1: \documentclass[12pt,preprint]{aastex}
2: %\documentclass[12pt,manuscript]{aastex}
3:
4: %\slugcomment{Submitted to ApJ (comments welcome)}
5:
6: \shorttitle{Variability of Narrow-Line Seyfert 1s}
7: \shortauthors{Klimek, Gaskell, \& Hedrick}
8:
9: \begin{document}
10:
11: \title{OPTICAL VARIABILITY OF NARROW-LINE SEYFERT 1 GALAXIES}
12:
13: \author{ELIZABETH S. KLIMEK, C. MARTIN GASKELL \& CECELIA H. HEDRICK}
14: \affil{Department of Physics \& Astronomy, University of Nebraska,
15: Lincoln, NE 68588-0111} \email{penumbra89@hotmail.com,
16: mgaskell1@unl.edu, piqueen314@hotmail.com}
17:
18: \begin{abstract}
19:
20: We present results of a broad-band photometric study of the
21: optical variability of six Narrow-Line Seyfert 1 (NLS1) galaxies
22: observed at 172 epochs. We searched for microvariability on 33
23: nights. Strong evidence for microvariability was found only for
24: our lowest luminosity object, NGC 4051, on one night. Weaker
25: evidence suggests such variability on a few other nights for two
26: other objects, but the data are not as convincing. Intra-night
27: variability in NLS1s is thus concluded to be rare and of low
28: amplitude. We give illustrations of how variable image quality
29: can produce spurious variability. We find that for well-studied
30: non-NLS1s there is a spread in the amplitude of seasonal
31: variability (i.e., in some years an AGN is more variable than in
32: others). We find that the means of the variability amplitudes of
33: non-NLS1s over several seasons vary from object to object (i.e.,
34: some AGNs are, on average, more variable than others). NLS1s also
35: show a spread in seasonal variabilities. The best-studied NLS1,
36: Ark 564, shows a range of amplitudes of variability from season to
37: season that is comparable to the range found in BLS1s, and in one
38: season Ark 564 was as variable as the most variable non-NLS1. The
39: seasonal amplitudes of variability for NLS1s are mostly in the
40: lower half of the range of non-BLS1 seasonal amplitudes, but the
41: absence of a suitable control sample makes a precise comparison
42: difficult. However, on long timescales (weeks to years) NLS1s as a
43: class are not {\it more} variable than non-NLS1s. The extreme
44: variability seen in the X-rays was not seen in the optical. This
45: has consequences for the models of AGNs in general as well as
46: NLS1s in particular.
47:
48: \end{abstract}
49:
50: \keywords{galaxies:active --- galaxies:quasars:general ---
51: X-rays:galaxies --- black hole physics --- accretion: accretion
52: disk}
53:
54: \section{INTRODUCTION}
55:
56: It has long been realized that some AGNs only have narrow emission
57: lines in their spectra while at the same time showing the
58: characteristic spectrum of the broad-line region (e.g., Phillips
59: 1977, Koski 1978, Davidson \& Kinman 1978). Thus the BLR (high
60: density) lines were narrower than usual. These objects came to be
61: called Narrow Line Seyfert 1s (NLS1s) (Gaskell, 1984; Osterbrock
62: \& Pogge, 1985, Goodrich 1989). Although there is no sharp
63: demarcation, an AGN is commonly call a NLS1 if the broad lines
64: have a FHWM $\le 2000$ km~s$^{-1}$. We will refer to AGNs with
65: FWHM $> 2000$ km~s$^{-1}$ as BLS1s (``broad-line Seyfert 1s'').
66: The properties of NLS1s lie at one end of a set of correlations
67: between AGN properties that is commonly called eigenvector 1
68: (Boroson and Green, 1992). In addition to the curiously narrow
69: ``broad'' lines, NLS1s tend to have steeper soft X-ray spectra
70: than non-NLS1s (Boller, Brandt \& Fink 1996). Many exhibit rapid
71: soft X-ray variability, which can also be large in amplitude, such
72: as a factor of $\sim 100$ in a day in IRAS 13224-3809 (Boller et
73: al. 1997). NLS1s show strong optical Fe II emission lines (Sargent
74: 1968) and sometimes the higher ionization Fe lines (Davidson \&
75: Kinman 1978), but while the Fe II equivalent widths are about the
76: same as those of non-NLS1s (Gaskell 1985), the H$\beta$ equivalent
77: widths are smaller than those of non-NLS1s (Gaskell 1985, Goodrich
78: 1989). The [OIII]/H$\beta$ line ratio is $< 3$, which is less than
79: the dividing line Shuder \& Osterbrock (1981) found between
80: Seyfert 1s and Seyfert 2s (i.e., the [OIII]/H$\beta$ ratio implies
81: that they are Seyfert 1s).
82:
83: A common trait possessed by all AGNs is that that they display
84: some degree of variability (Ulrich, Maraschi, \& Urry 1997;
85: Gaskell \& Klimek 2003). The most vigorous variability observed is
86: that of the X-rays (Mushotsky, Done, \& Pounds 1993). Variability
87: is valuable in that it provides clues to understanding what AGNs
88: actually are and how they work. Details of the inner workings of
89: AGNs remain poorly understood, despite over three decades of
90: research. Variability can help set constraints on the sizes of
91: different regions of AGNs and can give information about the
92: processes that are driving the variations. There has been much
93: effort in looking for optical variability, some of it in
94: conjunction with monitoring in other wavebands. If a link is
95: found between the variability in different wavebands, then the
96: processes behind each kind of variability are probably related.
97: Explanations of AGN variability include hotspots on accretion
98: disks, flares, and relativistic jets. BL Lac objects and
99: optically violently variable AGNs (OVVs) exhibit strong
100: variability on short timescales. This is believed to be a
101: consequence of relativistic beaming by jets. The beaming can
102: amplify intrinsic variations, which may or may not originate from
103: within the jet.
104:
105: Rapid (sub-diurnal) variability in the X-ray emission of AGNs is
106: well-known, but there have been conflicting claims about the
107: frequency of occurrence of rapid intra-night optical variability.
108: So far, observing campaigns searching for optical variability in
109: AGNs have typically yielded low amplitude variations on timescales
110: of no less than a few days (e.g., Webb \& Malkan 2000a). De
111: Ruiter \& Lub (1986) did not find any rapid ($< 1$ day) variations
112: of greater than 0.5 \% for any of the eight Seyfert galaxies they
113: observed, while other observers have reported low-level
114: microvariability on sub-diurnal timescales. For example,
115: Merkulova (2000) reported sub-diurnal variability at the 1\% level
116: on 60\% of nights for NGC 4151. Jang \& Miller (1995, 1997) found
117: that 8 out of 17 radio-quiet and 6 out of 7 radio-loud AGNs showed
118: intra-night variability on the order of a few percent (see also
119: Carini, Noble, \& Miller, 2003). Merkulova (2000) concluded that
120: intra-night variability is transient in character and has
121: manifested itself with different probabilities for different
122: galaxies. Even though there is some uncertainty over the
123: frequency of occurrence of sub-diurnal variability, it would seem
124: that extremely rapid and/or large amplitude optical variations are
125: rare.
