A&A 371, 79-92 (2001)
DOI: 10.1051/0004-6361:20010330
M. Dietrich1,2 - C. F. Bender3 - D. J. Bergmann3 - T. E. Bills3 - N. G. Bochkarev4 - A. Burenkov5 - C. M. Gaskell3 - D. D. Gutzmer3 - R. Grove3 - M. E. Hiller3 - J. P. Huchra6 - E. S. Klimek3 - C. Lund3 - N. Merkulova7,8 - S. Pebley3 - M. A. Poulsen3 - V. I. Pronik7,8 - S. G. Sergeev7,8 - E. A. Sergeeva7,8 - A. I. Shapovalova4 - V. V. Vlasyuk5 - B. Wilkes6
1 -
Department of Astronomy, University of Florida, 211 Bryant Space Science
Center, Gainesville,
FL 32611-2055, USA
2 -
Landessternwarte Heidelberg, Königstuhl, 69117 Heidelberg, Germany
3 -
University of Nebraska, Lincoln, Department of Physics and Astronomy,
Lincoln, NE 68588-0111, USA
4 -
Sternberg Astronomical Institute, Universitetskij Prospect, 13, 119899 Moscow,
Russia
5 -
Special Astrophysical Observatory, Russian Academy of Science, Nyzknij Arkhyz,
Karachaj-Cherkess Republic, 369167, Russia
6 -
Harvard-Smithsonian Center for Astrophysics, 60 Garden Street, Cambridge, MA
02138, USA
7 -
Crimean Astrophysical Observatory, p/o Nauchny, 98409 Crimea, Ukraine
8 -
Isaac Newton Institute of Chile, Crimean Branch, Chile
Received 17 November 2000 / Accepted 6 February 2001
Abstract
Results of a ground-based optical monitoring campaign on NGC 5548 in June
1998 are presented.
The broad-band fluxes (U, B, V), and the spectrophotometric optical
continuum flux
(5100 Å) monotonically decreased in flux
while the broad-band R and I fluxes and the integrated emission-line
fluxes of H
and H
remained constant to within 5%.
On June 22, a short continuum flare was detected in the broad band fluxes.
It had an amplitude of about
18% and it lasted only
90 min.
The broad band fluxes and the optical continuum
(5100 Å)
appear to vary simultaneously with the EUV variations.
No reliable delay was detected for the broad optical emission lines in
response to the EUVE variations.
Narrow H
emission features predicted as a signature of an accretion
disk were not detected during this campaign.
However, there is marginal evidence for a faint feature at
Å with
Å redshifted by
km s-1 with respect to H
.
Key words: galaxies: active - galaxies: Seyfert galaxies - galaxies: individual (NGC 5548)
The ultraviolet and optical continuum and the broad emission line flux of Seyfert 1 galaxies are known to be variable on timescales of a few days until years. Variations on timescales of hours and even less have been observed in X-rays (e.g. Barr & Mushotzky 1986; Leighly et al. 1996; George et al. 1998a, 1998b).
To study the physical processes which are responsible for the observed spectral energy distribution of an active galactic nuclei (AGN) multiwavelength monitoring campaigns have proven to be an excellent tool (cf. Peterson 1993; Netzer & Peterson 1997 for a review; Marshall et al. 1997; Nandra et al. 1998; Edelson et al. 2000). Thus, over the last decade, several large space-based and ground-based monitoring programs have been undertaken for nearby AGN, such as NGC 5548 (Clavel et al. 1991; Peterson et al. 1991, 1992, 1994, 1999; Maoz et al. 1993; Dietrich et al. 1993; Korista et al. 1995; Chiang et al. 2000), NGC 3783 (Reichert et al. 1994; Stirpe et al. 1994; Alloin et al. 1995), NGC 4151 (Crenshaw et al. 1996; Kaspi et al. 1996; Warwick et al. 1996; Edelson et al. 1996), Fairall 9 (Rodríguez-Pascual et al. 1997; Santos-Lleó et al. 1997), 3C 390.3 (Leighly et al. 1997; Dietrich et al. 1998; O'Brien et al. 1998), and NGC 7469 (Wanders et al. 1997; Collier et al. 1998).
A broad-line region (BLR) size of the order of less than a few light weeks
is indicated for Seyfert 1 galaxies by the correlated variations of the
broad emission-line flux and of the optical/ultraviolet continuum.
It is generally assumed that the central supermassive black hole is
surrounded by an accretion disk.
Such an accretion disk is the probable origin of a significant
fraction of the broad emission-line flux (e.g. Laor & Netzer 1989;
Dumont & Collin-Souffrin 1990; Halpern 1990;
Zheng et al. 1991;
Hubeny et al. 2000).
Recently, Kaspi et al. (2000) published results of a long
term monitoring campaign of quasars.
Including the results of Seyfert 1 galaxies (Wandel et al. 1999)
they suggested a relation between the BLR size and the optical luminosity
given by
.
The Seyfert 1 galaxy NGC 5548 has been continuously monitored in the
optical since late 1988 by the international AGN watch consortium
(Peterson et al. 1999; cf. Alloin 1994 for an
AGN watch overview).
In June 1998 this prominent Seyfert 1 galaxy was targeted for a
coordinated intense monitoring campaign using EUVE, ASCA, and RXTE to study
the high energy continuum emission and its temporal characteristics
(Chiang et al. 2000).
Chiang et al.'s observations indicate that the variations at 0.2 keV
(EUV) appear to lead similar variations at energies larger than
1 keV by 3-8 hrs. This was unexpected as it was generally assumed
that correlated variations of the EUV, UV, and optical emission would all be
due to reprocessed higher energy radiation.
Since this campaign provided the rare opportunity to access the high energy continuum variations with especially high temporal sampling, we organized a simultaneous ground-based campaign for the optical wavelength domain.
Assuming part of the BLR flux is emitted from the accretion disk, variations on timescales of one day or even less should be detected (Stella 1990). For this model of a relativistic Keplerian disk it was shown that weak narrow features should drift across the emission line profile. These weak structures are expected to start at high velocities i.e. in the outer profile wings, and move towards the line center to be due to the longer time delay at larger radii of the disk in response to the variable continuum emission. The timescale of the shift depends on the mass of the central black hole and it is for NGC 5548 of the order of less than several days (cf. Stella 1990). The detection of such structures would be a strong indication for the presence of an accretion disk in the innermost region of an AGN.
