1. Brennan Hughey MIT Kavli Institute Postdoc Symposium 4-3-09
The Burst Analyses: Searching for Unmodeled Gravitational Wave Transients with LIGO
Brennan Hughey
MIT Kavli Institute Postdoc Symposium
4-3-09

2. LIGO and Friends LIGO Livingston (LA) L1 – 4km LIGO Hanford (WA)
H1 – 4km H2 – 2km
Virgo 3 km (Italy)
Has data-sharing
agreement with LSC
GEO 600m (Germany)
Part of the LSC

3. Sources And Methods This talk Template-less methods Matched filter
Bursts
Stochastic Background
Template-less methods
Compact Binary Inspirals
Pulsars
Matched filter
Short
duration
Long
This
talk

4. Burst Analyses Sources of GW bursts include: Core-collapse supernovae
Merger of binary compact objects
Neutron star instabilities
Cosmic string cusps and kinks
Unexpected sources?
Approaches:
“All-sky” searches performed over whole sky and over all
Collected data
“Externally triggered” searches performed around the time
Of Gamma-Ray Bursts or Soft Gamma Repeaters

5. Analysis Methods Excess Power:
look for significant upward deviations from background expectation
Perform coincidence test with other interferometers in network to reduce background
Cross-correlation
Look for consistency in waveforms
observed in multiple interferometers
Fully coherent methods:
Reconstruct events hypothesizing
sky locations and accounting for
amplitude, time delay
Coherent WaveBurst sensitivity for LIGO
Network in Earth-centered, Earth-fixed
coordinates

6. Data Quality
Most triggers which stand out above background are of mundane origin
(glitches)
LIGO has hundreds of readout channels and auxiliary detectors to
provide additional information about event
seismic
wind
electromagnetic (power lines)
acoustic
Data quality flags remove periods
of dubious quality
Vetoes remove candidates in
direct coincidence with specific
channel

7. All-sky searches S5 all sky search 4 science runs of LIGO-GEO
instruments in
35 days triple coincidence
No event candidates
Upper limit (90%
confidence level on rate of
detectable events) set S2 S4
LIGO S4 search: CQG 24 (2007) 5343
Preliminary
S5 all sky search
Analysis underway, split
into 2 calendar years
Order of magnitude more
data and X2 sensitivity of S4
Exploiting coherent analysis
methods and full network
Detection Probability

8. High Frequency All Sky Searches
All-sky searches previously limited to 2 kHz or lower
New analyses extend range of analysis up to 6 kHz
Most source types previously listed can have >2 kHz emission
Use same methods as low frequency, adjusting as needed
Takes full advantage of Virgo
Clean, shot-noise
dominated background

9. Triggered Searches GRB searches –
137 GRBs (with at least two active interferometers) analyzed during S5
Mostly reported by Swift satellite
Analyzed using coherent methods
GRB –
Special case, earlier separate publication
LIGO excludes compact merger within M31 as source at 99% CL
SGR searches –
Individual SGR Burst Search
PRL 101, (2008)
Sensitive to neutron star f-modes
Stack-a-flare
And other analyses not listed here

10. S5 analyses wrapping up, several papers
coming out over next few months
S5 Year 1 High Frequency Search paper
should go to arXiv in late April or May

11. Data Analysis Prep for enhanced LIGO/Virgo+
Burst group is working on upgrades to pipelines to use with upgraded
detector in enhanced LIGO/Virgo+ joint science run
Biggest difference is move towards low-latency online analysis.
Candidate events minutes after initial processing
There are several goals behind this effort:
Assist detector characterization efforts
Expedite offline analysis
Work towards making LIGO/Virgo an integral part of the astronomical
Community
Quicker follow-up of events from other observatories
Produce event candidates for follow-up at other astronomical
observatories

12. Triggering other Detectors
EM observatories:
Wide-field optical telescopes: SkyMapper, ROTSE, TAROT, Quest, etc.
Radio: LOFAR, Arecibo, VLA, etc.
Gamma-ray/x-ray: Swift, etc.
Neutrino telescopes:
Super-k, IceCube, Antares, etc.
both supernova neutrinos and high energy
neutrino detectors don’t point, but sharing of data sets aids coincidence
Details vary from collaboration to collaboration
(e.g. Swift triggers will be much rarer than most of the ground-based telescopes)
In all cases, majority of triggers will be false alarms and we are clearly
communicating this
Independent confirmation that an astrophysically interesting event is happening
at the same time and place as our event candidate would greatly enhance our
confidence – especially for the first direct gravitational wave detection

13. Analysis Flowchart

14. Summary
Searches for gravitational wave transients have been conducted using
a variety of methods
Papers coming to publication over next few months
Enhanced LIGO/Virgo+ science run starting around July
“Online” low-latency analysis will enable much quicker turn-around
and facilitate EM follow-ups