126:
127: The most extreme X-ray variability in non-OVV AGNs is seen in
128: NLS1s. The shortest timescales for X-ray variability are about
129: 200 to 1000 s (Boller et al. 1993). Over a short period of time,
130: the amplitude can change by as much as a factor of 4, as seen in
131: the NLS1 IRAS 13224-3809, which varied by this much over a course
132: of hours (Boller et al. 1997). Large amplitude rapid X-ray
133: variability is one of the interesting properties of NLS1s. It has
134: been suggested that NLS1s are beamed (Boller et al. 1997), since
135: they display similarly strong variations in the X-rays.
136:
137: If rapid optical variations are present in AGNs other than BL
138: Lacs, it is possible that they would most likely be found in
139: objects displaying the most extreme X-ray variability, namely the
140: NLS1s. On long timescales Giannuzzo \& Stirpe (1996) and
141: Giannuzzo et al. (1997) compared the {\it Balmer-line} variability
142: of NLS1s with that of NGC 5548 and found the NLS1s to be less
143: variable. However, this was not a general comparison with
144: non-NLS1s, and it was possible that NGC 5548 was more variable
145: than the average BLS1. Young et al. (1999) unsuccessfully searched
146: for intra-night optical variability in IRAS 13224-3809, while
147: Miller et al. (2000) reported dramatic variability in the same
148: object on one night.
149:
150: The work presented here has two goals: (a) to look for evidence of
151: sub-diurnal optical continuum variations, especially those of
152: large amplitude, in a larger sample of NLS1s and (b) to see
153: whether NLS1s are more variable in the optical than non-NLS1s on
154: longer timescales. For one of the objects in our sample, Ark 564,
155: there are observations over a much longer period of time, so we
156: discuss this object at more length in a separate paper (Gaskell et
157: al. 2004).
158:
159: \section{OBSERVATIONS}
160:
161: The majority of the observations were made with the University of
162: Nebraska's 0.4-m Schmidt-Cassegrain telescope of the university's
163: Lincoln Observatory in a bright sky location. Images were
164: recorded with a Kodak KAF-0401 CCD, giving 0.887 arcseconds per
165: pixel through a f/5 focal reducer. All images were taken through
166: a standard Johnson V filter. Additional observations were made
167: with the 0.8-m Cassegrain at the university's Behlen Observatory
168: near Mead, Nebraska, in a darker sky location, with a Kodak
169: KAF-1001E chip giving 0.59 arcseconds per pixel with an f/9 focal
170: reducer through an identical filter. The NLS1s and the comparison
171: stars were measured through photometric apertures of identical
172: effective solid angles and we found that no scaling was necessary
173: between observations from the two telescopes.
174:
175: A sample of six NLS1s was chosen based only on the criteria that
176: they had to be bright enough to be observed with the 0.4-m
177: telescope and that they had to be high declination sources in
178: order to maximize the amount of time that they would be accessible
179: during the night. Thus, our selection was not based on any other
180: specific property or characteristic of this class of AGNs. In
181: particular, variability history was not taken into consideration.
182: The objects selected were Ark 564, Mrk 478, Mrk 493, Mrk 335, Mrk
183: 359, and NGC 4051.
184:
185: Observations for most objects were taken between May 2000 and June
186: 2003, with the exception of Ark 564, for which the observations
187: considered here began in August 1998. In searching for
188: intra-night variations, we observed objects continuously for most
189: of a night. Integration times were 10 to 15 minutes on the 0.4-m
190: telescope and 5 minutes on the 0.8-m telescope. Anywhere from
191: three to 26 images per object were obtained during any one night,
192: giving a range of 30 minutes to 6.5 hours of continuous data for
193: the most intensively observed objects.
194:
195: \section{DATA REDUCTION}
196:
197: \subsection{Comparison Stars}
198:
199: A summary of the adopted magnitudes for the comparison stars used
200: for the NLS1s (except Ark 564 –- see Gaskell et al. 2004) is given
201: in Table 1, and finding charts are shown in Figures 1 -- 5. Where
202: possible, comparison stars were chosen from the literature. For
203: Mrk 359, Mrk 493, \& Mrk 478 we calibrated our own comparison
204: stars from the count rates on a handful of our best nights. The
205: relative calibration of these stars is good, but the absolute
206: calibration is believed to be uncertain by ±0.2 magnitudes. This
207: uncertainty has no impact on this study, as we are only interested
208: in changes in magnitudes. In the case of Mrk 478, our magnitudes
209: agree with the relative magnitudes given by Webb and Malkan
210: (2000b).
211:
212: The detector noise is non-Gaussian because of hot pixels and
213: cosmic rays. If one comparison star gave anomalous magnitudes for
214: an image, it was not used for that image. Likewise, if the
215: average magnitude for an AGN was significantly off for an image on
216: a night, it was dropped from the nightly average. Between 5 \%
217: and 10 \% of comparisons were discarded. Visual inspection often
218: revealed obvious image problems in these cases.
219:
220: \subsection{Size of Photometric Aperture}
221:
222: In order to maximize the accuracy with which the counts from an
223: object are measured with respect to a sky background, the size of
224: the measuring aperture should be small in order to reduce the sky
225: background contribution, but at the same time, large enough to get
226: a good signal-to-noise ratio and to minimize the effect of
227: fluctuations in the measurement due to miscentering. A compromise
228: must then be chosen in order to obtain the best measurement.
229:
230: For point-sources in stellar photometry, these effects have been
231: studied by Howell (1989). For faint stellar sources, the optimum
232: aperture radius is about three pixels or about the FWHM. A number
233: of previous studies of AGN variability have followed the Howell
234: prescription (e.g., Carini et al, 1991; Jang \& Miller, 1995,
235: 1997), but, as Cellone, Romero, \& Combi (2000) pointed out, the
236: effects of the underlying host galaxy, especially in low
237: luminosity AGNs, should not be ignored. Because of the underlying
238: host galaxy even small seeing fluctuations can introduce spurious
239: variability in the AGN flux that can be mistaken for
240: microvariability. This is because poor seeing turns a point image
241: into a extended image, but a galaxy already has an extended image
242: and is therefore affected much less. This means that poor image
243: quality causes more light loss from a circular aperture for a star
244: than for a galaxy, and thus the galaxy will appear to be brighter
245: relative to the star. We provide illustrations of this effect
246: below. Similar conclusions in the spectroscopic case of choosing a
247: suitable aperture were found by Peterson et al (1995).
248:
249:
250: Since the ideal aperture is a function of the radial brightness
251: distribution of the galaxy and its brightness relative to the AGN,
252: we experimented with two aperture sizes to find the one most
253: appropriate for a given object. The larger aperture had a radius
254: of 9 pixels on the 0.4-m telescope, corresponding to 8-arcseconds,
255: and the smaller aperture had a radius of 5 pixels, or
256: 4.4-arcseconds. The same sky annulus was kept in both cases,
257: originally chosen to be appropriate for the larger aperture size,
258: in order to exclude most of the galaxy component from the sky
259: background measurement. The annulus had an inner radius of 13.5
260: pixels (12-arcseconds) and an outer radius of up to 22.5 pixels
261: (20-arcseconds).