Generally, current monitoring campaigns have not provided the necessary temporal and spectral resolution for detecting such substructures in the broad emission line profiles. However short term monitoring campaigns had been undertaken for NGC 4151 (Xanthopoulos & DeRobertis 1991; Crenshaw et al. 1996; Kaspi et al. 1996; Warwick et al. 1996; Edelson et al. 1996) and for a small sample of AGN (e.g. NGC 5548, NGC 4151, 3C 390.3, Arp102B, Mkn 6; Eracleous & Halpern 1993). As yet no significant short timescale broad emission line flux variation or profile variability has been detected.
In this paper, we present the results of the optical photometric and
spectroscopic observations that were obtained in June 1998 simultaneous with
the high energy campaign (Chiang et al. 2000).
In Sect. 2 we describe the optical observations and outline intercalibration
procedures by which a homogeneous set of photometric and spectroscopic
measurements can be achieved.
In Sect. 3 we present measurements of the broad-band flux as well as of the
broad Balmer emission-line flux.
We compare our results with the results of the simultaneous campaign of the
EUV wavelength range (Chiang et al. 2000).
We also discuss the shape of the H
and H
line profiles.
We summarize our results in Sect. 4.
Source | Code | Tel. | Photometry Aperture | Spectroscopy | ||||
[m] | U | B | V | R | I | Aperture | ||
Mt. Hopkins Observatory | C | 1.5 | -- | -- | -- | -- | -- |
![]() |
Crimean Astrophysical Obs. | D | 1.25 | 15 | 15 | 15 | 15 | 15 | -- |
Calar Alto Observatory | G | 2.2 | -- | -- | -- | -- | -- |
![]() |
Special Astrophysical Obs. | L1 | 1.0 | -- | -- | -- | -- | -- |
![]() |
Special Astrophysical Obs. | L2 | 6.0 | -- | -- | -- | -- | -- | ![]() |
Univ. of Nebraska 0.4 m | Q | 0.4 | -- | -- | 8 | -- | -- | -- |
Shajn reflector, Crimean Obs. | W | 2.6 | -- | -- | -- | -- | -- |
![]() |
A complete log of the photometric and spectroscopic observations is given in Tables 1A (available in electronic form at the CDS) and 2A.
Photometric observations of NGC 5548 were obtained by Gaskell et al.
(sample Q) and Merkulova (sample D).
The brightness of NGC 5548 was determined with respect to stars in
the field of NGC 5548 using the photometric sequence (stars 1 and 2)
defined by Penston et al. (1971), and star C1 and C
defined by Lyuty (1972). The bright star 1
east
to the galaxy located at PA = 240
at a distance of
9
with respect to NGC 5548 also was used. This star is
referred in the HST guidestar catalogue as gsc 0201001062
(RA = 17
17
29
66, Dec = +25
03
10
5
(2000.0)).
Star C1 in Lyuty is identical to star 4 in Penston et al.
The
magnitudes of C1 and 4 are identical within the errors and
differ by only
0.02 mag (B,V) and
0.17 mag (U)
(cf. Table 2). Star C is the bright star close to star 3 in Penston et al. located approximately 1
3 southeast of it.
band | 1a | 2a | 4,C1b | Cb | 0 |
U | 14.52 | 16.24 | 13.23 | 11.17 | -- |
B | 14.45 | 16.07 | 12.78 | 10.96 | -- |
V | 13.80 | 15.38 | 11.90 | 10.47 | 11.29 |
R | -- | -- | 11.17 | 9.96 | -- |
I | -- | -- | 10.64 | 9.57 | -- |
a From Penston et al. (1971), estimated
uncertainty 0.02 mag.
b From Lyuty (1972), estimated uncertainty 0.01 mag.
(U) of C1 0.02 mag.
The photometric observations of NGC 5548 of sample D obtained with the
1.25 m telescope of the Crimean Astrophysical Observatory were recorded
using a five channel version of the Double Image Chopping Photometer -
Polarimeter (Piirola 1973).
The observations were made simultaneously in five colors using dichroic
filters to split the light into five spectral bands.
The resulting passbands are close to the standard Johnson UBVRI photometric
system with effective wavelengths at 3600 Å, 4400 Å, 5300 Å,
6900 Å, and 8300 Å, respectively. A diaphragm with two equal
apertures (15
in diameter) was used in the focal plane of the
1.25 m telescope.
The distance between the aperture centers was 26
.
A rotating chopper
alternately closes one of the apertures, leaving the other free, thus the
photocathode is illuminated alternately by the galaxy (or the star) and the
sky apertures. Because the centers of the apertures are closely spaced,
background observations were also obtained at a distance of about 7
from the galaxy nucleus to correct the observational background for a
contribution of the outer regions of the galaxy. The telescope is fully
automated and an autoguider was used.
Observations of the nucleus of NGC 5548 were made on nights with good
atmospheric conditions - when the estimated seeing was in the range of
1
to 3
.
The positional accuracy by the autoguider during observations is better than
20% of the estimated seeing, i.e. better than 0
6 in the worst
case.
The galaxy nucleus was also positioned in the aperture using the autoguider,
with similar positioning errors.
The conventional technique of differential measurements was applied.
Two comparison stars (labelled as C1 and C) were taken from the list of
Lyuty (1972). Observations were performed against a comparison star
C1; and the second star, C, was used to check the results of the first
comparison.
In addition to stars C1 and C, secondary UBVRI standards by Neckel &
Chini (1980) were used for an absolute calibration.
The measurements were carried out by observing in the following sequence,
C-C1-sky-AGN-sky-C1-C-sky.
The time resolution was about 3.5 min. During a single observation, 8 integrations of 10 s each were made. Photon statistics (corrected for sky
background) were applied to calculate photometric errors, which generally were
the same as the rms errors obtained by averaging the 8 integrations.
Generally, the atmospheric seeing during the observations obtained with the
1.25 m CAO telescope averaged 2
.
