The warm absorber in NGC 5548

1
The warm absorber in NGC 5548

• Jelle Kaastra / Elisa Costantini
• SRON
• Katrien Steenbrugge
• CfA

2
Overview of the talk

1. Introduction to NGC 5548
2. Short description of WA modeling
3. Velocity structure of the outflow
4. How many absorption components?
5. Emission features
6. Time variability
7. Preliminary results of the 2005 observation
8. Conclusions

3
1. Introduction to NGC 5548

• One of the brightest S1 in UV and X-ray ?Well
  studied in UV and X-rays
• Low galactic absorption ? ideal for spectroscopy
• Moderately deep warm absorber ? blending not too
  large, but still strong WA signature

4
History 1 the low-resolution era

• ASCA data (Reynolds 1997)
• Spectrum power law
• Additional Fe-line
• Warm absorber (modeled with continuum edges)

Fe-K emission
Warm absorber
5
History 2 high spectral resolution lines!

• NGC 5548 first Seyfert ever observed at high
  spectral resolution (dec 1999, Chandra LETGS)
• Lots of absorption lines from different ions
• Shows importance of high resolution

Kaastra et al. 2000
6
2. Overview of available data
7
High-resolution X-ray and UV observations of NGC
5548

• This presentation is based on three groups of
  spectra
• 1. LETGS/HETGS/RGS single observations in
  1999/2000 (see next slide)
• 2. Large X-ray/UV campaign in 2002 (see later
  this talk)
• 3. 150 ks LETGS observation in 2005 (end of the
  talk, preliminary results)

8
NGC 5548 - need for campaign

• Chandra LETGS Dec 1999 - 86 ks
• Chandra HETGS Feb 2000 83 ks
• XMM-Newton Dec 2000 28 ks
• XMM-Newton Jul 2001 137 ks
• None with simultaneous UV
• HST/GHRS Feb/Aug 1996 18.2 ks
• HST/STIS Mar 1998 8.9 ks
• FUSE Jun 2000 25 ks
• None with simultaneous X-ray

9
Multi-? campaign in 2002

• Approved Chandra/HST/FUSE campaign
• ? Observed Jan 2002 (without FUSE, due to
  technical problems)
• HST/STIS 21 ks Crenshaw et al. 2003
• Chandra LETGSHETGS 510 ks
• Kaastra et al. 2004 - Time variability
• Steenbrugge et al. 2005 - Spectra

10
3. Short description of WA modeling
11
X-ray analysis

• Fit spectra using a power law modified
  blackbody continuum
• Where needed, add emission lines relativistic,
  BLR or NLR X-ray lines
• Fit warm absorber using a model (see next slide)
  ? ionic or total column densities
• Using photo-ionisation model, derive NH and ?
  distribution
• Spectral fits done with SPEX, global fits

12
Photoionisation models

• Models for transmission of a thin slab
• Continuum line absorption calculated
• slab model ion columns independent
• xabs model ion columns coupled through
  xstar/cloudy runs
• warm model continuous distribution of NH(?)

13
What is in the photo-ionisation models?

• Continuum opacities from Verner et al. 95
• Line absorption from many different sources
  Ferland Verner, HULLAC (Fe-L Fe-M, O-K inner
  shell) etc.
• Take account of line profile (Voigt)
• Allow for turbulent motion and systematic outflow
  velocities
• Self-consistent photo-ionisation model is under
  development

14
More info about SPEX

• See the web page
• www.sron.nl/divisions/hea/spex/index.html

15
4. Velocity structure of the outflow
16
X-ray line profiles 1999 spectrum

• Lines are broadened 270100 km/s
• Lines are blueshifted 280100 km/s
• Lines have tendency for extended blue wing
• Some lines (O VII res, O VIII Lya) have P Cygni
  -like profile
• ? outflowing, photoionised wind model

17
A comparison with UV lines

• High spectral resolution in UV
• Only few UV lines (H I, C IV, N V and O VI)
• At least 5 components seen in absorption,
  superimposed on broad emission lines
• UV lines narrow and resolved sv 20-80 km/s
• R.M.S. width of ensemble 160-260 km/s,
  consistent with LETGS
• UV lines are blueshifted, range 160-1060 km/s
• ? UV X-ray lines two manifestations of same
  phenomenon

18
Velocity structure in UV lines

• STIS spectra (Crenshaw Kraemer 1999) show 5
  velocity components
• Nr 1 high -1040 km/s
• Nr 2 med -667 km/s
• Nr 3 med -530 km/s
• Nr 4 med -336 km/s
• Nr 5 low -160 km/s

Radial velocity (km/s)
19
X-ray line profiles

• At long ?, LETGS has higher resolution C VI
  profile not fully resolved but consistent with UV
  structure

20
Outflow velocity versus ?

• UV shows 5 velocity components
• X-ray resolution insufficient to resolve them
• But high average v at high ?