262:
263: Each object was measured with both apertures and the resulting
264: errors were analyzed. We only give here the measurements
265: resulting from the aperture that produced the lowest rms variation
266: in the errors of the nightly means for each object. For all
267: objects but Mrk 478 (one of our faintest and most compact objects)
268: the larger aperture of 8-arcseconds was used. A summary of our
269: individual observations on nights we searched for microvariability
270: is presented in Tables 2 -- 6. We discuss the estimation of our
271: errors below. For other nights just the nightly means are given in
272: Table 7.
273:
274: \section{ANALYSIS}
275:
276: There are several sources that can give rise to spurious
277: variability. These sources include imperfect flat-fielding,
278: possible inconstancy of comparison stars, and changes in image
279: quality due to seeing and focus shift. The latter causes
280: fluctuation in the galaxy component measured (see above).
281:
282: \subsection{Estimation of Errors}
283:
284: Photon statistics alone underestimate errors so we used other
285: methods to estimate the errors. For the long-term light curves,
286: the error for each night was calculated by dividing the standard
287: deviation ($\sigma$) of the magnitudes by the square root of the
288: number of images (n) for the night. Nights with few images
289: typically have the greater error, but sometimes such nights had
290: fortuitously small errors (e.g., if two or three of the magnitudes
291: happened to be essentially the same). In these cases the
292: individual night's $\sigma$ was replaced by the median standard
293: deviation $\sigma_{med}$ for all of an object’s nights of
294: observation. The $\sigma_{med}$ was thus used as a more likely
295: estimator for calculating each of these night's errors in order to
296: prevent a serious underestimation of the errors.
297:
298: This method of estimating errors obviously cannot be used when
299: searching for intra-night variability, since it calculates the
300: errors under the assumption of no variability. Looking at the
301: differential light curves of stars comparable in brightness to the
302: AGN is a useful and important check but gives only a lower limit
303: to the error because the additional error due to the host galaxy
304: (see above) is not taken into consideration.
305:
306: In order to estimate the errors on nights that were searched for
307: intra-night variability, the following method was used. The
308: difference of adjacent magnitudes in the time series, $\Delta$$m_i
309: = m_i - m_{i-1}$, was used to find the point-to-point variation
310: during the night. The standard deviation of this difference was
311: then divided by the square root of two, since a difference of two
312: magnitudes was taken, to give the estimated error.
313:
314: If there is no variability during the course of the night, then
315: this error will be indistinguishable from the nightly standard
316: deviation. This was in fact the case for all of our intra-night
317: variability search nights, with the exception of the one night
318: where intra-night variations were detected for NGC 4051.
319:
320: The difference method could give erroneously large errors if any
321: rapid variations are present on a microvariable timescale of
322: $10 –- 15$ minutes. However, we consider this to be unlikely (see
323: below).
324:
325: \subsection{Searching For Variability}
326:
327: While it is difficult to rule out microvariability, it is easier
328: to detect or rule out variability on longer timescales (hours).
329: AGNs are already known to vary on timescales of more than a day.
330: Therefore, it can be reasoned that if AGNs show variability on a
331: smaller timescale of minutes, then they will show variability on
332: the intermediate timescale of hours.
333:
334: To search for variability on the timescale of a few hours, two
335: statistical tests were performed. First, the magnitudes for a
336: night were divided into first and second halves. A Student t-test
337: was used to evaluate the significance of differences in the mean
338: magnitude between the two equal halves of the night. The second
339: test was to evaluate the significance of the correlation
340: coefficient for the magnitudes versus time. A significant
341: correlation would mean a significant possibility that variations
342: occurred during a night. If either test gave a two-tailed
343: significance of $> 90$\% (i.e., a less than 10\% probability of
344: arising by chance), further investigation was conducted.
345:
346: The t-test and the correlation test are only sensitive to a
347: general trend during the night. Other types of variation would be
348: missed (e.g., a rise and fall or sinusoidal oscillations). To
349: search for general variations on a timescale of $30 -— 45$
350: minutes, an F-test was conducted to compare the variance of the
351: magnitudes during a night with the variance in their
352: point-to-point differences. This tells whether or not there is
353: significant correlated variability between exposures, which
354: correspond to $10 - 15$ minute intervals. Again, we conducted
355: further investigation if the F-test gave a significance of $>
356: 90$\%.
357:
358: \subsection{Systematic Errors}
359:
360: In the cases where the above-mentioned statistical tests showed a
361: high level of significance, we investigated the effects of
362: possible sources of systematic error.
363:
364: \subsubsection{Image-Quality Effects}
365:
366: As is well-known, good flat fielding is difficult to attain, even
367: when using sky flats. Variations of 1\% or more across the CCD
368: chip were not uncommon. However, this effect was minimized by
369: having several (typically four) comparison stars, by trying to
370: choose comparison stars that were close to the AGN, and by
371: positioning the AGN on the same part of the chip every night as
372: best as possible. Images that were found to have significant
373: flat-fielding problems were removed from our analysis. This
374: sometimes involved dropping entire nights in order to prevent
375: false detections of variations. Where there were signs of
376: possible microvariations, possible flat-fielding problems were
377: investigated by looking at the (x,y) position of the AGN on the
378: chip in addition to visual inspection. If the object ``wandered''
379: due to inconsistent positioning during a night, its position was
380: plotted versus time to see if its behavior correlated at all with
381: the object’s light curve.
382:
383: The effect of changes in the image quality due to seeing or focus
384: changes (see section 3) were examined by comparing the trend of
385: the average FWHM of the comparison stars during the night with the
386: object’s light curve over the same period of time. For each night
387: the AGN magnitudes were also plotted against seeing to further
388: check for any correlation. Such a correlation would signify that
389: any variations present were spurious. Two examples of this
390: analysis are given below.
391:
392: Only two of the objects, Mrk 359 and Mrk 478, showed a clear FWHM
393: dependence of the magnitudes. This dependence was only present in
394: the 4.4-arcsecond aperture. No effect was seen in the 8-arcsecond
395: aperture. Figs. 6 -- 10 show two examples of the seeing
396: dependence during a night of Mrk 359 and Mrk 478 observations.
397: Figs. 6 and 9 are plots of the light curves obtained by using a
398: 4.4-arcsecond aperture. Fig. 7 shows the average FWHM of the
399: comparison stars from image to image for Mrk 359. Figs. 8 and 10
400: confirm the correlations between the magnitudes of the AGNs and
401: the FWHMs. It can be seen that if the changes in image quality
402: were not taken into account, there would be a spurious detection
403: of variability. The statistical tests showed no significant
404: variation for the large apertures, while for Mrk 359, for example,
405: the Student t-test was highly significant for the smaller aperture
406: (98\% significance).
407:
408: Analysis of the errors for all the nights of observations for Mrk
409: 359 showed the least sub-diurnal variability for the larger
410: aperture, and this gave the smoothest long-term light curve.
411: However, in the case of Mrk 478, the smaller aperture gave the
412: most consistent errors between nights.
