Transcript Search for Bursts in the S1 Data

Real-time Astronomy with LIGO and
Virgo: Status and Prospects
Erik Katsavounidis
LIGO-MIT
for the LIGO Scientific Collaboration and the Virgo Collaboration
January 28, 2010
GWDAW14 – Rome
LIGO-G0901158
1
The path to gravitational wave
astronomy
In this conference:
 First generation individual detectors
first global
network
 First generation networks
second, third generation
and beyond
 First detections
source astrophysics
Actively engaging the full spectrum of electromagnetic and
particle astrophysics
LIGO-G0901158
2
The S6/VSR2 paradigm
 LIGO and Virgo are currently (and since July 7th 2009) in
their S6 and VSR2 science runs (respectively):
» An opportunity to improve data analysis infrastructure and organization with
respect to the previous runs of the instruments
» A test case for the advanced detectors regime
 Immediate goals (on the data analysis front):
» Near real-time Electromagnetic (EM) follow-up of outlier events (possibly
detection candidates) in order to catch the early light from astrophysical
sources
» Rapid follow-up with gravitational-wave (GW) detectors of external
triggers with the intention to provide a statement on gravitational wave
emission associated with them (see talks by A. Dietz, S. Marka)
» Prompt detector characterization: improve instruments and ultimate
scientific reach for the off-line searches by promptly identifying and
hopefully fixing noise sources affecting the sensitivity of the instruments
(see talks by N. Christensen, F. Robinet)
» Improve turn-around time of final science results and publications
LIGO-G0901158
3
Scientific rationale
 GW sources are reasonably likely to emit in the EM too
» UV/optical afterglows from supernovae/GRBs have been observed peaking on time
scales of hours to days
» prompt X-ray outbursts, bright X-ray afterglows have been observed in connection
with core-collapse supernovae, GRBs
 Also likely for GW sources to be nearby (so that to be detected with
initial instruments), EM detection generally easier
 EM signatures may be missed (e.g. beaming effects, simply because
not looking at the right time at the right place)
 Even if at the current sensitivity level, few -if any- detections are
expected, the scientific payoff in case of a discovery will be tremendous
» reinforce GW detection
» provide position of the source with much reduced error circle
» host galaxy, distance
 Connecting observations from many wavelengths and GWs 
understanding astrophysical processes  multi-messenger astronomy
LIGO-G0901158
4
The real-time implementation challenge
 Over the summer-fall 2009, the burst and binary inspiral search
groups in LIGO and Virgo have put in place near real-time, O(10min),
search pipelines looking for unmodeled bursts and low mass inspirals
(see talk by Damir Buskulic) in S6/VSR2
 Burst searches are as complete as possible and very similar to
searches run “offline”
» Use h(t) data calibrated online (~1 min latency) and transferred from all detector
sites to a single one
» They process 3-site data with ~10min latency, identify significant events and
threshold them according to user-defined criteria
» Establish background in order to assess events’ significance
» Apply data quality criteria
» Provide reconstruction of the events’ sky position
» 500-600 CPUs are employed in order to keep up with data and within the O(10min)
mark
LIGO-G0901158
5
Position reconstruction of burst sources



Formal study over a broad range of simulated signals added on S5/VSR1 and
S6/VSR2 instrument data (the “Position Reconstruction Challenge”)
Performance varies significantly with signal-to-noise ratio (SNR), morphology,
analysis parameters
5-10 degrees near threshold, 1-2 degrees for “loud” signals
Position error areas hide the fact that they may be broken down to many disjoint patches
LIGO-G0901158
6
Galaxy targeting
 Position reconstruction error area can be further reduced if priors
on the directional origin of a GW burst are assumed: known
galaxy in our nearby universe, the Milky Way!
