Multi epoch variability of αox in the CDFS

Transcription

1 Multi epoch variability of αox in the CDFS Fausto Vagnetti most results by: Marco Antonucci + Dario Trevese, Maurizio Paolillo,... XMM CDFS Team meeting - Cervia, 2012 May 31- June 1

2 outline present work on the XMM CDF-S data is part of a research on the variability of αox with analyses of more samples the αox-luv relation dispersion and variability samples with simultaneous X-UV measurements XMMSSC+XMMOMSUSS Grupe /Swift XMM deep survey in the CDF-S archival data - epochs - sources light-curves tracks in the αox-luv plane structure functions

3 the αox-luv relation UV studied by many authors, e.g. Avni & Tananbaum 1986,Vignali et al 2003, Strateva et al 2005, Steffen et al 2006, Just et al 2007, Lusso et al 2010, Young et al 2010, Grupe et al 2010 etc α ox = log[l ν(2kev)/l ν (2500Å)] log[ν(2kev)/ν(2500å)] =0.384 log L X L UV X Just et al 2007 more luminous objects are relatively weaker in X-rays αox = ( ± 0.007) log LUV + (2.705 ± 0.212) BUT: Gibson Brandt & Schneider 2008: large dispersion, possibly due to variability and/or non-simultaneity of X and UV

4 samples to analyse αox variability wish to have samples with multi-epoch, simultaneous X/UV measurements

5 samples to analyse αox variability wish to have samples with multi-epoch, simultaneous X/UV measurements XMM serendipitous source catalogs Vagnetti et al 2010, 192 radioquiet non-bal sources, 41 multiepoch (2-9 epochs) Nepo

6 samples to analyse αox variability wish to have samples with multi-epoch, simultaneous X/UV measurements XMM serendipitous source catalogs Vagnetti et al 2010, 192 radioquiet non-bal sources, 41 multiepoch (2-9 epochs) Swift sample by Grupe et al low z sources re-analysed by us, 67 multi-epoch (2-8 epochs), in progress Nepo

7 samples to analyse αox variability wish to have samples with multi-epoch, simultaneous X/UV measurements XMM serendipitous source catalogs Vagnetti et al 2010, 192 radioquiet non-bal sources, 41 multiepoch (2-9 epochs) Swift sample by Grupe et al low z sources re-analysed by us, 67 multi-epoch (2-8 epochs), in progress XMM deep survey in the CDF-S 20 multi-epoch sources (2-33 epochs), in progress Nepo

8 samples to analyse αox variability wish to have samples with multi-epoch, simultaneous X/UV measurements XMM serendipitous source catalogs Vagnetti et al 2010, 192 radioquiet non-bal sources, 41 multiepoch (2-9 epochs) Swift sample by Grupe et al low z sources re-analysed by us, 67 multi-epoch (2-8 epochs), in progress XMM deep survey in the CDF-S 20 multi-epoch sources (2-33 epochs), in progress multi-epoch survey of the same sky area ideal to get detailed information on individual sources Nepo

9 samples in the L-z plane

10 samples in the L-z plane 1 Vagnetti et al

11 samples in the L-z plane 1 Vagnetti et al Grupe et al

12 samples in the L-z plane CDF-S 1 Vagnetti et al Grupe et al

13 XMM serendipitous source catalogs Vagnetti Turriziani Trevese Antonucci 2010 XMMSSC XMMOMSUSS slope SDSS DR5 Quasar Cat 241 sources 315 observations in total 192 radio-quiet non-bal sources 41 with multiple epochs radio loud AGNS and BAL QSOs are removed slope slope the slope of the correlation can be compared with other authors we are more interested in the dispersion slope 0.217

14 dispersion and variability of αox dispersion σ=0.12: similar to previous non-simultaneous analyses artificial αox variability (non-simultaneity): not important intrinsic αox variability (true change of αox): important its contribution can be called intra-source dispersion another important contribution is due to intrinsic differences in the average αox values from source to source, which we call inter-source dispersion how much intra-source and how much inter-source?

