The Dynamic Radio Sky and Future Instruments

Jim Cordes Cornell University

AAS MeetingNashville

28 May 2003

Dynamic Radio Sky

• We know enough about the DRS to know that there is a great deal yet to be discovered

• c.f. the high energy universe, optical, etc.• What is in the DRS?• What are the prospects for new discoveries?

Astrophysical parameters Extrinsic effects RFI

• Instruments & surveys that will reveal the DRS

TRANSIENT SOURCES TRANSIENT SOURCES Sky Surveys:Sky Surveys:The X-and-The X-and--ray skies have been monitored highly successfully -ray skies have been monitored highly successfully with wide FOV detectors with wide FOV detectors (e.g. RXTE/ASM, CGRO/BATSE).

Neutrino/gravitational wave detectors are ‘all sky.’Neutrino/gravitational wave detectors are ‘all sky.’

Optical transient surveys (ROTSE, RAPTOR, LSST) are/will Optical transient surveys (ROTSE, RAPTOR, LSST) are/will revolutionalize our knowledge of the optical transient sky and revolutionalize our knowledge of the optical transient sky and will drive the trend toward data mining of will drive the trend toward data mining of » » petabyte databases.petabyte databases.

The transient radio sky (e.g. t < 1 month) is largely unexplored.The transient radio sky (e.g. t < 1 month) is largely unexplored.

New objects/phenomena are likely to be discovered as well as New objects/phenomena are likely to be discovered as well as extreme cases in predictable classes of objects.extreme cases in predictable classes of objects.

Ingredients for transient detection

AT needs to be “large”

A = collecting area

= solid angle covered (instantaneous FOV)

T = time per sky position

Issue: to dwell (stare), or tile the sky, or be triggered?

Successes in transient astronomy

RXTE/ASM

VelaROTSE

RAPTOR

Why is the Dynamic Radio Sky Largely Uncharted?

• Large collecting areas, A, needed for sensitivity• Typically A is small enough that telescope

throughput is small• Telescope time is expensive so dwell times are short• Sources cover a wide range of time scale and sky

density

insufficient sky and temporal coverage

Giant pulse from the Crab pulsar

S ~ 160 x Crab Nebula

~ 200 kJy

Detectable to ~ 1.5 Mpc with Arecibo

Arecibo

2-ns giant pulses from the Crab: (Hankins et al. 2003)

Giant Pulses seen from B0540-69 in LMC (Johnston & Romani 2003)

Giant pulses are the fastest known transients

• Giant pulses from Crab detectable to ~1.5 Mpc with Arecibo @ 1/hour

• 2-ns wide `nano-Giant pulses’ identified from Crab (Hankins et al. 2003)

• GPs seen from Crab clone in LMC (B0540-69) by Johnston & Romani (2003) w/ similar intrinsic amplitude

• GPs from two millisecond pulsars• Radio GPs in pulse components also

seen in X-rays• GP-emitting objects have ~ same B

fields at their light cylinders

Nano-giant pulses (Hankins et al. 2003)

Arecibo

5 GHz

0.5 GHz bw

coherent dedispersion

STARE 611 MHz 3-station radio transient detector (Katz, Hewitt, Corey, Moore 2003)

Solar Radio Bursts

GRB 980519 variability (Frail et al. 2000)

Interstellar scintillations

TRANSIENT SOURCES TRANSIENT SOURCES TARGET OBJECTS: TARGET OBJECTS:

• Atmospheric/lunar pulses from neutrinos & cosmic rays

• Accretion disk transients (NS, blackholes)• Neutron star magnetospheres• Supernovae• Gamma-ray burst sources• Brown dwarf flares (astro-ph/0102301)• Planetary magnetospheres & atmospheres• Maser spikes• ETI

TRANSIENT SOURCES TRANSIENT SOURCES TARGET PROCESSES:TARGET PROCESSES:

• Intrinsic:incoherent:

( inverse Compton brightness limit)coherent: (virtually no limit)continuum: low frequencies favored

spectral line: masers

• Extrinsic:scintillationmaser-maser amplificationgravitational lensingabsorption events