21
Velocity structure

• Strongest lines fit in three components with
  fixed v high, med and low v
• High v column density increases more rapidly with
  ? than med or low

UV/HST Simultaneous
UV/FUSE Non-simultaneous.
22
5. How many absorption components?

• First glance at 1999 LETGS spectrum strong lines
  of O VII, O VIII and others from photoionized
  plasma (Kaastra et al. 2000)
• Is there more than 1 component?

23
Why we can constrain the warm absorber in NGC
5548 so well
K-shell region

• High flux low NH ? bright in soft X-ray band
• ?also detection of WA at long ? from L-shell
  transitions Mg, Si, S etc.
• ? redundancy in determining WA structure

L-shell region
Sample of 1.5-100 Å spectra (here the 2002
spectrum)
24
The oxygen region why important?

• Good diagnostic region because
• For almost any ? there is a diagnostic ion
• Oxygen is the most abundant metal

25
Long wavelength transitions

• Our photoionisation model at long ? quite
  succesfull
• See 2002 LETGS spectrum
• Only global fitting works here low S/N
• Atomic data need update in this region

26
A more detailed look to 1999 data three
ionization components

• Ionization structure is not a simple 1-phase
  medium
• fits to LETGS data require at least 3 ionization
  components
• log ? 0.5
• log ? 1.9
• log ? 2.9

Kaastra et al.2002
27
Decomposition into separate ? evidence for 5
components

• Use column densities Fe ions from RGS data
• Measured Nion as sum of separate ? components
• LETGS results similar
• Need at least 5 components

28
Separate components in pressure equilibrium?

• Not all components in pressure equilibrium (same
  ??/TF/p)
• Division into ? comps often poorly defined
• ? Continuous NH(?) distribution see next slide

29
Column density versus ?
30
Column density versus ?
Fe at low T DR rates?
31
Continuous ionization distribution

• Continuous distribution over at least 3.5 orders
  in ?
• dNH/dln??a, with a0.400.05
• Adopt streamer-like geometry
• Take dNHn(s)ds with s distance from axis
• ?L/n(s)r², r and L constant
• ?n(s)1/(1s/s0)ß, ß1/(1a)
• ss0 n(s)1/s0.71

32
Mass loss through the wind
v (km/s) -166 -1040
?1 0.0007 0.0001
?1000 0.7 0.1
33
6. Emission features
34
Relativistic lines in NGC 5548

• Evidence for relativistic lines of O VIII and N
  VII in 1999 spectrum
• Lines weak EW 0.6 and 1.1 Å
• Significant at 3s
• Inclination 46º consistent with Fe-K (Yaqoob et
  al.)
• Inner radius lt2.6GM/c² ? Kerr hole?
• Also seen by BeppoSAX before? (Nicastro et al.)

35
Broad emission lines

• NGC 5548 C VI Lya
• FWHM 10000 km/s (Kaastra et al. 2002)
• Also seen in O VII triplet in NGC 5548
  (Steenbrugge et al. 2005) and Mrk 279 Costantini
  et al. 2005)

36
Narrow emission lines

• O VII forbidden line strongest narrow line
• No significant red/blueshift
• Low sv lt 300 km/s ? from distant region
• Not variable between 1999 and 2002 (LETGS, RGS)

37
7. Time variability
38
Long term variability of NGC 5548

• Difference between LETGS spectra
• Dec 1999 - Jan 2002
• Difference in red wing broad C VI lines (_at_ 2000
  to 3000 km/s, FWHM1000 km/s)
• Difference in O V line ? log ?-0.2

39
Short term continuum variability
40
8. New LETGS data april 2005
41
New LETGS data april 2005

• New data 150 ks
• Taken in april 2005
• NGC 5548 was in a very low state 4.5 x weaker
  than in 2002
• Continuum is very hard

42
Changes in warm absorber

• Plot shows scaled spectra
• Significant change in O V- OIII (see plot)
• OV is deeper and broader trend of 2002 continued
• Very deep O III
• Other ions no significant change

43
Interpretation of WA changes

• Luminosity drop of factor 4.5 since 2002
• But not simply scaling with ?L/nr²
• Columns O III, O IV and O V much larger
• Also larger width s440 km/s versus 70-140 km/s
  in 2002

44
Forbidden line O VII time variable

• Forbidden line constant during 1999-2002
  Chandra/XMM observations
• Fluxes in ph/m2/s
• 1999 0.810.19
• 2002 0.880.12
• 2005 0.350.08
• ? forbidden line formed at pc scale

45
Conclusions/questions

• Warm absorber in NGC 5548 and other sources more
  consistent with continuous NH(?) distribution
  then separate components in pressure equilibrium
• Outflow should occur in narrow, density
  stratified streamers
• What determines maximum ionization parameter?