15 intra-source and inter-source dispersion variations are on different time scales, to compare use Structure Function : SF(τ) = π 2 α ox(t + τ) α ox (t) SF is increasing, both for αox and for the residuals Δαox greatest change at 1 yr, σintra-source ~ 0.07 Δαox=αox αox(luv) σ 2 tot =σ 2 intra-source+σ 2 inter-source σ 2 intra-source ~30% σ 2 tot x comparable contributions: αox L logluv

16 Swift sample by Grupe et al low redshift sources (z<0.35) 74 multi-epoch, 67 radio-quiet analysis in progress

17 Swift sample by Grupe et al low redshift sources (z<0.35) 74 multi-epoch, 67 radio-quiet analysis in progress tracks in the αox-luv plane are not much different from serendipitous sample

18 Swift sample by Grupe et al low redshift sources (z<0.35) 74 multi-epoch, 67 radio-quiet analysis in progress tracks in the αox-luv plane are not much different from serendipitous sample the SF of αox is also consistent with the serendipitous sample, with σintra-source ~

19 XMM-Newton deep survey in the CDFS preliminary analysis of XMM- Newton archive on HEASARC also in the same field 8 epochs PI Bergeron epochs PI Comastri great advantage: many epochs 8+25=33 XMM-Newton Master Log & Public Archive obsid status name ra dec time duration pi lname pi fname public date data in heasarc Search Offset archived AXAF Ultra Deep F :03: Jacqueline Bergeron Y (CDFS) archived AXAF Ultra Deep F :48: Jacqueline Bergeron Y (CDFS) archived AXAF Ultra Deep F :14: Jacqueline Bergeron Y (CDFS) archived AXAF Ultra Deep F :08: Jacqueline Bergeron Y (CDFS) archived AXAF Ultra Deep F :38: Jacqueline Bergeron Y (CDFS) archived AXAF Ultra Deep F :23: Jacqueline Bergeron Y (CDFS) archived AXAF Ultra Deep F :26: Jacqueline Bergeron Y (CDFS) archived AXAF Ultra Deep F :24: Jacqueline Bergeron Y (CDFS) archived AXAF Ultra Deep F :03: Jacqueline Bergeron Y (CDFS) archived AXAF Ultra Deep F :42: Jacqueline Bergeron Y (CDFS) archived AXAF Ultra Deep F :47: Jacqueline Bergeron Y (CDFS) archived AXAF Ultra Deep F :48: Jacqueline Bergeron Y (CDFS) archived CDFS :27:51 Comastri Andrea Y (CDFS) archived CDFS :54: Comastri Andrea Y (CDFS) archived CDFS :44: Comastri Andrea Y (CDFS) archived CDFS :02: Comastri Andrea Y (CDFS) archived CDFS :13: Comastri Andrea Y (CDFS) archived CDFS :57: Comastri Andrea Y (CDFS) archived CDFS :45: Comastri Andrea Y (CDFS) archived CDFS :24: Comastri Andrea Y (CDFS) archived CDFS :21: Comastri Andrea Y (CDFS) archived CDFS :11: Comastri Andrea Y (CDFS) archived CDFS :39: Comastri Andrea Y (CDFS) archived CDFS :48: Comastri Andrea Y (CDFS) archived CDFS :13: Comastri Andrea Y (CDFS) archived CDFS :16: Comastri Andrea Y (CDFS) archived CDFS :59: Comastri Andrea Y (CDFS) archived CDFS :27: Comastri Andrea Y (CDFS) archived CDFS :37: Comastri Andrea Y (CDFS) archived CDFS :12: Comastri Andrea Y (CDFS) archived CDFS :59: Comastri Andrea Y (CDFS) archived CDFS :50: Comastri Andrea Y (CDFS) archived CDFS :39: Comastri Andrea Y (CDFS) archived CDFS :58: Comastri Andrea Y (CDFS) archived CDFS :34: Comastri Andrea Y (CDFS) archived CDFS :36: Comastri Andrea Y (CDFS) archived CDFS :07: Comastri Andrea Y (CDFS) archived CDFS :07: Comastri Andrea Y (CDFS) archived CDFS :36: Comastri Andrea Y (CDFS) archived CDFS :55: Comastri Andrea Y (CDFS) archived CDFS :59: Comastri Andrea Y (CDFS) archived CDFS :44: Comastri Andrea Y (CDFS) archived CDFS :22: Comastri Andrea Y (CDFS) archived CDFS :01: Comastri Andrea Y (CDFS) archived CDFS :29: Comastri Andrea Y (CDFS) archived CDFS :00: Comastri Andrea Y (CDFS) archived CDFS :40: Comastri Andrea Y (CDFS) archived CDFS :22: Comastri Andrea Y (CDFS)

