worsens the effect

• How to handle the BHs: excision…
• Push the limits of computer power
• Multiple spatial and temporal scales adaptive mesh refinement
• Extraction of gravitational waves…
• Novel approaches: Lazarus…
• What quantities can be calculated that will help in confirming

detection of various sources?

12

LIGO’s sensitivity to coalescing binaries and burst sources . . .

13

LISA will observe a rich variety of sources, from MBHs to Galactic binaries . . .

14

Challenges for data analysis.…

• What strategies could be used to detect collapses and

mergers?

– matched filtering requires sufficient number of templates… – time-frequency, slope-detector, excess power statistic,….

• What strategies could be used to identify such

sources?

– specific type: merger versus collapse – specific case:

• NS/NS merger vs. NS/BH merger….
• collapse of 10 M∃ star vs. 20 M∃ star….
• For LISA, will MBH/MBH coalescences obscure

sources at smaller S/N?

– remove bursts from data stream “lose” this data – need to know rates…