It was better for two nights
(June 27/28 and June 30/July 1) and worse for the night June 22/23
(2-3
).
The V-band observations of Gaskell et al. (sample Q) were obtained with the
University of Nebraska 0.4 m telescope in Lincoln, Nebraska.
The frames were recorded with a ST-7 CCD camera.
The flux of NGC 5548 was determined for an 8
diameter aperture.
The sky was measured in an annulus of radius 15 to 20
.
The V-band magnitude was derived relative to the comparison stars 1 and
gsc0201001062 in the field of NGC 5548 (cf. Penston et al. 1971).
The exposure time of an individual frame was 3 min.
Generally, to increase the signal-to-noise ratio of the flux measurements,
the V-band magnitudes of three subsequently recorded frames were averaged.
The seeing during the observations was typically 4
.
There was no
correlation detected between the derived magnitude of NGC 5548 and the
seeing.
To combine the broad band flux measurements of sample D and Q, the apparent
magnitudes were transformed into fluxes. The conversion was performed using
the following equations (Allen 1973; Wamsteker 1981):
![]() |
= | -0.4 mU - 8.361, | (1) |
![]() |
= | -0.4 mB - 8.180, | |
![]() |
= | -0.4 mV - 8.439, | |
![]() |
= | -0.4 mR - 8.759, | |
![]() |
= | -0.4 mI - 9.080. |
![]() |
Figure 1:
Optical broad-band light curves for NGC 5548. Fluxes
in broad-band U, B, V, R, and I are in units of
10-15 erg s-1 cm-2 Å-1.
In the top panel the normalized V-band flux of star 0 (+) and
C (x) is shown which was contant within ![]() |
Open with DEXTER |
The flux calibration of AGN spectra can be accomplished in several ways.
In variability studies, it has become common practice to normalize the flux
scale to the fluxes of strong forbidden narrow emission lines which are
assumed to be constant over timescales of at least several decades
(e.g. Peterson 1993).
This assumption is justified by the large spatial extent and low gas density
of the narrow-line region (NLR), but is only valid if the aperture used
is larger than the NLR.
Light travel-time effects and the long recombination timescale
(
years for
cm-3)
damp out short timescale variability.
As a major sources of uncertainty in inhomogenous samples of variability data
is the use of different instrumental settings, it is important to take
aperture effects into account (Peterson & Collins 1983).
The seeing-dependent uncertainties which are introduced by the aperture
geometry can be minimized by using large apertures.
It has been shown that apertures of 5
7
5 can
reduce seeing-dependent photometric errors to no more than a few percent
in the case of nearby AGNs (Peterson et al. 1995).
The spectra of NGC 5548 obtained at the 1 m telescope of SAO RAS
(sample L1 Table 1) were extracted for a slit aperture of
and
,
respectively. For these large apertures seeing losses can be neglected.
To measure the
F([O III] 5007 Å) a linear pseudo-continuum was fitted beneath
the [O III] emission line and the flux was measured for the
wavelength range
Å. For the extraction windows used for sample
L1 we determined a flux of F([O III] 5007 Å)
= (
)
10-13 erg s-1 cm-2
(
)
and
F([O III] 5007 Å)
= (
)
10-13 erg s-1 cm-2
(
), respectively. These values are
in good agreement with F([O III] 5007 Å) measured for NGC 5548
during the last 10 years (cf. Peterson et al. 1999). For large
apertures of comparable size (e.g.
)
F([O III] 5007 Å)
= 5.1 to 5.9 10-13 erg s-1 cm-2 is mentioned
(Peterson et al. 1992).
To be constistent with the H
line flux measurements of NGC 5548
obtained since 1988 (Peterson et al. 1999) we adopted
F([O III] 5007 Å)
= (
)
10-13 erg s-1 cm-2.
The broad band flux variations in U, B, V, R, I are displayed in Fig. 1.
In addition to the broad band flux variations, the light curve of the
measured flux of the comparison stars C and 0 (cf. Table 2) are shown.
The normalized flux remained constant within 0.57%.
The broad band flux of NGC 5548 decreases throughout the campaign in June
1998.
This trend is clearly visible in the U- and B-band, and the ratio of the
flux levels observed in early and late June becomes smaller for
increasing wavelength. In the I-band the flux decay is negligible.
The merged V-band light curve of sample D and Q shows different internal
scatter of the measurements recorded during individual nights (Figs. 1, 2).
This difference might be caused by different seeing conditions of the
observations. Merkulova (sample D) measured the brightness of NGC 5548
with an 15
aperture under a typical seeing of 1-3
.
The
seeing amounts to
4
for the measurements Gaskell
et al. (sample Q) observed for an 8
aperture.
Modelling the host galaxy of NGC 5548 Romanishin et al. (1995)
showed that the light loss of the AGN is of the order of 10% for sample Q
and only slightly larger for the host galaxy.
However, the different scatter may be also intrinsic. The temporal
resolution of the samples amounts to
4 min (D) and nearly
16 min (Q), respectively.
![]() |
Figure 2:
The strong variation in the broad band fluxes as measured on June 22
(JD = 2450987). In all 5 broad band fluxes a rapid increase and
subsequent decay was detected within ![]() |
Open with DEXTER |
The most obvious structure in the broad band light curves is a flare
like event. This variation was detected on JD = 2450987 (June 22). The
observations recorded during this night are shown in more detail in Fig. 2.
In all 5 broad band measurements a short intense increase of the flux level
occured. The broad band flux increased by % within
30 min and decreased within
60 min to the flux level of the
beginning of this event. The duration of this flare lasted only
90 min. This remarkable flux increase can not be due
to temporal variations of a standard star which was used for relative flux
calibration. The flux calibration is based on two stars which were in cross
checked with secondary standard stars taken from Neckel & Chini
(1980) in addition.
As can be seen in Figs. 1 and 2 the used comparison stars were constant
within less than 0.57%.
The standard deviations of the primary star used for calibration during
this night amount to 0.018, 0.014, 0.009, 0.023, 0.001 mag in
U, B, V, R, I.