Phase Space for Transients: SpkD2 vs. W

W

W

Pulse

Process

Spk

log

Sp

kD

2

log W

W = light travel time

brightness temperature:

SpkD2

Tb = ------------- 2k (W)2

Phase Space for Transients: SpkD2 vs. W

W

W

Pulse

Process

Spk

log

Sp

kD

2

log W

Lines of constant brightness temperature

Phase Space for Transients: SpkD2 vs. W

W

W

Pulse

Process

Spk

log

Sp

kD

2

log W

Solar system

+

local galactic sources

Phase Space for Transients: SpkD2 vs. W

W

W

Pulse

Process

Spk

log

Sp

kD

2

log W

OH masers

+

Pulsars (including giant pulses)

Phase Space for Transients: SpkD2 vs. W

W

W

Pulse

Process

Spk

log

Sp

kD

2

log W

Cosmological sources:

AGNs (including IDV sources)

+

GRB afterglows

Phase Space for Transients: SpkD2 vs. W

W

W

Pulse

Process

Spk

log

Sp

kD

2

log W

Phase Space for Transients: SpkD2 vs. W

W

W

Pulse

Process

Spk

log

Sp

kD

2

log W

Interstellar scintillations = apparent fast variations of IDVs & GRBs

New instruments can cover this phase space

W

W

Pulse

Process

Spk

log

Sp

kD

2

log W

Exploring the Transient Radio Sky:Striving for large AT

• Pilot observations: Arecibo: single pixel and multibeam (ALFA) STARE and similar multisite arrays GBT: single pixel and multibeam arrays ATA: 2.5 deg FOV, ~8 array beams EVLA (wideband, high sensitivity & spatial resolution)

• LOFAR: low frequencies (< 240 MHz)• SKA: broad frequency range (0.15 to 25 GHz)

Giant pulses from M33Arecibo observations

(Maura Mclaughlin & Cordes, submitted to ApJ, astro-ph

Galactic Center TransientsVLA 0.33 GHz

Hyman et al. 2002

Exploring the Transient Radio Sky:Covering the Sky

• Staring vs. mosaicing (tiling)?• Radio sky needs both;

Fast transients: too fast to raster scan the sky (< hours to months) (e.g. GPs)

Slower transients: raster scan (e.g. for objects showing radio only) trigger from other wide-field instruments (GRB

afterglows)

TRANSIENT SOURCES TRANSIENT SOURCES Sure detectionsSure detections::

• Analogs to giant pulses from the Crab pulsar out to ~5 – 10 Mpc

• Flares from brown dwarfs out to at least 100 pc.

• GRB afterglows to 1 µJy in 10 hours at 10 .

PossibilitiesPossibilities::

• -ray quiet bursts and afterglows?-ray quiet bursts and afterglows?

• Intermittent ETI signals?Intermittent ETI signals?

• Planetary flares?Planetary flares?

Isolated pulsarRe-ignition of pulsar action in mergers?

Hansen & Lyutikov 2000

RFI Editing in the f-t planeRFI dynamic spectra (from AO monitoring program)

Dynamic spectrum of pulsar scintillation

Working Around Radio Frequency Interference

• Single-dish/single-pixel transient detection:• Very difficult to separate terrestrial & astrophysical

transients (significant overlap in signal parameter space)

• Multiple beam systems (Parkes, Arecibo, the GBT):• Simultaneous on/offs partial discrimination

Multiple site systems (a la LIGO, PHOENIX)• Very powerful filtering of RFI that is site specific or

delayed or Doppler shifted between sites

LOFAR = Low Frequency Array

Stations of dipoles

30 to 240 MHz

Large AT

Optimal for coherent continuum transients

SKA = Square Kilometer Array SKA = Square Kilometer Array

Current ConceptsCurrent Concepts

China KARSTChina KARST

Canadian Canadian aerostataerostat

US Large NUS Large N

Australian Australian Luneburg Luneburg LensesLenses

Dutch fixed Dutch fixed planar arrayplanar array

(cf. Allen Telescope Array,

Extended VLA)

(cf. LOFAR = Low Freqency Array)

