Monitoring the Seyfert Galaxy Mkn 766 Continuum and

Monitoring the Seyfert Galaxy Mkn 766 Continuum and Fe line variability Liz Puchnarewicz and Keith Mason Mullard Space Science Laboratory, University College, London Ian Mc. Hardy, University of Southampton Paul O’Brien, University of Leicester Mkn 766 is a highly variable Seyfert 1 galaxy. The richness of the flux and spectral changes are invaluable for testing models of accretion disks, their feedback with the hot corona surrounding the disk and the nature of the dusty, warm ionized gas along the line of sight. Here we present our initial analysis of data from the continuous 130 ksec observation taken in Guaranteed Time, focussing on the overall spectral changes. The Seyfert was in a high state (relative to the earlier, shorter PV observation) and varied by a factor of 4 in 36 hours. We analyse the variability in total X-ray flux, X-ray colours and mean spectral slope every 100 seconds. There is no evidence for periodic changes in the flux. We compare 0. 3 -10 ke. V spectra in the high, medium and low states and compare N, O and Fe emission line parameters. Spectral fits: power-law, lines and warm absorber high medium/high Lightcurve of the MOS 2 data in the 0. 3 to 12 ke. V range. Continuum variability – broadband spectral fits every 100 seconds. lightcurve medium low/high low a warm n. H high state 1. 19 0. 14 49 e. V 52 e. V 1 ke. V 1. 4 medium state 1. 22 0. 13 44 e. V 92 e. V 700 e. V 1. 5 1. 00 0. 04 69 e. V 137 e. V 1. 5 ke. V 1. 7 Spectrum low state k. T(bb) 49 e. V N(0. 5 ke. V) O(0. 65 ke. V) Fe(~6. 7 ke. V) chin 2 The total spectrum was divided into three different flux states, high, medium and low (top left). Each spectrum was fit with a 7 -component model – Galactic column (fixed to 1. 7 e 20 cm-2), warm absorber (columns in the table are in units of 1022 cm-2), power-law and disk lines (the ‘laor’ model in XSPEC) for the carbon (0. 36 ke. V), nitrogen (0. 50 ke. V), oxygen (0. 65 ke. V) and Fe lines (~6. 7 ke. V). The normalization of the carbon line converged to zero in all fits. The inclination angles of the disk for the Laor lines were linked and converged to 30 degrees for all states – all lines have similar profiles in all states. The EWs of the Fe lines are unrealistically large – the profiles are shallow and broad which suggests that there may be an additional convex component around this region. A more detailed investigation is underway. In the low state, this 7 -component model gave a poor fit but the fit improved significantly with the addition of a ~50 e. V blackbody component. The power-law slope flattened slightly and the EWs of the emission lines increased, as expected if the continuum falls away. The column density of the warm absorber also appears to have fallen although any physical significance of this must be addressed by fits to the RGS data. These changes are all confirmed by medium/high and low/high state flux ratios (top right): a slight softening in the medium state, but a significant hardening as the spectral flux falls. spectral index, a We have also parameterized changes in ‘curvature’ of the spectrum to test whether the soft (0. 3 -2 ke. V) part of the spectrum changes independently of the hard (2 -12 ke. V), and if so, how. For this analysis, the spectrum was divided into 1000 second bins to increase the signal to noise. The results (plotted below) show that the soft end (dark red) changes very little throughout the entire 130 ksec. This implies that the emergence of the blackbody in the low state offsets the hardening of the underlying power-law. soft slope hard slope In each 100 second bin, low resolution spectra over 6 X-ray colours were constructed (0. 30. 5, 0. 5 -1. 0, 1. 0 -2. 0, 2. 0 -5. 0, 5. 0 -8. 0 and 8. 012. 0 ke. V) and fit with a simple power-law model. The change in spectral index, a, as a function of time is plotted top left (dark blue). The total count rate has also been scaled and plotted for comparison. The spectral index shows remarkable stability, particularly for the first 80 ksec when the flux continually falls and rises by ~50%. After 100 ksec though, the slope begins to flatten (harden), but this is only as the total flux becomes particularly low. There is also the suggestion that the flux is lagging slightly behind the spectral hardening but the significance of this has not been tested. lightcurve The largest spectral changes occur in the hard part of the spectrum (dark blue). The 2 -12 ke. V slope hardens significantly after 100 ksec as the total flux (scaled and plotted green) falls away. Indeed there is the suggestion of a correlation between hard X-ray slope and total flux (ie the hard part of the spectrum hardens as the flux falls away) but again, this has not been rigorously tested. www. mssl. ucl. ac. uk