Broad emission-line fluxes were integrated over a wavelength range of
4850-5020 Å for H
and
6500-6800 Å for H
.
A local linear continuum fit was interpolated under each emission line. In
the case of the H
region, the continuum was defined by the flux measured
in two windows (10 Å width) at 4845 Å and 5170 Å in
the observed frame (H
)
and at 6350 Å and 6965 Å (H
).
No attempt was made to correct any of the measured emission-line fluxes
for their respective narrow-line contributions.
The optical continuum flux
(5100 Å) was determined as the
average value in the range 5185-5195 Å (cf. Table 4).
The measured H
and H
emission line fluxes were adjusted to take
into account the different slit sizes (Table 1).
The measurements of both H
and H
were normalized by a similar
factor
![]() |
(2) |
![]() |
Figure 3:
Light curves for the emission lines H![]() ![]() ![]() |
Open with DEXTER |
![]() |
(3) |
where
is the adopted absolute flux of
F([O III] 5007 Å), the quantity in brackets is the observed
continuum to F([O III] 5007 Å) flux ratio measured from the
spectrum, and G is an aperture-dependent correction for the host-galaxy flux.
The sample L1 which uses a relatively large aperture
(8
0
6
6) was adopted as the standard
(i.e.,
,
G = 0 by definition), and other data sets were merged
progressively by comparing measurements based on observations made during
the same night.
Note that this means that any real variability that occurs on timescales this
short will be somewhat suppressed by the process that allows us to combine the
different data sets.
The scaling factor
and the additive scaling factor G for the
various samples are given in Table 3.
sample | add. constant | point source | extended source |
V-maga | correction
![]() |
correction Gb | |
C | -- |
![]() |
![]() |
D | 1.000 | -- | -- |
G | -- |
![]() |
![]() |
L1 | -- | 1.00 | 0.00 |
L2 | -- |
![]() |
![]() |
Q |
![]() |
-- | -- |
W | -- |
![]() |
![]() |
a Relative to sample D.
b Relative to sample L1.
The resulting light curves of the optical continuum (5100 Å) and
of the H
and H
emission line fluxes are displayed in Fig. 3
and presented in Table 3A (available in electronic form at the CDS). The
optical continuum light curve
is very similar to the broad-band V light curve (cf. Figs. 1 and 3).
The difference in the flux levels is probably due to the larger aperture used
for the broad band flux determination relative to that used for the optical
continuum fluxes.
A final check of the uncertainty estimates was performed by examining the ratios of all pairs of photometric and spectroscopic observations which were separated by 0.05 days or less.
There are more than a dozen independent pairs of measurements separated by
less than 0.05 d.
For H,
the uncertainty estimate is dominated by the spectra taken at
Calar Alto Observatory, while for H
more than 20 pairs could be used.
The dispersion about the mean (unity), divided by
,
provides an
estimate of the typical uncertainty in a single measurement
(
).
The observational uncertainties (
)
assigned to the spectral
line flux measurements were estimated from the error spectra which were
calculated by the intercalibration routine, as well as from the
signal-to-noise ratio within the spectral range near the individual emission
lines.
For example the mean fractional flux error is
for
the H
line, and the average fractional uncertainty from the internal
statistical estimate is
.
For the H
line the values are
and
which implies that the error estimates for both
lines are probably quite good.
Generally, the estimated errors (
)
are of the same order as,
but slightly smaller than, the observational uncertainties
(
)
derived directly from multiple measurements (Table 5).
The large difference of
and
for
the broad band photometric fluxes is mainly due to the measurements
obtained on
(June 22), which we expect is due to a real
variation.
feature |
![]() |
![]() |
feature |
![]() |
![]() |
U | 0.008 | 0.021 | I | 0.006 | 0.024 |
B | 0.008 | 0.020 | ![]() |
0.020 | 0.016 |
V | 0.010 | 0.020 | H![]() |
0.022 | 0.010 |
R | 0.004 | 0.016 | H![]() |
0.040 | 0.025 |
a Observational uncertainty based on uncertainties assigned to individual
points.
b Mean fractional uncertainty based on point-to-point differences between
closely spaced (i.e.,
d) measurements.
The average interval between measurements of the combined broad band light
curves is about
days for U, B, R, I, and for V
days.
However, if the large gaps (>5 days) are not taken into account, the
sampling rate drops to
days (U, B, R, I) and
days (V) (Table 6).
The sampling of the H
,
H
,
and the optical continuum flux
(5100 Å) can be obviouly devided into two regimes.
Taking into account all measurements the sampling is about
(H
=
days for H
and
(5100 Å).
Neglecting the large temporal gap of nearly two weeks at the beginning of the
monitoring campaign (full moon),
is
(H
=
days.
For the H
light curve the corresponding sampling rates are
(H
days and
(H
days, respectively.
feature | N | interval | intervala |
[days] | [days] | ||
U | 120 |
![]() |
![]() |
B | 120 |
![]() |
![]() |
V | 178 |
![]() |
![]() |
R | 120 |
![]() |
![]() |
I | 120 |
![]() |
![]() |
![]() |
52 |
![]() |
![]() |
H![]() |
52 |
![]() |
![]() |
H![]() |
38 |
![]() |
![]() |
a Neglecting temporal gaps larger than 5 days.
The broad H
and H
emission lines, the optical continuum
flux
(5100 Å) and the U, B, V, R, I broad band fluxes
appear to exhibit small amplitude variations on timescales of days
(Figs. 1-3).
The broad band light curves U, B, V, as well as the
(5100 Å)
flux show a decreasing flux from the beginning of this campaign to the end.
The broad band flux at longer wavelengths (R and I) appeared to be nearly
constant.
The H
and H
light curves show weak evidence for small
amplitude variations on timescales of only a few days. However, this
is statistically not significant; within the
errors the H
and H
emission line flux can be regarded as
nearly constant. This can be seen by deriving the variability parameter
.
The variability parameters
and
have been
calculated for the broad-band flux variations as well as for the
broad emission lines and the optical continuum
(cf. Clavel et al. 1991;
Rodríguez-Pascual et al. 1997).