413:
414: \subsubsection{Constancy of Comparison Stars}
415:
416: The constancy of the comparison stars was checked by plotting the
417: magnitudes of each individual star with respect to the averages of
418: the other comparison stars for each night. For NGC 4051 Star 5
419: showed evidence of a slow change of 0.03 magnitudes over three
420: observing seasons and was therefore not used. For NGC 4051, for
421: which the only instance of intra-night variability was found (see
422: below), it was especially important to have dependably constant
423: stars. There was also slight possible variations for two
424: comparison stars of Ark 564 (see Gaskell et al. 2004) but these
425: are comparable to our measuring errors and have a negligible
426: effect on our results, since for Ark 564 we are averaging
427: comparisons with four stars.
428:
429:
430: \section{RESULTS}
431:
432: \subsection{Sub-Diurnal Timescales}
433:
434: All nights containing at least a dozen images of the same object,
435: amounting to at least three hours’ worth of continuous monitoring,
436: were searched for evidence of microvariability. The significance
437: was evaluated as described in 4.2. The results are shown in Table
438: 8. For each test we give the probability of the possible
439: variability arising by chance (i.e., the null hypothesis is no
440: variability). As can be seen, the F-test reveals no correlated
441: variation on the time scale of $\sim 30$ minutes. The Pearson
442: correlation coefficient test and the Student t-test reveal trends
443: with one-tailed significances of $> 90$\% on a time scale of
444: several hours on five nights. However, because we are looking at
445: 33 nights of monitoring, we would expect to get a significance of
446: 1 in 10 about three times.
447:
448: Allowing for the number of nights of observation, it is apparent
449: from Table 8 that there is only one night with a high probability
450: of microvariability: NGC 4051 on 2003 Feb 20. There are four
451: other nights where the correlation and t-tests indicate possible
452: variability at the $\sim 90 –- 95$\% significance level. Apparent
453: variations on these nights were checked for spurious causes.
454:
455: A 0.045 magnitude change was found in the light curve of NGC 4051
456: during the course of one night (see Fig. 11). Both the
457: correlation and t-tests gave greater than 99.6\% confidence that
458: the variations were not due to chance. We checked for possible
459: instrumental causes. Fig. 11 shows the behavior of NGC 4051 on
460: 2003 February 20 with respect to star 2, the brightest comparison
461: star, and with respect to the average of the other three
462: comparison stars. Both curves are in good agreement, attesting to
463: the reliability of the comparison stars. The larger-than-average
464: fluctuation in the third image is caused by a problem with star 2.
465: Such problems are not unusual (see 3.1) and are consistent with
466: our detector noise characteristics. Likewise, statistically there
467: could also be a problem with the AGN in one or both of images 8 or
468: 9.
469:
470: Figures 12 and 13 show plots of the y-position on the chip and the
471: FWHM as a function of time. Although there are slight changes in
472: the y-position during the night, comparable changes in the
473: position on the chip are seen on other nights without affecting
474: the magnitude. The two largest changes in the light curve (Fig.
475: 11) and the y-position (Fig. 12) on the chip occur simultaneously.
476: However, after this point the two plots behave differently. There
477: were no peculiarities found in either the flat-fielding or in the
478: dark subtraction. The x-position change (not shown) also shows no
479: detailed correlation with the light curve. The image quality
480: improved during the night, but the FWHM does not correlate in
481: detail with the shape of the light curve. The 8-arcsecond radius
482: aperture magnitudes show no correlation with FWHM on any other
483: night for NGC 4051 (or indeed for any other object). When the
484: 4.4-arcsecond radius aperture magnitudes do show a correlation
485: with FWHM (see Figs. 8 and 10), it is in the {\it opposite} sense
486: to what was found here. We are thus unable to find artificial
487: explanations of the apparent sub-diurnal variations of NGC 4051 on
488: 2003 February 20.
489:
490: With the exception of the one night for NGC 4051 we therefore find
491: very little evidence for variability on an intra-night timescale.
492: A study of microvariability in Seyferts (Carini, Noble, and
493: Miller, 2003) found only one Seyfert, the BLS1 Ark 120, in a
494: sample of eight, which showed signs of sub-diurnal activity. Their
495: result for the NLS1 Mrk 335 was consistent with the null result
496: obtained from our study. Jang and Miller (1997) also found no
497: evidence of microvariability in Mrk 335 in their study of
498: radio-quiet versus radio-loud quasars. Webb \& Malkan (2000a)
499: searched for intra-night variability in Seyfert 1s. No evidence
500: for such variability was found, but they typically had only two
501: images per night, compared with the more than twelve in this
502: study, and their observational errors were about ± 0.03
503: magnitudes. Their sample included Mrk 478, an object for which we
504: found no microvariations.
505:
506: \subsection{Longer Timescales}
507:
508: Long-term light curves from the nightly mean magnitudes given in
509: Table 7 are plotted in Figs. 14 -- 18. As can be seen,
510: variability is present for all six objects on a timescale of
511: months to years.
512:
513: In all photometric and spectroscopic studies of continuum
514: variability there is contamination from the host galaxy. This
515: varies with the size of the photometric or spectroscopic aperture
516: used. In order to make a legitimate comparison, our magnitudes
517: were scaled to small aperture sizes, which have been used in most
518: spectrophotometric studies such as those of the {\it International
519: AGN Watch}. This was done by taking some of our best images for
520: each object (i.e., images having the lowest FWHM) and measuring
521: the counts for successively smaller apertures. The ratio between
522: the original aperture we used and the smallest practical aperture,
523: 2.7 arcseconds, was found. This ratio was then used to scale our
524: data by removing the extra host galaxy contribution in the larger
525: aperture. This effectively gives a multiplicative correction
526: factor for the amplitude of variability. These scale factors are
527: given in Table 9. In Table 10 we give scaled seasonal standard
528: deviations for our six NLS1s and also seasonal standard deviations
529: for NGC 4051 from the International AGN Watch observations of
530: Peterson et al. (2000). In Table 11 we give the seasonal standard
531: deviations for eight non-NLS1 AGNs observed by the International
532: AGN Watch and others.
533:
534: A comparison of the seasonal standard deviations of the NLS1s with
535: the sample of BLS1s is given in Fig. 19. From this we note the
536: following:
537:
538: \begin{enumerate}
539:
540: \item For well-studied BLS1s (e.g., NGC 5548, NGC 4151, 3C 273),
541: there is a range in standard deviation from season to season.
542:
543: \item Some BLS1s appear on average to have higher seasonal
544: standard deviations than others.
545:
546: \item NLS1s also show a spread in seasonal standard deviations.
547:
548: \item The NLS1 seasonal standard deviations are in the lower-half
549: of the range of BLS1 standard deviations.
550:
551: \item The well-studied NLS1 Ark 564 shows a range of seasonal
552: standard deviations comparable to that of a BLS1 and in one season
553: was as variable as the most variable BLS1 (see Gaskell et al. 2004
554: for more detailed discussion).
555:
556: \end{enumerate}
557:
558: \section{DISCUSSION}
559:
560: \subsection{Short-Timescale Variability}
561:
562: Our study clearly shows that high-amplitude variability in NLS1s
563: is rare. As noted already, Miller et al. (2000) did however find
564: an example of large-amplitude, rapid optical variability in IRAS
565: 13224-3809. With this exception, NLS1s do not display rapid
566: high-amplitude variability in the optical such as is seen in BL
567: Lac objects or in the X-rays of NLS1s. Such short timescale
568: variability as there is must usually be of low amplitude. The
569: study of Ferrara et al. (2001) also supports this conclusion.
