Voeventnet.caltech.edu VOEventNet By Matthew J. Graham (Caltech)

voeventnet.caltech.edu

VOEventNet

By Matthew J. Graham (Caltech)

VOEvent II6 December 2005

What is VOEventNet? Real-time astronomy with a rapid-response

telescope grid A peer-to-peer cyberinfrastructure to enable

rapid and federated observations of the dynamic night sky

A network of telescopes and computers working synergistically, under the watchful eye of humans, to find and study interesting astronomical events

A transportation of events to interested subscribers, automatically in seconds or minutes after discovery

VOEvent II6 December 2005

What is VOEventNet? $600k 3 year NSF-funded project

under the DDDAS (Dynamic Data-Driven Applications Systems) initiative involving Caltech, UC Berkeley and LANL

Personnel:

^

really

Roy Williams (PI)Joshua BloomGeorge DjorgovskiShri KulkarniThomas Vestrand

Matthew GrahamAshish MahabalAndrew DrakeDerek FoxPrzemek Wozniak

VOEvent II6 December 2005

Event Synthesis Engine

Pairitel

Raptor

PQ next-daypipelines

catalog

Palomar-Quest

knownvariables

knownasteroids

SDSS2MASS

PQ Event Factory

remote archives

baselinesky

eStar

VOEventNet

GRBsatellites

VOEventdatabase

Palomar 60”

Architecture

VOEvent II6 December 2005

Palomar-Quest Survey

Synoptic sky survey using the 48” Palomar Samuel Oschin Schmidt telescope and the 112-CCD, 161-Megapixel Quest II camera

Collaboration between Caltech, Yale/Indiana U., NCSA, and JPL; et al. VO compliance/standards built in from the start Two modes: drift scan with UBRI/rizz or multiple repeated snapshots in

one filter ~70GB of data/night 15000 deg2 observed a minimum of 8 times with baselines minutes to

years

VOEvent II6 December 2005

A 152 ft 20 ft mural produced for Griffith Observatory from PQ survey BRI images

A swath of 15.2 2.0 swath through the center of the Virgo cluster, sampled at 0.4 arcsec/pixel, giving a 136,800 18,000 pixel image

Computed at CACR using HyperAtlas and a custom data cleaning pipeline

Reproduced on 114 steel-backed porcellain plates, expected to last many decades

Will be seen by millions of visitors Associated website will include NVO outreach

Real PQ data: The Big Picture

VOEvent II6 December 2005

The Big Picture: detailThe Big Picture: detail

VOEvent II6 December 2005

The Big Picture:Tile C12 (M87)Zoom-in

The Big Picture: more detail

VOEvent II6 December 2005

Transients in the Big Picture

Tile b07

740 Cantabia

VOEvent II6 December 2005

Event Synthesis Engine

Pairitel

Raptor

PQ next-daypipelines

catalog

Palomar-Quest

knownvariables

knownasteroids

SDSS2MASS

PQ Event Factory

remote archives

baselinesky

eStar

VOEventNet

GRBsatellites

VOEventdatabase

Palomar 60”

Architecture

VOEvent II6 December 2005

Palomar-Quest Event Factory Real-time pipeline to process raw data

streaming from telescope: Remove detector signatures including glitches

masquerading as transient events: meteors, airplanes, glints from satellites and junk, etc.

Apply basic photometric and astrometric calibration

Extract detected sources and measure attributes Compare with baseline data (catalogs/images) to

identify new, transient or highly variable sources Compare with dbs of known variables, asteroids,

etc.

VOEvent II6 December 2005

Event Synthesis Engine

Pairitel

Raptor

PQ next-daypipelines

catalog

Palomar-Quest

knownvariables

knownasteroids

SDSS2MASS

PQ Event Factory

remote archives

baselinesky

eStar

VOEventNet

GRBsatellites

VOEventdatabase

Palomar 60”

Architecture

VOEvent II6 December 2005

Event Synthesis Engine New input arrives from PQ/elsewhere:

Establish event “portfolio” to archive and federate all subsequent data and analysis

Send initial event notification to subscribers Launch query against external dbs via NVO

Classify and prioritize: Evaluate likelihood probabilities of event being

associated with possible astrophysical sources using machine learning techniques (‘Thinking Telescope’)

Evaluate urgency of desired follow-up Send out VOEvent

VOEvent II6 December 2005

Event Synthesis Engine

Pairitel

Raptor

PQ next-daypipelines

catalog

Palomar-Quest

knownvariables

knownasteroids

SDSS2MASS

PQ Event Factory

remote archives

baselinesky

eStar

VOEventNet

GRBsatellites

VOEventdatabase

Palomar 60”

Architecture

VOEvent II6 December 2005

VOEventNet Communication Fabric Author Publisher (aggregator):

Stores packet and assigns identifier Distributes to subscribers based on pre-defined

criteria using one-way web services Two dbs - Caltech for PQ and Los Alamos for

Raptor - harvest each other Subscriber:

Gets event from publisher, evaluates it and causes scheduling in telescope observing queue or an archive search

VOEvent II6 December 2005

Event Cycling An event can be injected back into the

same decision/classification engine that published it but supplemented with data from elsewhere

Events dynamically cycle through follow-up observation and computation (no humans in loop) with subscribers make judgments and adding value until convergence

VOEvent II6 December 2005

Robotic Telescopes RAPTOR

Stereoscopic sky monitoring with follow-up ‘fovea’ telescope

PAIRITEL Meter-class IR follow-up

P60 Principal follow-up facility for PQ Events

eSTAR

VOEvent II6 December 2005

Timeline Year one - proof-of-concept system

consisting of: Source of VOEvents A VOEvent store Basic event discriminator Robotic telescopic capable of responding

to VOEvent: P60 and Pairitel Year two - prototype system Year three - production system