The quantity
is the ratio of the maximum to the minimum
flux. The quantity
is an estimation of the fluctuations of the
intrinsic variations relative to the mean flux. Therefore, the rms of the
light curves has been corrected with respect to the uncertainties introduced
by the observations.
Values of
and
for the broad-band measurements
are given in Table 7.
The value of
for the optical continuum
(5100 Å)
agrees well with
for the V-band.
Furthermore, for the broad band variations and the optical continuum
we reproduce the well known trend that the amplitude of the variation
decreases with increasing wavelength (Table 7).
The broad band fluxes and the
(5100 Å) continuum flux
were not corrected for the host galaxy contribution since this
fraction is only of the order of less than 10%
(Romanishin et al. 1995).
The values of
determined for the H
and H
emission line flux variations indicated that the Balmer line flux can be
taken as constant within the errors, even if one is tempted to glimpse hints
of small amplitude variations on timescales of a few days
(cf. Table 7, Fig. 3).
feature | mean flux | rms flux |
![]() |
![]() |
U | 24.80 | 2.37 | 1.48 | 0.0865 |
B | 20.08 | 1.56 | 1.44 | 0.0697 |
V | 17.76 | 0.91 | 1.38 | 0.0421 |
R | 17.54 | 0.78 | 1.34 | 0.0406 |
I | 13.15 | 0.69 | 1.44 | 0.0557 |
![]() |
14.4 | 0.9 | 1.36 | 0.040 |
H![]() |
10.5 | 0.2 | 1.11 | -0.005 |
H![]() |
38.7 | 1.2 | 1.13 | -0.006 |
EUV | 0.0993 | 0.0376 | 59 | 0.2035 |
![]() |
Figure 4:
The normalized EUVE light curve from the monitoring program for
NGC 5548 (Chiang et al. 2000) is shown in the bottom
panel. In the top panel the normalized variations of ![]() |
Open with DEXTER |
For comparison we also calculated
and
for the rapid
variations detected by Chiang et al. (2000) in the extrem
ultraviolet domain (Table 7).
In Fig. 4 we show the normalized light curve of
(5100 Å)
and of the EUV variations (kindly provided by Chiang and collaborators).
The
(5100 Å) continuum flux show the same overall trend as
the EUV variations. But in the optical continuum the rapid variations
visible in the EUV are smeared out or the amplitude of the variations at
Å is so small that it is hidden by the measurement
uncertaintities.
The disappearance of small scale variations with increasing wavelength was
also observed for NGC 4151 in Dec. 1993
(cf. Edelson et al. 1996).
However, the H
and H
emission line fluxes show only marginal
indications for variations. Within the errors the emission line flux was
constant within 5% which is also indicated by
(cf. Table 7).
In spite of this we calculated cross-correlation functions.
We used the extrem ultraviolet light curve measured with EUVE for the driving
continuum kindly provided by Chiang and collaborators.
We computed the ICCF (cf. White & Peterson 1994) in the formalism
of the so called "local CCF'' (Welsh 1999). This approach takes
into account the influence of the bias to underestimate the delay as has been
shown by Welsh (1999).
To correct for the influence of low frequency power in the flux variations
we calculated a linear trend which was subtracted. The removal of a linear
trend helps to reduce the bias of the observed delay towards values smaller
than the real delay (Welsh 1999). We also applied the
cross-correlation analysis to the detrended light curves, and we calculated
the discrete correlation function (DCF) for comparison.
Uncertainties in the ICCF results for the cross-correlation maxima
and centroids were computed through Monte Carlo techniques
(cf. Peterson et al. 1998).
The sampling characteristics of each of the light curves are given in
Table 8. The name of the feature and the total number of points, N,
in the light curves are given that were used in computing the
auto- and cross-correlation functions. The width (FWHM) of the ACF
based on the original and of the detrended light curves are given as
well.
The corresponding widths of ICCF computed by cross-correlation with the
EUV continuum at 0.2 keV is listed too.
![]() |
Figure 5:
Cross-correlation functions for U, B, V, R, Ifluxes, ![]() ![]() ![]() ![]() |
Open with DEXTER |
feature | N | FWHM (ACF) | FWHM (ICCF) | ||
orig. | detr. | orig. | detr. | ||
U | 120 | 6.8 | 3.3 | 7.1 | 5.0 |
B | 120 | 5.7 | 4.2 | 6.6 | 5.9 |
V | 178 | 1.7 | 0.2 | 9.3 | 2.2 |
R | 120 | 1.8 | 1.6 | 4.1 | 5.9 |
I | 120 | 0.2 | 0.2 | 3.6 | 3.7 |
![]() |
52 | 30 | 0.4 | 6.1 | 3.3 |
H![]() |
52 | 1.7 | 0.6 | 4.0 | 2.4 |
H![]() |
38 | 1.2 | 0.2 | 1.1 | 1.4 |
The results of the cross-correlation analysis are given in Table 9.
The first column indicates the "responding'' light curve (i.e., the light
curve that is assumed to be responding to the driving light curve). The
second and third column provide the peak value of the correlation coefficient
for the ICCF (original and detrended). The position of the
peak of the cross-correlation functions
was measured by
determing the location of the peak value of the ICCF and DCF; these values
are given in the next four columns, respectively,
The centroids
of the ICCF of the original and detrended
light curves were computed using the points in the cross-correlation
function with values greater than
.
The error estimate for the position of the cross-correlation peak
and the ICCF centroid
for the original and detrended light curves are also given
(cf. Table 9 columns
and
).
The broad-band (U, B, V, R, I) and the optical continuum flux
variations appear to be simultaneous within the errors, relative to the EUV
continuum variations at 0.2 keV (Fig. 5).
Taking into account the small variability amplitude expressed by
(Table 7), the temporal coverage of the time series of 30
days only, with dense temporal sampling solely for the second half,
the small CCF amplitudes, and the uncertainty of the location of the
peak and centroid of the CCFs of
3 days (Table 9)
no reliable delay was detected for the broad emission lines of H
and
H
in response to the EUV variations (Fig. 5).
The mean and rms spectra of the samples C, L1, L2, W are identical within
less than 3% (H)
and 1.5% (H
).