570: Interestingly, the one NLS1 in which we find reasonable convincing
571: evidence for intra-night variability is also the least luminous
572: AGN in our sample. It is therefore the one with the least massive
573: black hole (Peterson et al. 2000) and the one for which the most
574: rapid variability would be expected. It is also perhaps
575: interesting that this microvariability in NGC 4051 occurred when
576: there was a rapid (1-day) drop in the flux (see Fig. 14).
577:
578: Factors contributing to spurious artificial variations are
579: numerous, and great care must be taken to account for them. The
580: inhomogeneity of data in previous variability studies is a
581: problem. The errors must be well-determined, and all sources of
582: systematic error must not be ignored. It is important to reduce
583: errors because true variations could be hiding in the noise.
584: Convincingly detecting microvariations is no small task.
585:
586: Because of uncertainties in studies of non-NLS1 microvariability,
587: it is not yet possible to make as detailed a comparison of the
588: microvariability of NLS1s and BLS1s as we would like. As usual,
589: more data are needed to further strengthen our findings. With all
590: the factors that can lead to a false positive detection of
591: microvariability, it is especially beneficial to have multiple
592: observers to verify a detection of microvariability. This has
593: generally not happened in the past.
594:
595: \subsection{Long-Term Variability}
596:
597: From Fig. 19 it might seem that, on average, NLS1s are less
598: variable than BLS1s, but before such a conclusion can be made,
599: selection effects need to be understood. None of our objects were
600: chosen on the basis of optical variability characteristics, but
601: BLS1s are typically chosen for studies with a hope that they will
602: show high-amplitude variations. This selection is based on prior
603: variability history, and, as can be seen from Fig. 19, some
604: objects tend to be more variable than others. Thus, there is a
605: selection effect if the non-NLS1 sample consists primarily of the
606: most variable BLS1s, in which case the comparison between the
607: variabilities of the two classes is not a valid one. Giannuzzo \&
608: Stirpe (1996) and Giannuzzo et al. (1998) compared the Balmer-line
609: variability of NLS1s with that of NGC 5548 and Fig. 19 shows that
610: NGC 5548 is indeed more variable than the average BLS1.
611:
612: We believe that, because of the lack of an adequate control sample
613: of BLS1s, it is not yet possible to conclude that NLS1s are less
614: variable in the optical than BLS1s. There is no evidence that the
615: amplitude of variability of NLS1s and BLS1s is different.
616:
617: \subsection{Implications}
618:
619: That NLS1s vary in amplitude by so much so rapidly in the X-rays
620: without any similar behavior in the lower energy end of the
621: spectrum is a remarkable result and says something important about
622: the structure of AGNs. The fact that NSL1s probably do not vary
623: any differently than non-NLS1s in the visible continuum region is
624: in itself another interesting finding.
625:
626: These two findings rule out a situation where the X-rays and the
627: optical variations have a simple common origin. The source
628: producing the X-ray variability cannot be the same as that
629: producing the optical variations. If they both come from an
630: accretion disk, for example, then they must be in distinctly
631: different parts. If orientation plays a role in the NLS1
632: phenomenon, as it does through relativistic beaming, then only the
633: X-ray emission is being enhanced.
634:
635: Our results also do not support simple reprocessing, since the
636: optical band should respond in some manner to the X-ray changes.
637: That is, if the X-rays vary with a larger amplitude in NLS1s than
638: in BLS1s, then the optical variations, if present, should also
639: have a larger amplitude in NLS1s, even if the increase is small.
640: We know that in general NLS1s are highly variable in the X-rays,
641: constantly varying rapidly, (this is certainly true for one of our
642: objects, Ark 564, Turner et al. 2001, Shemmer et al. 2001, Edelson
643: et al. 2003) but we do not see any significant rapid optical
644: variations in NLS1s (with the exception of one night for one
645: object).
646:
647: \section{CONCLUSIONS}
648:
649: Our data show that, as a class, there is no evidence that NLS1s
650: behave any differently than non-NLS1s in terms of variability.
651: NLS1s can exhibit signs of variability on a timescale of hours,
652: but such events are rare and of low amplitude. They do not
653: exhibit the sort of remarkable variability seen in the X-rays.
654: Longer term variability over a timescale of days to months is the
655: norm, as is with non-NLS1s.
656:
657: \acknowledgments
658:
659: Financial support for this work was provided in part by the
660: University of Nebraska Layman Fund, the University of Nebraska
661: Undergraduate Creative Activities and Research Experiences
662: program, and by the National Science Foundation through grant AST
663: 03-07912. We are grateful to Beccie Grove and Matt Poulsen for
664: assistance in obtaining some of the observations.
665:
666: \begin{thebibliography}{}
667: \bibitem[]{666} Boller, T., Brandt, W. N., Fabian, A. C. \& Fink, H. H. 1996,
668: MNRAS, 289, 393
669:
670: \bibitem[]{669} Boller, T., Brandt, W. N., \& Fink, H. H. 1996, A\&A, 305, 53
671:
672: \bibitem[]{671} Boller, T., Truemper, J., Molendi, S., Fink, H., Schaeidt, S.,
673: Caulet, A., \& Dennefeld, M. 1993, A\&A, 279, 53
674:
675: \bibitem[]{672} Boroson, T. A. \& Green, R. F. 1992, ApJS, 80, 109
676:
677: \bibitem[]{674} Carini, M. T., Noble, J. C., \& Miller, H. R. 2003, AJ, 125,
678: 1811
679:
680: \bibitem[]{677} Cellone, S. A., Romero, G. E. \& Combi , J. A. 2000, AJ, 119,
681: 1534
682:
683: \bibitem[]{680} de Ruiter, H. R. \& Lub, J. 1986, A\&AS, 63, 59
684:
685: \bibitem[]{682} Edelson, R. Turner, T. J., Pounds, K., Vaughan, S.,
686: Markowitz, A., Marshall, H., Dobbie, P., \& Warwick, R. 2002, ApJ,
687: 568, 610
688:
689: \bibitem[] {684} Ferrara, E. C., Miller, H. R., McFarland, J. P., Williams, A. M., Wilson,
690: J. W., Fried, R. E., \& Noble, J. C. 2001, in Probing the Physics
691: of Active Galactic Nuclei, ed. Bradley M. Peterson, Richard W.
692: Pogge, \& Ronald S. Polidan, ASP Conference Proceedings, Vol. 224,
693: (San Francisco: Astronomical Society of the Pacific), 319
694:
695: \bibitem[]{686} Gaskell, C. M. 1984, ApL, 24, 43
696:
697: \bibitem[]{688} Gaskell, C. M. 1985, ApJ, 291, 112
698:
699: \bibitem[]{690} Gaskell, C. M., Doroshenko, V. T., Klimek, E. S.,
700: Campbell, J. S., Crowley, K. A., George, T. A., Goosmann, R. W.,
701: Grove, R., Hiller, M. E., Peterson, B. W., \& Poulsen, M. A. 2004,
702: in preparation
703:
704: \bibitem[]{695} Gaskell, C. M. \& Klimek, E. S. 2003, AAT, 22, 661
705:
706: \bibitem[]{697} Giannuzzo, M. E. \& Stirpe, G. M. 1996, A\&A, 314, 419
707:
708: \bibitem[]{699} Giannuzzo, M. E., Mignoli, M., Stirpe, G. M., \& Comastri, A.