 Galaxy catalogs have been developed by LIGO-Virgo
collaborators for the purpose of compact binary searches and the
EM follow-up (Kopparapu et al. 2008, ApJ, 675, 1459; Daw,
Dhillon, and White, in preparation)
 These catalogs give positions, as well as estimates for distances
and blue luminosities, for over 50,000 galaxies out to a distance
of 100 Mpc
 About 20% of 0.5x0.5 sq. deg. fields intersect a known galaxy
within 100 Mpc
LIGO-G0901158
7
Initial partner telescopes
 Swift satellite mission:
» capable of performing multi-wavelength observations
of GRBs and their aftermath. Three different
telescopes record gamma-ray, X-ray, ultra-violet, and
optical light
» UV/optical telescope: 0.4x0.4 sq. deg. FOV
» X-ray telescope: 0.3x0.3 sq. deg. FOV
» 3 Target-of-Opportunity (ToO) observations granted for
Cycle 5 (April 2009-March 2010)
Swift
 TAROT robotic observatories in France and
Chile
» 1.85x1.85 sq. deg. FOV
» follows up GRBs, carries out a supernova search
 QUEST in Chile
Pi of the sky
» 4.6x4.1 sq. deg. FOV
» survey telescope for supernova searches, etc.
 Pi of the sky
» 20x20 sq. deg FOV run by the U of Warsaw
 Signed MOUs for collaborative work (and
accepted to image ~1 target/day)
 Awarded ToO time on the Liverpool telescope
for the April 19th – June 15th time window
LIGO-G0901158
TAROT8
Putting it all together…
 Starting ~20th of December and through January 8th the first EM followup program took off the ground!
 The real-time burst search identified outlier events that were pushed all
the way to obtaining images with our initial telescope partners
 Burst events had to meet certain significance requirements when
compared with background, be localizable in the sky with decent
probability and within the instruments’ FOV budget and meet basic data
quality requirements
 Operation driven by scripts collecting events from the search methods in
real-time, processing them and communicating them via e-mail to a
designated and volunteered group of LIGO-Virgo collaborators
 Event-by-event decisions (to follow up in EM or not) could be reached
typically within 30minutes
 Tight (and crucial) coordination with the instrument control rooms,
telescope principals, LIGO-Virgo search groups altogether
 EM data collected and are currently being analyzed in collaboration with
telescope principals
 Extremely fruitful exercise on numerous fronts!
LIGO-G0901158
9
Plans for the remaining of
S6/VSR2
 LIGO and Virgo have short and long commissioning breaks planned for
most of Winter/Spring 2010, but likely to come back in triple-coincidence
mode of operations in later Spring/early Summer 2010
 Identify lessons-learned and address shortcomings of the first EM followup program
 Expect initial partners to continue working with us (Swift, TAROT, QUEST,
Pi of the sky)
 Establish additional (if) telescope partners – expand in other frequencies
(radio: NRAO/VLA ToO?, LOFAR and MWA ToO-style of observations)
 Prepare for a restart of such program that includes bursts and compact
binary inspiral triggers feeding into it
 Challenges never stop!
» Understand EM backgrounds
» Understand role of priors assumed for gravitational-wave transient signal (e.g., origin
from within known mass in the Universe or not, morphology and polarization etc)
LIGO-G0901158
10
Summary and Outlook
 LIGO and Virgo have completed a ~3-week long near-real time search for
bursts that allowed the follow-up of outlier events in the optical and X-rays
 “Near-real-time” requirements for such a project have been largely met, with
outlier events becoming available for EM observation within 30-60 minutes
after data were collected
 Many lessons were learned from this exercise and a rich to-do list with more
challenges in it awaits us
 Efforts so far to catch EM counterparts to gravitational-wave transients have
been opportunistic in nature
 Important path-finding exercise for routine detections expected with
advanced detectors in 2014-2015 and the new astronomical messenger
they will bring in
 Need to start getting ready now:
» put the right software tools in place
» think out well how the coordinated GW-EM searches will be performed
LIGO-G0901158
11