![]() |
Figure 6:
The mean and root-mean-square (rms) spectra the of the
H![]() ![]() ![]() ![]() ![]() ![]() ![]() |
Open with DEXTER |
![]() |
Figure 7:
The average echelle spectra of the H![]() ![]() |
Open with DEXTER |
The mean and rms spectra of the H
and H
region are presented
for the spectra of subsample C (Fig. 6). It provides 19 epochs during the
studied period and a daily sampling with homogeneous settings for the second
half of June 1998.
The rms spectrum of H
and H
indicates continuum variations of
the order of less than 5%. Furthermore, weak broad emission line flux
variations are superimposed but the amplitude is less than 2%.
Thus, these features in the rms-spectra again should be taken only as
weak indications for emission line flux variations on timescales of days.
However, the width of the broad HeII4686 emission line in the rms spectrum
is similar to that detected by Peterson et al. (2000) for the
narrow-line Seyfert 1 galaxy NGC 4051. The contribution of FeII emission
(mulitplets 37, 38) should be negligible since there is no indication of
variable FeII emission (multiplets 48, 49) in the rms spectrum in the
wavelength range
Å.
![]() |
![]() |
![]() |
error.est. | |||||||
feature | orig. | detr. | ICCF | DCF | ICCF | [days] | ||||
orig. | detr. | orig. | detr. | orig. | detr. |
![]() |
![]() |
|||
U | 0.86 | 0.57 | 0.0 | 0.9 | 0.0 | 1.0 | -0.1 | -0.1 | 0.6 | 1.1 |
B | 0.83 | 0.65 | -0.1 | -1.0 | 0.0 | 1.0 | -0.4 | -0.6 | 0.6 | 0.6 |
V | 0.71 | 0.40 | 1.0 | 1.0 | -2.0 | 1.0 | 1.4 | 1.0 | 2.1 | 2.1 |
R | 0.60 | 0.62 | -1.2 | -1.3 | 1.0 | 1.0 | -0.0 | -0.1 | 0.9 | 0.8 |
I | 0.52 | 0.58 | 0.5 | 0.5 | 1.0 | 1.0 | 0.3 | 0.6 | 5.2 | 2.0 |
![]() |
0.83 | 0.53 | -0.1 | -0.1 | 0.0 | 0.0 | 0.5 | 0.1 | 1.4 | 3.2 |
H![]() |
0.56 | 0.04 | 6.1 | 6.5 | 6.0 | 7.0 | 6.1 | 7.8 | 3.0 | 4.5 |
H![]() |
0.26 | 0.29 | 9.8 | 9.7 | 10. | 10. | 9.8 | 9.7 | 6.0 | 2.7 |
The spectra obtained at Calar Alto Observatory have a significant higher
spectral resolution ( km s-1) than the spectra of the
other samples (
km s-1).
Hence, we used the spectra of sample G to search for small features as
suggested by Stella (1990). The H
profile was filtered
with a narrow gaussian curve (FWHM=0.1 Å) to improve the signal-to-noise
ratio.
The average spectra of each night of the 15 epochs obtained at Calar Alto
Observatory are presented in Fig. 7 as well as the mean spectrum of the
samples C, L1, L2, W.
The shape of the H
profile is smooth and no narrow features have been
detected during this short period.
There is no obvious narrow moving feature visible in the H
profile.
However, at
Å there might be a marginal indication of
a weak structure which is visible in the mean spectrum and in most of the
individual spectra of sample G. It is redshifted by
1100 km s-1
and the width amounts to
6 Å corresponding
km s-1.
We used also the spectra of the samples with lower spectral resolution to
search for this weak feature. It is also detectable in the mean spectrum
based on these samples as can be seen in Fig. 7.
Unfortunately, the redshift of
km s-1 places this
weak feature in the red wing of the prominent [N II]6583
emission line. Hence, it is not possible to detect this structure in
the H
line profile to provide further evidence for the existence
of this weak feature.
Acknowledgements
This work has been supported by SFB328D (Landessternwarte Heidelberg), by the NASA grant NAG5-3234 (University of Florida), the Russian Basic Research Foundation grant N94-02-4885a, N97-02-17625 (Sternberg Astronomical Institute, Special Astrophysical Observatory), by the Smithsonian Institution, and by INTAS grant N96-032. We would like to thank the FLWO remote observers P. Berlind and M. Calkins, and also S. Tokarz for help in reducing and archiving the FLWO data.