20 exposure times most epochs have X-ray exposures above 100 ks but the O/UV exposures are shorter, divided among the 6 OM filters and the adopted filters vary from epoch to epoch UVW1 and U have most often the longest exposures Xray,ks UVW2 UVM2 UVW1 U BV n obsid time

21 exposure times most epochs have X-ray exposures above 100 ks but the O/UV exposures are shorter, divided among the 6 OM filters and the adopted filters vary from epoch to epoch UVW1 and U have most often the longest exposures exposure (ks) Bergeron Xray,ks UVW2 UVM2 UVW1 U BV n obsid time X-ray UVW UVM UVW U B V time Comastri

22 source numbers as a result, the number of detected sources varies from epoch to epoch, both in X-ray and O/UV we compare source coordinates with tables of known redshifts, and search for sources with simultaneous X-ray/UV detections hundreds of sources are detected in UV and not in X-rays, mostly galaxies several tens of sources are detected in X-rays and not in UV, mostly AGNs few sources, up to 20, are detected at each epoch in both X-ray and UV X-ray O X+O detected sources time

23 sources this corresponds to 65 source with simultaneous X-ray/UV measurements for a number of epochs between 1 and unclassified sources 5 possible/uncertain AGNs 13 single-epoch AGNs 20 multi-epoch AGNs # i z Nepo class X-class RAJ2000 DEJ2000 RA DEC Src_num HEX AGN BLAGN QSO BLAGN QSO UAGN BLAGN QSO LEX AGN BLAGN AGN BLAGN AGN BLAGN QSO BLAGN AGN BLAGN AGN BLAGN AGN UAGN BLAGN AGN HEX AGN LEX QSO BLAGN QSO BLAGN QSO BLAGN QSO BLAGN QSO

24 200

25 203

26 228

27 241

28 249

29 276

30 289

31 319

32 328

33 337

34 tracks in the plane αox-luv detailed information on individual variability

35 tracks in the plane αox-luv detailed information on individual variability but with large errors

36 tracks in the plane αox-luv detailed information on individual variability but with large errors comparison with previous samples

37 structure functions individual SF: a few cases where we are able to subtract variability due to photometric errors (brightest sources) 1 SF(τ) = π 2 α ox(t + τ) α ox (t) 2 σ 2 n in most cases the error is larger than the measure the same occurs for the ensemble SF

38 spectral variability parameter describes changes of spectral index divided by changes of luminosity introduced by Trevese & Vagnetti 2002 in the optical β(τ) = β>0: harder when brighter can be used also in αox-luv relation: can be estimated by average slope of αox-luv relation for individual source α(t + τ) α(t) log L(t + τ) log L(t) β<0: softer when brighter β ox = α ox log L UV however, many correlations have slopes approaching -0.38, because αox is intrinsically correlated with LUV by the definition αox=0.38log (LX/LUV). if LX stays constant and only LUV changes, slope will be better evaluate LX-LUV correlations best cases (Pnull<0.03) are the sources: 237 (βox= 2.3±0.7) 228 (βox=-0.16±0.09) 319 (βox=-0.46±0.05) most sources have negative βox. softer when brighter means that optical varies more than X-ray source 237 shows an opposite behavior

39 next steps try to decrease the errors get optimised X-ray fluxes from the collaboration check OM images improve ensemble and individual structure functions improve evaluation of βox with errors try X-ray/UV cross correlations to determine lags

40 backup slides disk-corona relation CIV blueshift 3 bright sources in L-z plane Nepo histograms cross-correlations optical and X-ray images

41 X-ray/optical interplay α ox 0.38 log L X L UV if αox changes, then X and UV must vary differently, one more than the other serendipitous sample larger variations at long time scales, according to SF expected to be driven by optical fluctuations 0.1 Grupe et al irradiation 0.1 inverse Compton propagating fluctuations (Lyubarskii 1997, Arevalo 2006) CDFS

42 CIV Equivalent Width and blueshift αox is anti-correlated with CIV blueshift and correlated with CIV Equivalent Width (Richards et al 2011, Kruczek et al 2011)

43 L-z plane

44 Nepo histograms

45 β, r and P(>r) beta p r

46 cross-correlation with quite large number of epochs, it could be possible to try cross-correlations to determine time-lags between X-ray and optical methods to estimate probabilities within the gaps could be applied (Zu et al 2011)

47 optical image, VEIS

48 X-ray image, 4Ms Chandra