709: 1998, A\&A, 330, 894
710:
711: \bibitem[]{702} Goodrich, R. W. 1989, ApJ, 342, 224
712:
713: \bibitem[]{704} Jang, M. \& Miller, H. R. 1995, ApJ, 452, 582
714:
715: \bibitem[]{706} Jang, M. \& Miller, H. R. 1997, AJ, 114, 565
716:
717: \bibitem[]{708} Koski, A. T. 1978, ApJ, 223, 56
718:
719: \bibitem[]{710} Merkulova, N. I. 2000, AJ, 119, 631
720:
721: \bibitem[]{712} Miller, H. R., Ferrara, E. C., McFarland, J. P., Wilson, J.
722: W., Daya, A. B., \& Fried,
723: R. E., 2000, New Astronomy Reviews, 44, 539
724:
725: \bibitem[]{716} Mushotzky, R. F., Done, C., \& Pounds, K. A. 1993,
726: ARA\&A, 31,717
727:
728: \bibitem[]{719} Osterbrock, D. E. \& Pogge, R. W. 1985, ApJ, 297, 166
729:
730: \bibitem[]{721} Peterson, B. M., Pogge, Richard W., Wanders,
731: Ignaz, Smith, Sean M., \& Romanishin, W. 1995, PASP, 107, 579
732:
733: \bibitem[]{721} Peterson, B. M., et al. 2000, ApJ, 542, 161
734:
735: \bibitem[]{723} Phillips, M. M. 1977, ApJ, 215, 746
736:
737: \bibitem[]{725} Shemmer, O., et al, 2001, ApJ, 561, 162
738:
739: \bibitem[]{727} Shuder, J. S. \& Osterbrock, D. E. 1981, ApJ, 250, 55
740:
741: \bibitem[]{729} Turner, T. J., Romano, P., George, I. M., Edelson, R., Collier, S. J., Mathur, S., \& Peterson, B.
742: M. 2001, ApJ, 561, 131
743:
744: \bibitem[]{732} Ulrich, M.-H., Maraschi, L., \& Urry, C. M. 1997,
745: AA\&A, 35, 445
746:
747: \bibitem[]{735} Webb, W. \& Malkan, M. A. 2000a, ApJ, 540, 652
748:
749: \bibitem[]{737} Webb, W. \& Malkan, M. A. 2000b, ApJS, 130, 165
750:
751: \bibitem[]{739} Young, A. J., Crawford, C. S., Fabian, A. C., Brandt, W. N.,
752: O'Brien, P. T. 1999,
753: MNRAS, 304, L46
754: \end{thebibliography}
755:
756: \clearpage
757: %TABLE 1
758: \begin{deluxetable}{lrrlrlrrrrrr}
759: \tablewidth{0pt} \tabletypesize{\small} \tablecaption{Comparison
760: Stars\label{tbl 1-A_a}} \tablehead{ \colhead{Object} &
761: \colhead{Star} & \colhead{V (magnitude)} & \colhead{Reference}}
762: \startdata
763: & & & \\
764: Ark 564 & 0 & 12.175 & Gaskell et al. (2004) \\
765: & 1 & 13.658 & \\
766: & 3 & 14.176 & \\
767: & 5 & 14.490 & \\
768: & & & \\
769: Mrk 478 & 2 & 14.255 & uncertain to +/- 0.2 \\
770: & 3 & 13.495 & \\
771: & & & \\
772: Mrk 493 & 2 & 13.107 & uncertain to +/- 0.2 \\
773: & 4 & 13.684 & \\
774: & 5 & 13.933 & \\
775: & 8 & 14.429 & \\
776: & & & \\
777: Mrk 335 & 4 & 13.679 & Shrader et al. (1990) \\
778: & 6 & 13.794 & \\
779: & & & \\
780: Mrk 359 & 1 & 13.473 & uncertain to +/- 0.2 \\
781: & 3 & 13.533 & \\
782: & 4 & 13.725 & \\
783: & 5 & 14.540 & \\
784: & & & \\
785: NGC 4051 & 1 & 14.010 & Penston, Penston, \& Sandage 1971 \\
786: & 2 & 11.120 & \\
787: & 4 & 13.390 & \\
788: & 5 & 14.760 & our calibration (not used) \\
789:
790: \enddata
791: \end{deluxetable}
792:
793:
794: \newpage
795: %TABLE 2
796: \begin{deluxetable}{lrrrrc}
797: \tablewidth{0pt}
798: %\tabletypesize{\scriptsize}
799: \tablecaption{V-Magnitudes During Nights Searched For
800: Microvariability [Electronic Table]} \tablehead{ \colhead{Mrk 335}
801: & \colhead{UT} & \colhead{JD}
802: & \colhead{V (magnitude)} & \colhead{Error} & \colhead{FWHM}\\
803: \colhead{ } & \colhead{ } & \colhead{ } & \colhead{ } &
804: \colhead{$\pm$ } & \colhead{(arcsec)}}
805: \startdata
806:
807: contact author for data
808:
809: \enddata
810:
811: \end{deluxetable}
812: \clearpage
813: %TABLE 3
814: \begin{deluxetable}{lrrrrcrrrrrr}
815: \tablewidth{0pt}
816: %\tabletypesize{\scriptsize}
817: \tablecaption{V-Magnitudes During Nights Searched For
818: Microvariability [Electronic Table]} \tablehead{ \colhead{Mrk 359}
819: & \colhead{UT} & \colhead{JD} & \colhead{V (magnitude)} &
820: \colhead{Error} & \colhead{FWHM}\\ \colhead{ } & \colhead{ } &
821: \colhead{ } & \colhead{ } & \colhead{$\pm$ } & \colhead{(arcsec)}}
822: \startdata
823:
824: contact author for data
825:
826: \enddata
827:
828: \end{deluxetable}
829: \clearpage
830: %TABLE 4
831: \begin{deluxetable}{lrrrrcrrrrrr}
832: \tablewidth{0pt}
833: %\tabletypesize{\scriptsize}
834: \tablecaption{V-Magnitudes During Nights Searched For
835: Microvariability [Electronic Table]} \tablehead{ \colhead{Mrk 478}
836: & \colhead{UT} & \colhead{JD} & \colhead{V (magnitude)} &
837: \colhead{Error} & \colhead{FWHM}\\ \colhead{ } & \colhead{ } &
838: \colhead{ } & \colhead{ } & \colhead{$\pm$ } & \colhead{(arcsec)}}
839: \startdata
840:
841: contact author for data
842:
843: \enddata
844:
845:
846: \end{deluxetable}
847: \clearpage
848: %TABLE 5
849: \begin{deluxetable}{lrrrrcrrrrrr}
850:
851: \tablewidth{0pt}
852: %\tabletypesize{\scriptsize}
853: \tablecaption{V-Magnitudes During Nights Searched For
854: Microvariability [Electronic Table]} \tablehead{ \colhead{Mrk 493}
855: & \colhead{UT} & \colhead{JD} & \colhead{V (magnitude)} &
856: \colhead{Error} & \colhead{FWHM}\\ \colhead{ } & \colhead{ } &