civil date | jul.date | code |
![]() |
range | aperture | PA | see. | Res. | filename |
mid. | [s] | [Å] | [arcsec] |
![]() |
[arcsec] | [Å] | |||
98 Jun. 01 | 2450965.67125 | C | 180 | 3655-7529 | 3.0 ![]() |
91 | 2.0 | n50965f | |
98 Jun. 01 | 2450966.42281 | W | 3600 | 4312-5612 | 3.0 ![]() |
90 | 2.0 | 7-8 | n50966wb |
98 Jun. 02 | 2450966.67348 | C | 180 | 3654-7529 | 3.0 ![]() |
91 | 3.0 | n50966f | |
98 Jun. 16 | 2450980.75052 | C | 180 | 3655-7530 | 3.0 ![]() |
90 | 2.0 | n50980f | |
98 Jun. 17 | 2450981.71094 | C | 180 | 3655-7530 | 3.0 ![]() |
90 | 1.5 | n50981f | |
98 Jun. 17 | 2450982.33950 | W | 3720 | 4298-5598 | 3.0 ![]() |
90 | 3.5 | 7-8 | n50982wba |
98 Jun. 17 | 2450982.40486 | G | 3600 | 5600-7100 | 2.05 ![]() |
0 | 2 | 0.12 | n50982da1 |
98 Jun. 17 | 2450982.40486 | G | 3600 | 4375-5375 | 2.05 ![]() |
0 | 2 | 0.12 | n50982db1 |
98 Jun. 17 | 2450982.41389 | L1 | 1200 | 4090-5800 | 8.0 ![]() |
0 | 2 | 9 | n50982llb1 |
98 Jun. 17 | 2450982.41389 | L1 | 1200 | 4090-5800 | 8.0 ![]() |
0 | 2 | 9 | n50982llb4 |
98 Jun. 17 | 2450982.42847 | L1 | 1200 | 4090-5800 | 8.0 ![]() |
0 | 2 | 9 | n50982llb2 |
98 Jun. 17 | 2450982.42847 | L1 | 1200 | 4090-5800 | 8.0 ![]() |
0 | 2 | 9 | n50982llb5 |
98 Jun. 17 | 2450982.44514 | L1 | 1200 | 4090-5800 | 8.0 ![]() |
0 | 2 | 9 | n50982llb3 |
98 Jun. 17 | 2450982.44514 | L1 | 1200 | 4090-5800 | 8.0 ![]() |
0 | 2 | 9 | n50982llb6 |
98 Jun. 17 | 2450982.45000 | G | 3600 | 5600-7100 | 2.05 ![]() |
0 | 2 | 0.12 | n50982da2 |
98 Jun. 17 | 2450982.45000 | G | 3600 | 4375-5375 | 2.05 ![]() |
0 | 2 | 0.12 | n50982db2 |
98 Jun. 17 | 2450982.48648 | W | 5505 | 4296-5586 | 3.0 ![]() |
90 | 3.4 | 7-8 | n50982wbb |
98 Jun. 17 | 2450982.49514 | G | 3600 | 5600-7100 | 2.05 ![]() |
0 | 2 | 0.12 | n50982da3 |
98 Jun. 17 | 2450982.49514 | G | 3600 | 4375-5375 | 2.05 ![]() |
0 | 2 | 0.12 | n50982db3 |
98 Jun. 18 | 2450982.53958 | G | 3600 | 5600-7100 | 2.05 ![]() |
0 | 2 | 0.12 | n50982da4 |
98 Jun. 18 | 2450982.53958 | G | 3600 | 4375-5375 | 2.05 ![]() |
0 | 2 | 0.12 | n50982db4 |
98 Jun. 18 | 2450982.79700 | C | 180 | 3656-7530 | 3.0 ![]() |
90 | 1.5 | n50982f | |
98 Jun. 18 | 2450983.33958 | L1 | 1200 | 4090-5800 | 8.0 ![]() |
0 | 2 | 9 | n50983llb1 |
98 Jun. 18 | 2450983.33958 | L1 | 1200 | 4090-5800 | 8.0 ![]() |
0 | 2 | 9 | n50983llb2 |
98 Jun. 18 | 2450983.37845 | W | 4235 | 4284-5586 | 3.0 ![]() |
90 | 3.0 | 7-8 | n50983wb |
98 Jun. 18 | 2450983.39861 | G | 3600 | 5600-7100 | 2.05 ![]() |
0 | 2 | 0.12 | n50983da1 |
98 Jun. 18 | 2450983.39861 | G | 3600 | 4375-5375 | 2.05 ![]() |
0 | 2 | 0.12 | n50983db1 |
98 Jun. 18 | 2450983.44375 | G | 3600 | 5600-7100 | 2.05 ![]() |
0 | 2 | 0.12 | n50983da2 |
98 Jun. 18 | 2450983.44375 | G | 3600 | 4375-5375 | 2.05 ![]() |
0 | 2 | 0.12 | n50983db2 |
98 Jun. 18 | 2450983.49236 | G | 3600 | 5600-7100 | 2.05 ![]() |
0 | 2 | 0.12 | n50983da3 |
98 Jun. 18 | 2450983.49236 | G | 3600 | 4375-5375 | 2.05 ![]() |
0 | 2 | 0.12 | n50983db3 |
98 Jun. 19 | 2450983.53681 | G | 3600 | 5600-7100 | 2.05 ![]() |
0 | 2 | 0.12 | n50983da4 |
98 Jun. 19 | 2450983.53681 | G | 3600 | 4375-5375 | 2.05 ![]() |
0 | 2 | 0.12 | n50983db4 |
98 Jun. 19 | 2450983.84502 | C | 180 | 3655-7530 | 3.0 ![]() |
90 | 1.5 | n50983f | |
98 Jun. 19 | 2450984.30069 | L1 | 2400 | 4090-5800 | 8.0 ![]() |
0 | 4 | 9 | n50984llb1 |
98 Jun. 19 | 2450984.30069 | L1 | 2400 | 4090-5800 | 8.0 ![]() |
0 | 4 | 9 | n50984llb4 |
98 Jun. 19 | 2450984.34167 | L1 | 2400 | 5630-7250 | 8.0 ![]() |
0 | 4 | 9 | n50984llr1 |
98 Jun. 19 | 2450984.34167 | L1 | 2400 | 5630-7250 | 8.0 ![]() |
0 | 4 | 9 | n50984llr3 |
98 Jun. 19 | 2450984.37847 | L1 | 2400 | 4090-5800 | 8.0 ![]() |
0 | 4 | 9 | n50984llb2 |
98 Jun. 19 | 2450984.37847 | L1 | 2400 | 4090-5800 | 8.0 ![]() |
0 | 4 | 9 | n50984llb5 |
98 Jun. 19 | 2450984.39514 | G | 3600 | 5600-7100 | 2.05 ![]() |
0 | 1 | 0.12 | n50984da1 |
98 Jun. 19 | 2450984.39514 | G | 3600 | 4375-5375 | 2.05 ![]() |
0 | 1 | 0.12 | n50984db1 |
98 Jun. 19 | 2450984.41389 | L1 | 2400 | 5630-7250 | 8.0 ![]() |
0 | 4 | 9 | n50984llr2 |
98 Jun. 19 | 2450984.41389 | L1 | 2400 | 5630-7250 | 8.0 ![]() |
0 | 4 | 9 | n50984llr4 |
98 Jun. 19 | 2450984.43958 | G | 3600 | 5600-7100 | 2.05 ![]() |
0 | 1 | 0.12 | n50984da2 |
98 Jun. 19 | 2450984.43958 | G | 3600 | 4375-5375 | 2.05 ![]() |
0 | 1 | 0.12 | n50984db2 |
civil date | jul.date | code |
![]() |
range | aperture | PA | see. | Res. | filename |
mid. | [s] | [Å] | [arcsec] |
![]() |
[arcsec] | [Å] | |||
98 Jun. 19 | 2450984.44403 | L1 | 1200 | 4090-5800 | 8.0 ![]() |
0 | 4 | 9 | n50984llb3 |
98 Jun. 19 | 2450984.44403 | L1 | 1200 | 4090-5800 | 8.0 ![]() |
0 | 4 | 9 | n50984llb6 |
98 Jun. 19 | 2450984.48403 | G | 3600 | 5600-7100 | 2.05 ![]() |
0 | 1 | 0.12 | n50984da3 |
98 Jun. 19 | 2450984.48403 | G | 3600 | 4375-5375 | 2.05 ![]() |
0 | 1 | 0.12 | n50984db3 |
98 Jun. 20 | 2450984.52778 | G | 3600 | 5600-7100 | 2.05 ![]() |
0 | 1 | 0.12 | n50984da4 |
98 Jun. 20 | 2450984.52778 | G | 3600 | 4375-5375 | 2.05 ![]() |
0 | 1 | 0.12 | n50984db4 |
98 Jun. 20 | 2450984.80627 | C | 180 | 3655-7530 | 3.0 ![]() |
90 | 1.5 | n50984f | |
98 Jun. 20 | 2450985.29444 | L1 | 2400 | 5630-7250 | 8.0 ![]() |
0 | 2 | 9 | n50985llr1 |
98 Jun. 20 | 2450985.29444 | L1 | 2400 | 5630-7250 | 8.0 ![]() |
0 | 2 | 9 | n50985llr3 |
98 Jun. 20 | 2450985.36319 | L1 | 2400 | 4090-5800 | 8.0 ![]() |
0 | 2 | 9 | n50985llb1 |
98 Jun. 20 | 2450985.36319 | L1 | 2400 | 4090-5800 | 8.0 ![]() |
0 | 2 | 9 | n50985llb3 |
98 Jun. 20 | 2450985.39653 | L1 | 2400 | 5630-7250 | 8.0 ![]() |
0 | 2 | 9 | n50985llr2 |
98 Jun. 20 | 2450985.39653 | L1 | 2400 | 5630-7250 | 8.0 ![]() |
0 | 2 | 9 | n50985llr4 |
98 Jun. 20 | 2450985.40174 | G | 4500 | 5600-7100 | 2.05 ![]() |
0 | 1 | 0.12 | n50985da1 |
98 Jun. 20 | 2450985.40174 | G | 4500 | 4375-5375 | 2.05 ![]() |
0 | 1 | 0.12 | n50985db1 |
98 Jun. 20 | 2450985.43056 | L1 | 2400 | 4090-5800 | 8.0 ![]() |
0 | 2 | 9 | n50985llb2 |
98 Jun. 20 | 2450985.43056 | L1 | 2400 | 4090-5800 | 8.0 ![]() |
0 | 2 | 9 | n50985llb4 |
98 Jun. 20 | 2450985.45660 | G | 4500 | 5600-7100 | 2.05 ![]() |
0 | 1 | 0.12 | n50985da2 |
98 Jun. 20 | 2450985.45660 | G | 4500 | 4375-5375 | 2.05 ![]() |
0 | 1 | 0.12 | n50985db2 |
98 Jun. 20 | 2450985.51215 | G | 4500 | 5600-7100 | 2.05 ![]() |
0 | 1 | 0.12 | n50985da3 |
98 Jun. 20 | 2450985.51215 | G | 4500 | 4375-5375 | 2.05 ![]() |
0 | 1 | 0.12 | n50985db3 |
98 Jun. 21 | 2450985.70999 | C | 180 | 3655-7530 | 3.0 ![]() |
90 | 1.5 | n50985f | |
98 Jun. 22 | 2450986.68704 | C | 180 | 3654-7529 | 3.0 ![]() |
90 | 2.0 | n50986f | |
98 Jun. 23 | 2450987.69251 | C | 180 | 3654-7529 | 3.0 ![]() |
90 | 1.5 | n50987f | |
98 Jun. 24 | 2450988.65380 | C | 210 | 3654-7529 | 3.0 ![]() |
90 | 1.5 | n50988f | |
98 Jun. 25 | 2450989.64822 | C | 180 | 3655-7529 | 3.0 ![]() |
90 | 1.5 | n50989f | |
98 Jun. 25 | 2450990.42708 | L2 | 1800 | 4090-5800 | 2.0 ![]() |
0 | 2 | 8 | n50990llb |
98 Jun. 26 | 2450990.65554 | C | 180 | 3658-7532 | 3.0 ![]() |
90 | 1.5 | n50990f | |
98 Jun. 26 | 2450991.42708 | L2 | 1800 | 4090-5800 | 2.0 ![]() |
0 | 2 | 8 | n50991llb |
98 Jun. 27 | 2450991.67628 | C | 180 | 3657-7531 | 3.0 ![]() |
90 | 1.5 | n50991f | |
98 Jun. 28 | 2450992.68902 | C | 180 | 3657-7531 | 3.0 ![]() |
90 | 1.5 | n50992f | |
98 Jun. 29 | 2450993.67738 | C | 180 | 3656-7531 | 3.0 ![]() |
90 | 1.5 | n50993f | |
98 Jun. 29 | 2450994.32743 | W | 3600 | 4294-5594 | 3.0 ![]() |
90 | 2.2 | 7-8 | n50994wba |
98 Jun. 29 | 2450994.46063 | W | 3660 | 4298-5586 | 3.0 ![]() |
90 | 2.8 | 7-8 | n50994wbb |
98 Jun. 30 | 2450994.65447 | C | 180 | 3656-7531 | 3.0 ![]() |
90 | 1.5 | n50994f | |
98 Jun. 30 | 2450995.45581 | W | 5500 | 4298-5588 | 3.0 ![]() |
90 | 1.7 | 7-8 | n50995wb |
98 Jul. 01 | 2450995.66466 | C | 180 | 3659-7534 | 3.0 ![]() |
90 | 1.5 | n50995fa | |
98 Jul. 01 | 2450995.77620 | C | 180 | 3659-7534 | 3.0 ![]() |
90 | 1.5 | n50995fb |