Current Baseline SpecificationsParameter Design Goal Comments Sensitivity A/T = 2 x 104 m2 / K 20x Arecibo, 75x VLA Surface brightness 1K at 0.1 arcsec (cont) Point sources 0.5 Jy 10 in 1 day, 100 MHz Frequency range 0.15 – 22 GHz Redshift coverage Z < 8.5 HI,

Z > 4.2 CO (1 0)

L* galaxies Zmax ~ 2 HI, ~ 20 CO FOV (imaging) 1 degree2 at 1.4 GHz Multibeams > 100 Ang. Resolution 40 mas at 1.4 GHz VLBI: SKA enables all-sky

phase referencing Pixels 108 Instantaneous bandwidth

20% at high frequencies

Spectra channels 104 Image Dynamic Range

106 at 1.4 GHz

Polarization isolation

-40 dB

Methods with LOFAR & SKAMethods with LOFAR & SKA

I. Target individual objects

II. Blind Surveys: trade FOV against gain by multiplexing SKA into subarrays.

III. Allow rapid response to triggers

IV. Exploit coincidence tests to ferret out RFI, use multiple beams.

Primary beam & station Primary beam & station synthesized beamssynthesized beams

Station subarrays for Station subarrays for larger FOVlarger FOV

One station of many in SKAOne station of many in SKA

Blind Surveys with SKA

• Number of pixels needed to cover FOV:Npix~(bmax/D)2 ~104-109

• Number of operationsNops~ petaops/s

• Post processing per beam:e.g. standard pulsar periodicity analysis

Summary Transient science is unexplored territory for radio astronomy:

New looks at known sources Entirely new classes of sources: LOFAR will survey transients at f < 240 MHz; SKA for 0.15 GHz < f < 25 GHz (or more)

Implications for SKA design: Rapid imaging/mosaicing of sky (days) Large instantaneous FOV desired for short time scales (e.g.

hemispheric). US Plan: Subarrays to allow coincidence tests and maximal

sky coverage. Versatile imaging/beamforming/signal processing modes.

Similar implications for pulsar science

Radio Pulsars

• ~1400 known (doubled by Parkes MB)• ~100 millisecond pulsars• 2 to 3 with planets

• ~5 NS-NS binaries (Porb > 8 hr)

• MSPs have exceedingly stable spins, suitable for seeking gravitational wave perturbations

Why more pulsars?• Extreme Pulsars:

• P < 1 ms P > 8 sec

• Porb < hours B > 1013 G (link to magnetars?)

• V > 1000 km s-1

• NS-NS & NS-BH binaries

• Population & Stellar Evolution Issues

• The high-energy connection (e.g. GLAST)

• Physics payoff (GR, Gwaves, EOS, LIGO, GRBs…)

• Serendipity (strange stars, transient sources)

• New instruments (AO, GBT, LOFAR, SKA) will dramatically increase the volume searched (galactic & extragalactic)

Parkes MB Feeds

Parkes MB Feeds

ALFA Science Goals: Massive Surveys

Drift scan surveys (14 sec across 3.5 arcmin)

Deep Galactic plane survey (GPS) (5-10min, |b| < 5 deg, 30 < l < 80 + anticenter)

Medium latitude surveys ( 5 < |b| < 25 deg)

Targeted: globular clusters, high EM/DM HII regions, SNRs, Galactic chimneys, M33, X/ -ray selected objects

(long dwell times, up to 2.5 hr)

Surveys Surveys with Parkes, with Parkes, Arecibo & Arecibo & GBT.GBT.

Simulated & Simulated & actualactual

Yield ~ 1000 Yield ~ 1000 pulsars.pulsars.

ALFA Surveys at Arecibo• ALFA surveys can be viewed as part of a long-

term, grander effort (“Full Galactic Census”) (LOFAR, SKA, )

• RFI mitigation required and provides general purpose tools

• Data & data products = long term resources data management policy & resources~ 1 petabyte of survey raw data~ 1 petabyte of data products

• Exploit telescope time fully (transients, piggybacking)

SKA pulsar SKA pulsar surveysurvey

600 s per 600 s per beambeam

~10~1044 psr’s psr’s