857: \colhead{ } & \colhead{ } & \colhead{$\pm$ } & \colhead{(arcsec)}}
858: \startdata
859:
860: contact author for data
861:
862: \enddata
863:
864:
865: \end{deluxetable}
866: \clearpage
867: %TABLE 6
868: \begin{deluxetable}{lrrrrcrrrrrr}
869: \tablewidth{0pt}
870: %\tabletypesize{\scriptsize}
871: \tablecaption{V-Magnitudes During Nights Searched For
872: Microvariability [Electronic Table]} \tablehead{ \colhead{NGC
873: 4051} & \colhead{UT} & \colhead{JD} & \colhead{V (magnitude)} &
874: \colhead{Error} & \colhead{FWHM}\\ \colhead{ } & \colhead{ } &
875: \colhead{ } & \colhead{ } & \colhead{$\pm$ } & \colhead{(arcsec)}}
876: \startdata
877:
878: contact author for data
879:
880: \enddata
881:
882: \end{deluxetable}
883:
884:
885:
886: \clearpage
887:
888: \begin{deluxetable}{lrrrrcrrrrrr}
889: %TABLE 7
890: \tablewidth{0pt}
891: %\tabletypesize{\scriptsize}
892: \tablecaption{Long-Term Light Curves [Electronic Table]}
893: \tablehead{ \colhead{Object} & \colhead{UT} & \colhead{JD} &
894: \colhead{Number of} & \colhead{V (magnitude)} & \colhead{Error} & \colhead{FWHM}\\
895: \colhead{ } & \colhead{ } & \colhead{ } & \colhead{Images } &
896: \colhead{ }& \colhead{$\pm$ } & \colhead{(arcsec)}}
897: \startdata
898:
899: contact author for data
900:
901: \enddata
902:
903:
904: \end{deluxetable}
905:
906: \clearpage
907: %TABLE 8
908: \begin{deluxetable}{llrrrc}
909: \tablewidth{0pt}
910: %\tabletypesize{\scriptsize}
911: \tablecaption{Results of Statistical Tests\label{tbl 1-A_b}}
912: \tablehead{ \colhead{Name} & \colhead{Night} & \colhead{t-test} &
913: \colhead{Pearson}& \colhead{F-test}& \colhead{Microvariability}}
914: \startdata
915: Ark 564 & 98-Oct-19 & 97.4 & 67.4 & 60.1 & \\
916: & 98-Nov-24 & 6.20 & 8.50 & 99.3 & ? \\
917: & 02-Oct-20 & 77.8 & 42.4 & 22.8 & \\
918: Mrk 335 & 02-Oct-13 & 66.1 & 83.4 & 83.0 & \\
919: & 02-Oct14 & 28.9 & 33.7 & 66.8 & \\
920: & 02-Oct-21 & 85.9 & 77.2 & 96.4 & \\
921: & 02-Oct-22 & 27.6 & 36.3 & 71.1 & \\
922: & 02-Nov-04 & 1.30 & 5.60 & 41.7 & ? \\
923: & 02-Nov-12 & 26.3 & 11.6 & 97.6 & \\
924: Mrk 359 & 02-Nov-07 & 57.0 & 81.0 & 72.4 & \\
925: & 02-Nov-25 & 50.4 & 49.6 & 56.8 & \\
926: & 02-Dec-10 & 52.7 & 39.0 & 68.8 & \\
927: & 03-Jan-13 & 98.1 & 28.0 & 49.5 & \\
928: Mrk 478 & 02-Mar-12 & 48.3 & 81.8 & 48.6 & \\
929: & 02-Apr-23 & 32.4 & 12.6 & 79.1 & \\
930: & 02-May-03 & 8.10 & 10.3 & 74.7 & \\
931: & 02-May-14 & 41.7 & 71.9 & 61.0 & \\
932: & 02-May-18 & 6.60 & 10.5 & 18.2 & \\
933: & 02-May-20 & 86.4 & 54.2 & 83.8 & \\
934: & 02-May-21 & 3.60 & 4.90 & 81.9 & ? \\
935: & 02-May-31 & 27.7 & 43.0 & 95.1 & \\
936: & 02-Jun-01 & 89.8 & 39.0 & 86.8 & \\
937: & 02-Jun-06 & 97.3 & 86.2 & 3.10 & \\
938: & 02-Jun-09 & 92.8 & 58.9 & 94.1 & \\
939: & 02-Jun-14 & 70.8 & 47.2 & 53.9 & \\
940: Mrk 493 & 02-Jun-27 & 8.30 & 7.30 & 77.2 & \\
941: & 02-Jun-28 & 47.8 & 6.60 & 82.1 & \\
942: & 03-Apr-09 & 37.6 & 11.0 & 96.9 & \\
943: NGC 4051 & 03-Feb-13 & 78.3 & 77.2 & 66.8 & \\
944: & 03-Feb-20 & 0.400 & 0.300 & 16.5 & Yes \\
945: & 03-Feb-21 & 39.6 & 5.90 & 26.2 & \\
946: & 03-Mar-06 & 62.2 & 71.9 & 68.8 & \\
947: & 03-Mar-14 & 16.7 & 9.90 & 73.9 & \\
948:
949: \enddata
950:
951: \end{deluxetable}
952:
953: \clearpage
954: %TABLE 9
955: \begin{deluxetable}{lr}
956: \tablewidth{0pt}
957: %\tabletypesize{\scriptsize}
958: \tablecaption{Flux Ratios Scale Factors} \tablehead{
959: \colhead{Object} & \colhead{Scale Factor}} \startdata
960:
961: Mrk 335 & 1.14 \\
962: Mrk 359 & 1.93 \\
963: Mrk 478 & 1.09 \\
964: Mrk 493 & 1.38 \\
965: NGC 4051 & 1.55 \\
966: Ark 564 & 1.24 \\
967:
968: \enddata
969:
970: \end{deluxetable}
971:
972:
973: \clearpage
974: %TABLE 10
975: \begin{deluxetable}{lrrrrrrr}
976:
977: \tablewidth{0pt}
978: %\tabletypesize{\scriptsize}
979: \tablecaption{Root-Mean-Square Seasonal Variability of NLS1s}
980:
981: \tablehead{\colhead{ } & \colhead{NGC} & \colhead{NGC} &
982: \colhead{Ark} & \colhead{Mrk} & \colhead{Mrk} & \colhead{Mrk}&
983: \colhead{Mrk}\\
984:
985: \colhead{ } & \colhead{4051} & \colhead{4051} & \colhead{564} &
986: \colhead{335} & \colhead{359} & \colhead{478}& \colhead{493}\\}
987:
988: \startdata
989:
990: & 0.036 & 0.014 & 0.025 & 0.023 & 0.026 & 0.025 & 0.050 \\
991: & 0.045 & & 0.054 & & & & \\
992: & 0.048 & & 0.147 & & & & \\
993: & & & 0.045 & & & & \\
994: & & & 0.014 & & & & \\
995: & & & & & & & \\
996: Ave (scaled) & 0.043 & 0.022 & 0.071 & 0.026 & 0.050 & 0.027 & 0.069 \\
997:
998: \enddata
999:
1000:
1001: \end{deluxetable}
1002:
1003:
1004:
1005: \clearpage
1006: %TABLE 11
1007: \begin{deluxetable}{lrrrrrrrrr}
1008:
1009: \tablewidth{0pt}
1010: %\tabletypesize{\scriptsize}
1011: \tablecaption{Root-Mean-Square Seasonal Variability of BLS1s}
1012: \tablehead{\colhead{ } & \colhead{NGC} & \colhead{NGC} &
1013: \colhead{NGC} & \colhead{NGC} & \colhead{Mrk} & \colhead{Mrk}&
1014: \colhead{3C390.3} & \colhead{3C 273 } &
