Update on TKSP imaging results

Update on Transient KSP Imaging Results Alexander van der Horst
University of Amsterdam Jess Broderick University of Southampton Martin Bell, Valeriu Tudose, Dan Calvelo, Dan Plant, Richard Plotkin, Dario Carbone, Adam Stewart, Gosia Pietka, Antonia Rowlinson, John Swinbank, Evert Rol, Bart Scheers, Jason Hessels, James Miller-Jones, Rob Fender, Ralph Wijers and the LOFAR Transients KSP

LOFAR imaging pipeline

TKSP data overview • PSR B0329+54 (MarCn Bell/Jess Broderick) •
PKS 1510-‐089 (Adam Stewart/Gosia Pietka/Jess Broderick) • 3C 454.3 (Jess Broderick) • M101 (Antonia Rowlinson) • SS433 (Jess Broderick) • GRS 1915+105 (Richard Plotkin/Dario Carbone) • Cygnus X-‐1 (Dan Calvelo/Dan Plant) • Cygnus X-‐3 (Valeriu Tudose)

PSR B0329+54 ﬁeld • 11 x 12 hour and 8
x 6 hour observaCons • April 2010 – November 2011 • 7 – 20 staCons • Field of view: ~20 deg2 • Noise: 4-‐15 mJy/beam • 2 transient candidates: – 1 turns out to be artefact at edge of image (Cimager performed poorly at image edge) – 1 with higher conﬁdence, but what is it?

Transient candidate All observaCons Aug – Oct 2010 ObservaCons Dec
2010 WENSS 325 MHz

Transient light curve • Non-‐detecCons in Oct/Nov; noise: 10
mJy/beam • No radio, opCcal, X-‐ray or gamma-‐ray counterpart so far • WSRT observaCons at 1.4 GHz and 325 MHz scheduled: December 16-‐18 with simultaneous LOFAR observaCons

PKS 1510-‐089 • Blazar with record ﬂux levels in
gamma rays (Fermi LAT) and cm/mm radio (Eﬀelsberg/IRAM) on October 20 & 21 (Atel 3694, 3698) • 11 HBA LOFAR observaCons (Oct 23 – Nov 11): – TesCng LOFAR with variable bright sources – TesCng LOFAR with a low DeclinaCon source (-‐9) – Core and remote staCons; last 7 also internaConal • Sky model: gsmuCls.py script, ~600 sources including PKS 1510-‐089

• Only core staCons used – good ﬁrst calibraCon since
VLSS/NVSS sky model maybe not adequate for longest baselines • Standard cookbook parameters with few tweaks • One good image from one observaCon • Averaged from 21 sub-‐ bands • Post BBS ﬂagging of 25% – to be reduced! • Flux: 4.4 Jy • Noise: 138 mJy/beam PKS 1510-‐089: current status

One sub-‐band of nice averaged image Best of the
worst: One sub-‐band of other obs. Results can vary wildly with same parameters!

• Only one observaCon so far with a good final
image • InvesCgaCons: – Effects of relaCvely low DeclinaCon (-‐9 deg) – Sun only 20 degrees away – needed in sky model? – InternaConal baselines affecCng core data? – CalibraCon poor with very high amplitude spikes post BBS • ConCnued observaCons, hopefully recording flux peak PKS 1510-‐089: current & future work

3C 454.3 • Blazar with large flares from high-‐energy gamma
rays down to low radio frequencies • LOFAR 6 hour HBA test observaCons in August • Sky model: gsmuCls.py script, 3C 454.3 removed • First image with 6 sub-‐bands: 119.1-‐152.3 MHz • Remote staCons flagged • Individual images combined in image plane • ResoluCon 195x168 arcsec • StaCon beam correcCon at center of field

3C 454.3: ﬁrst image • 3C 454.3 detected: 3.4 Jy/beam
• Noise: 70 mJy/beam; calibraCon needs to be checked

SN2011fe in M101 • Youngest Type Ia SN ever
recorded • Most nearby Type Ia SN in decades (6.4 Mpc) • No detecCon at high radio frequencies • LOFAR 4 hour HBA observaCon • 4 days aoer discovery (Aug 28) • Sky model from imaging pipeline used WSRT at 1.4 GHz (Horesh et al. 2011)

SN2011fe in M101: careful cleaning Before …
… and aoer

SN2011fe in M101: zoom in Before …
… and aoer

SN2011fe in M101: comparison LOFAR HBA NVSS (1.4 GHz) •
Non-‐detecCon not unexpected based on higher frequency observaCons and modeling • Good exercise in cleaning procedures – to be improved!

SS433 / W50 • Famous GalacCc microquasar at 5.5 kpc
• Mildly-‐relaCvisCc precessing jets from SS433 inﬂate structures in supernova remnant W50 VLA 1.4 GHz (Dubner et al. 1998) VLA 74 MHz (Miller-‐Jones et al. 2007)

SS433 / W50: new LOFAR observaCons • Three 6 hour
HBA observaCons: – Jun 25: 19 core + 7 Dutch remote – Aug 16: 18 core + 8 Dutch remote – Sep 11: 20 core + 8 Dutch remote • Sky model: PyBDSM & shapelets on 74 MHz VLA map • SS433 not included in sky model • Shortest baselines (< 50 m) excluded from BBS because of unsaCsfactory W50 model – not flagged in imaging • Dutch remote staCons flagged completely because of unsaCsfactory model on long baselines • Beam correcCon applied at field center

SS433 / W50: June observaCons • 59 sub-‐bands, 115.0-‐148.2 MHz;
280x150 arcsec • Noise: 15 mJy/beam; SS433: 0.75 +/-‐ 0.15 Jy

SS433 / W50: August observaCons • 103 sub-‐bands, 115.0-‐152.1 MHz;

SS433 / W50: September observaCons • 145 sub-‐bands, 115.0-‐144.7 MHz;

SS433 / W50: results & future work • CalibraCon seems
accurate to 10-‐20% • Marginal detecCon of variability? • No obvious transients when comparing observaCons • Future work: – Sky model improvements – ConCnued observaCons – Re-‐analysis of April HBA data (2 x 2 hours) – Analyze LBA data – Image data with AWimager and use self-‐calibraCon

SS433 / W50: ﬁrst AWimager results • June observaCons •
Core staCons only • Noise: 250 mJy/beam • 500x180 arcsec • Noise in individual sub-‐bands factor 2 worse than in CASA

GRS 1915+105 • Famous GalacCc microquasar • Complicated ﬁeld with
H II regions and SN remnants • LOFAR 6 hour HBA observaCon • Standard pipeline and cookbook parameters • First image with 15 sub-‐bands • Individual sub-‐band images combined in image plane

GRS 1915+105: comparison LOFAR HBA WSRT at 325 MHz (Miller-‐Jones
et al. 2007)

GRS 1915+105: future work • Explore diﬀerent ﬂagging cuts •
Improve the sky model: – Now only VLSS (74 MHz), spectral index ~0.7 assumed – Near future : include 325 MHz image (James Miller-‐Jones) • Use all 256 sub-‐bands; close to possible detecCon?

Cygnus X-‐1 • LOFAR 6 hour HBA observaCon • 49
sub-‐bands, average at 128 MHz • Images for each sub-‐band combined in image plane • Remote staCons not included • ResoluCon: 3x6 arcmin • Sky model: – PyBDSM on a WENNS mosaic: 1050 model sources – Cyg A model based on VLA image at 325 MHz (3 arcsec) • Noise: 12-‐20 mJy, depending on closeness to bright sources (sidelobe related artefacts)

Cygnus X-‐1: wide-‐ﬁeld image LOFAR HBA Sky model

Cygnus X-‐1: zoom in LOFAR HBA Sky model

H II region close to Cygnus X-‐1 LOFAR contours on
opCcal map (courtesy of David Russell) LOFAR contours (60,90 mJy) on WENNS 325 MHz (20,30,40,60 mJy)

Cygnux X-‐3 • Image of one sub-‐band looks promising •
Sky model very important: Cyg A at 7 deg, W66 at 5 deg

Cygnus X-‐3: ﬁrst results • Noise: 0.1 Jy/beam • Possible
0.2 Jy source, 2 arcmin from Cyg X-‐3 posiCon

Imaging challenges • Find best way to generate iniCal sky
model • Start with core staCons for iniCal model, then remote ones with self calibraCon? • ContribuCon of A-‐team sources and demixing • Take direcCon dependent eﬀects into account • Try global solver to improve calibraCon (mulCple sub-‐ bands)

Flagger tesCng • Make sure that transients are not mistaken
for RFI • Add model transients to real data to test the AOflagger • Example: 2 Jy source on for half hour in 6 hour observaCon of PSR B0329+54  flagged • Test the limits of pipeline parameters in terms of amplitude and duraCon of sources • InvesCgate various parameters to minimize unwanted flagging of bright transients • SoluCon: separate pipelines for transient searches?

Summary & future work • Good progress being made with
the various ﬁelds • Several new commissioners have been trained • TKSP input in general commissioning important • AWimager tesCng: – TKSP-‐speciﬁc tests – Imaging Busy Week (Jan 30 – Feb 3) • TesCng of source extractors (PYSE / PyBDSM) • Quality control metrics for trip/trap + sky model tesCng • MSSS preparaCon • Imaging team increased, but volunteers welcome!

Update on TKSP imaging results

Update on TKSP imaging results

More Decks by transientskp

Other Decks in Science

Featured

Transcript