The MIXR sample: Radio-loud AGN through the eyes of 3XMM, WISE, and FIRST/NVSS Beatriz Mingo University of Leicester ARCHES Scientific Workshop Collaborators: Paris, Nov 30 – Dec 2 2015 Mike Watson, Gordon Stewart, Simon Rosen, Andrew Blain (Leicester), Martin Hardcastle (Hertfordshire),Angel Ruiz, Silvia Mateos, Francisco Carrera (IFCA)
Why radio-loud AGN? ● How relevant are radio-loud AGN? – RL ~10-20% of the total AGN population, but (presumably) all AGN have been RL – RL phase affects environment dramatically and is fundamental to understand accretion ● Many Seyferts with jets and lobes! (see e.g. Croston + 2008, Mingo + 2011, 2012) ● Advantages for AGN science: – Select dimmer AGN than optical/X-ray/IR cuts – Low sky density, clean selection – Only selects very nearby starburst galaxies Beatriz Mingo - Radio-loud AGN
The (LM)XRB link LERG/HERG =/= FRI/FRII FRII RG same jets in LERG/HERG rad efficient → jet can't be originated in classical accr. disk (HERG) AGN can develop jets while in this region, XRB, pulsars, WD, etc cannot, as they can't get enough magnetic flux accretion rate (Tchekhovskoy+ FRI RG 2011, Sikora & rad inefficient Begelman 2013) (ADAF/RIAF) (LERG) RQ rad efficient winds Beatriz Mingo - Radio-loud AGN
The Idea ● 3XMM + WISE + NVSS/FIRST – What does the WISE c/c plot look like? ● Data quality criteria: – S/N>5 in w1, w2, S/N>3 in w3 – No X-ray extended – “good” radio ● ARCHES X-matching Lake + 2012 code (FX Pineau's talk) Beatriz Mingo - Radio-loud AGN
The goals ● The MIXR sample: build the largest (to date) uniform Mid-IR + X-ray + Radio catalogue of AGN + star-forming galaxies. ● Establish diagnostics to quickly characterise sources even when no z available – Useful for big data (SKA/Athena/JWST/E-ELT...) ● Address science questions – Diagnose/disentangle SF & AGN activity? – Lbol vs Q: origin of the scatter? – How biased are MIR/X-ray selected samples? Beatriz Mingo - Radio-loud AGN
1) Collapse FIRST sources within 30” (Magliocchetti et al. 1998) 2) USE FIRST flux unless NVSS larger by 5 sigma It's complicated... 3) Exclude extended X-ray sources 4) Xmatch Beatriz Mingo - Radio-loud AGN
All that work to get just ~1400 sources 1008 AGN (577 z) W3 S/N>3 eliminates 45% of the sample 145 Ellipticals (94 z) SDSS z eliminates another 45% (still good result!) 222 Spirals (149 z) 174 Starburst (114 z) Beware: populations overlap! Beatriz Mingo - Radio-loud AGN
Diagnostics: Flux/mag triage Strong SF correlation, also tells us about “radio-shy” AGN Wide range of radio fluxes for given mid-IR magnitude → jet regulation? Need L to verify Beatriz Mingo - Radio-loud AGN
Diagnostics: Flux/mag triage SF correlation, messier than IR/radio, but still reliable AGN correlation: many “radio-shy” AGN from the previous plot are X-ray bright LERG/LINER have systematically lower X-ray fluxes Beatriz Mingo - Radio-loud AGN
Get some z's For each pop. The W3 rejected sources distrib. peaks at larger z → distance, not evolution Beatriz Mingo - Radio-loud AGN
Luminosity diagnostics (1) Radio/IR correlation for SF looks wonderful! LERG clearly separated from the rest AGN: wide range of radio L for a given mid-IR L → Mingo+ 2014 Beatriz Mingo - Radio-loud AGN
Luminosity diagnostics (2) SF, LERG/LINER and AGN populations clearly separated Wide range of radio L for given X-ray L Several Spirals harbouring AGN! Beatriz Mingo - Radio-loud AGN
RL AGN, RQ AGN and LERG/LINER RL AGN: 1.5 sigma above SF correlation in Lmid-IR/Lradio RQ AGN: Below 1.5 sigma in the SF correlation, but L2-10 keV >= 5x10^41 erg/s LERG/LINER: RL AGN with L2-10 keV < 5x10^41 erg/s Beatriz Mingo - Radio-loud AGN
Rawlings & Saunders 1991 = wrong? ...probably? CAUTION AGN When plotting sources with unclear radio emission! Timescales probably introduce a lot of scatter in this correlation (more than we suspected in Mingo et al. 2014) LERG/LINER certainly don't follow the same trend! Beatriz Mingo - Radio-loud AGN
Beware of the bias! ● Make sure to compare like with like! – Most MIR AGN samples ignore Seyferts – Many X-ray AGN samples ignore RL sources – More cat. = more diagnostics, but also more bias! Beatriz Mingo - Radio-loud AGN
MIXR Conclusions ● Diagnostic plots: useful when you don't have z! – SF radio/IR correlation is tight and fairly reliable ● Useful also to identify RQ AGN ● Lbol/Q does have large scatter – Still unclear why... if we can understand the timescales involved we can also shed light on the AGN vs star formation conundrum. ● Beware of the bias! – w1-w2>0.5 leaves out lots of interesting AGN – X-ray samples biased against RL sources – More data, please! :) Beatriz Mingo - Radio-loud AGN
Recommend
More recommend