1015: \colhead{Average}\\\colhead{ } & \colhead{3783} & \colhead{4151} &
1016: \colhead{5548} & \colhead{7469} & \colhead{ 279} & \colhead{509}&
1017: \colhead{ } & \colhead{ } & \colhead{}\\\colhead{ } &
1018: \colhead{IAW} & \colhead{IAW} & \colhead{IAW} & \colhead{IAW}&
1019: \colhead{IAW} & \colhead{IAW} &
1020: \colhead{see ref}& \colhead{see ref} & \colhead{ }\\
1021: \colhead{Ref*} & \colhead{a} & \colhead{b} & \colhead{c} &
1022: \colhead{d}& \colhead{e} & \colhead{f} & \colhead{g}& \colhead{h}
1023: & \colhead{ }}
1024: \startdata
1025:
1026: & 0.085 & 0.032 & 0.137 & 0.026 & 0.087 & 0.132 & 0.061 & 0.030 & \\
1027: & & 0.080 & 0.136 & & 0.079 & 0.076 & 0.151 & 0.016 & \\
1028: & & 0.090 & 0.099 & & & 0.142 & 0.107 & 0.059 & \\
1029: & & 0.041 & 0.172 & & & 0.087 & & 0.073 & \\
1030: & & 0.106 & 0.100 & & & 0.086 & & 0.026 & \\
1031: & & 0.037 & 0.112 & & & & & 0.082 & \\
1032: & & & 0.085 & & & & & 0.040 & \\
1033: & & & 0.161 & & & & & 0.067 & \\
1034: & & & 0.113 & & & & & 0.014 & \\
1035: & & & 0.107 & & & & & 0.023 & \\
1036: & & & 0.165 & & & & & 0.032 & \\
1037: & & & 0.173 & & & & & 0.077 & \\
1038: & & & 0.130 & & & & & & \\
1039: & & & & & & & & & \\
1040: Ave & 0.085 & 0.064 & 0.130 & 0.026 & 0.083 & 0.104 & 0.106 & 0.045 & 0.078 \\
1041:
1042:
1043: \enddata
1044:
1045:
1046:
1047:
1048: \end{deluxetable}
1049:
1050:
1051:
1052:
1053: \clearpage
1054:
1055: \begin{figure}
1056: \plotone{f1_sm.ps} \caption{Comparison stars for Mrk 335}
1057: \end{figure}
1058: \begin{figure}
1059: \plotone{f2_sm.ps} \caption{Comparison stars for Mrk 359}
1060: \end{figure}
1061: \begin{figure}
1062: \plotone{f3_sm.ps} \caption{Comparison stars for Mrk 478}
1063: \end{figure}
1064: \begin{figure}
1065: \plotone{f4_sm.ps} \caption{Comparison stars for Mrk 493}
1066: \end{figure}
1067: \begin{figure}
1068: \plotone{f5_sm.ps} \caption{Comparison stars for NGC 4051}
1069: \end{figure}
1070: \begin{figure}
1071: \plotone{f6_sm.EPS} \caption{Apparent microvariability of Mrk 359
1072: when measured with a 4.4 arcsecond aperture on 2002 December 10.}
1073: \end{figure}
1074: \begin{figure}
1075: \plotone{f7_sm.EPS} \caption{Variation of image quality for Mrk
1076: 359 on 2002 December 10.} \end{figure}
1077: \begin{figure}
1078: \plotone{f8_sm.EPS} \caption{Variation of measured magnitude with
1079: image quality for Mrk 359 when measured with a 4.4 arcsecond
1080: aperture.} \end{figure}
1081: \begin{figure}
1082: \plotone{f9_sm.EPS} \caption{Apparent microvariability of Mrk 478
1083: on 2002 May 18 when measured with a 4.4 arcsecond aperture.}
1084: \end{figure}
1085: \begin{figure}
1086: \plotone{f10_sm.EPS} \caption{Variation of measured magnitude with
1087: image quality for Mrk 478 on 2002 May 18.} \end{figure}
1088: \begin{figure}
1089: \plotone{f11_sm.EPS} \caption{V-band light curve for NGC 4051 on
1090: 2003 February 20. The solid circles are the magnitudes measured
1091: relative to just star 2; the open circles are the magnitudes
1092: relative to the other comparison stars.}
1093: \end{figure}
1094: \begin{figure}
1095: \plotone{f12_sm.EPS} \caption{Variation of the y-position of the
1096: nucleus of NGC 4051 on the chip during the night of 2003 February
1097: 20.}
1098: \end{figure}
1099: \begin{figure}
1100: \plotone{f13_sm.EPS} \caption{Variation of the image quality for
1101: images of NGC 4051 during the night of 2003 February 20.}
1102: \end{figure}
1103: \begin{figure}
1104: \plotone{f14_sm.EPS} \caption{Long-term light curve for NGC 4051.}
1105: \end{figure}
1106: \begin{figure}
1107: \plotone{f15_sm.EPS} \caption{Long-term light curve for Mrk 335.}
1108: \end{figure}
1109: \begin{figure}
1110: \plotone{f16_sm.EPS} \caption{Long-term light curve for Mrk 359.}
1111: \end{figure}
1112: \begin{figure}
1113: \plotone{f17_sm.EPS} \caption{Long-term light curve for Mrk 478. }
1114: \end{figure}
1115: \begin{figure}
1116: \plotone{f18_sm.EPS} \caption{Long-term light curve for Mrk 493.}
1117: \end{figure} \clearpage
1118: \begin{figure}
1119: \epsscale{.85} \plotone{f19_sm.EPS} \caption{Distributions of
1120: root-mean-square V-band variability amplitudes for NLS1s (open
1121: squares) and BLS1s (hashed squares). For individual objects the
1122: seasonal variability is shown. For the NLS1 average and BLS1
1123: average each square represents the mean of the variability over
1124: the seasons given in Tables 10 and 11.}
1125: \end{figure}
1126:
1127: \clearpage
1128:
1129: Fig. 19.--- Distributions of root-mean-square V-band variability
1130: amplitudes for NLS1s (open squares) and BLS1s (hashed squares).
1131: For individual objects the seasonal variability is shown. For the
1132: NLS1 average and BLS1 average each square represents the mean of
1133: the variability over the seasons given in Tables 10 and 11.
1134:
1135: \end{document}
1136: