: 
; 
; 
: 
; 
; 
.
and 
data
since these were measured at the JKT only. These differences in the
magnitudes and colours of SAO91568 are within the errors expected in the
three data sets and so are unlikely to indicate systematic differences.
Intercomparison of the mean results for SAO91577, however, indicates a large
discrepancy between SO and JKT in respect of (U-B) colour
(
). This is uncomfortably large to be
attributed to photometric errors. It is possible, of course, that the bluer
photometric transformations, particularly U, are uncertain at the extreme
red end of their range. On the other hand the mean (U-B) colours for
II Peg itself are in good agreement (cf. Fig. 1 (click here)). At
present this anomaly is unresolved.
Figure 1: Nightly mean V and (B-V), (U-B), and
light curves for II Peg in September 1992. The filled
squares represent data from JKT in September, the open squares those
from Stephanion Observatory in July, the open triangles those from the
same source in September and the stars those from the Bulgarian
National Observatory. The point bracketted and labelled "F" is that measured
during the flare of 12 September
The V light curve, as presented in Fig. 1 (click here), is a
superposition of the data from all three sources. Its scatter is larger
than might be expected from the intercomparison of the mean magnitudes for
SAO91568. It can, nevertheless, be readily seen that there is a minimum
near phase 
0.03 and a maximum near 0.4. Thus the
light curve is clearly asymmetric, with a much steeper rise than fall.
However, there is a relatively slow fall between 
,
followed by a much more rapid decline to minimum. The amplitude of the
modulation, 
is 
.
The scatter in the B-V curve is reasonably uniform with phase and is of
amplitude 
. Overall, the B-V curve shows
evidence of a low-level modulation (
) in phase with
the V variation.
The U-B curve shows a uniform scatter with amplitude of
apart from one point which stands out from the
rest (near 
). This is associated with the large optical
flare on 5 September which will be discussed further below
(Sect.3.2 (click here)). It is indicated in both the U-B and B-V
colour curves and in the V light curve by an "F". We note that the
scatter in the (U-B) measurements is much larger than would be expected on
the basis of the uncertainty in the measurements themselves (i.e.
0.03). This may be due to low-level flaring as has been suggested in
many active late-type stars (see e.g. Byrne 1983 and references
therein).
Both the near-IR colour curves mirror the V variation accurately in
phase. V-R, however, has an amplitude of only 
,
while 
. Both are again in phase with V,
consistent in a general sense, with the spot origin of the variations.
Figure 2: The J magnitude and (J-H) and (J-K) colour curves for II Peg
at the end of July and the first half of August 1992 as observed at the
Teide Observatory of the IAC. Phases have been calculated according to the
ephemeris of Vogt (1981), i.e.
The light curves for the IR JHK bands will be found in
Fig. 2 (click here) in the form of J magnitude and (J-H) and (J-K)
colour diagrams. The J light curve is similar to V but with an
amplitude of 
. There is no evidence of systematic
variation in either of the IR colours themselves, but, given the different
amplitudes in V and J, there is a strong variation in (V-J). Because,
however, simultaneous V and JHK measurements are not available, it is
not possible to get a direct measure of the (V-J) colour.
Figure 3: Light curve of the large optical flare observed in the U-band at
Stephanion Observatory on 5 September 1992
Two optical U-band flares were recorded, details of which will be found in Table 8 (click here). The light curve of the largest of these will be found in Fig. 3 (click here).
| Date | UT |  U | ED | EU |
| 1992 | max | s | erg | |
| 05 Sep. | 20:39 | 0.32* | >3925 | >1.75 10 |
| 12 Sep. | 21:49 | 0.29 | 59 | 2.63 10 |
We have used the Equivalent Duration (ED) method of Gershberg (1972) in calculating the total, time-integrated energy in each flare. ED is defined as the time during which the quiescent star emits the same energy as the flare in the same passband. Thus the measurement of the flare's energy is referred to the quiescent pre-flare level as measured immediately preceding the flare itself. This avoids many of the uncertainties involved in the absolute calibration of the flare observations.
The energy of the flare is thus defined as
where 
is the quiescent energy emitted in the U band per
second. This, in turn, is related to the star's quiescent U magnitude
(taken as 9.3, see Table1 (click here)) by
where d is the distance to the star, taken as 29 pc (Strassmeier
et al. 1993), and 
is the conversion factor from U magnitude
to energy, taken as 
(Bessell 1979). This yields 
. Deriving the energy of each flare in
absolute terms is then a relatively simple procedure, the results of which
will be found in Table 8 (click here).
The resulting spectra were then corrected to the rest frame of the K star
primary of II Peg. This was achieved by shifting a strong, isolated
photospheric line (
) to its laboratory
wavelength. The spectra were then normalized by fitting a spline function to
the local continuum and dividing through by this spline. The points at
which the spline was fit were those judged to be free of lines when
comparison was made with spectra of a number of slowly rotating K giants,
taken with the same instrument and broadened to II Peg's rotational
(
, Vogt (1981)).
Figure 4 (click here) shows the mean of all the H spectra.
Figure 4: The overall mean profile of II Peg's H emission line derived
from four nights' data (upper solid curve). Note that all spectra have
been wavelength corrected to the rest frame of the K-star primary and the
vertical line gives the rest wavelength of the H line. The
dashed curve gives the ratio of the mean 1992 spectrum to that for 1991 taken
from Paper I and displaced downward to aid visibility. Both spectra have been
smoothed by a gaussian of 
, less than the nominal
resolution of the spectrograph
From Fig. 4 (click here) it will be seen that the mean H line is
strongly in emission with a peak intensity 1.45 times that of the local
continuum and has a 
. Furthermore, it is
asymmetric, in the sense that it shows an "absorption reversal'' whose red
peak is depressed relative to the blue peak, and a blue wing which exceeds
the red in total flux.
emission line were extracted and analyzed similarly to those at
H (Sect. 3.3 (click here)). They are shown after correction to
the rest frame of the K-star primary in Fig. 5 (click here) in the form of the
overall mean spectrum derived from the data on all three nights of
observation.
Figure 5: The overall mean profile of II Peg's H emission line derived
from all three nights' data (upper curve). The dashed curve gives
the mean 1991 H profile from Paper I which have been displaced downward
to aid visibility. Note that all spectra have been wavelength corrected to
the rest frame of the K-star primary and the vertical line gives
the rest wavelength of the H line
As remarked in Paper I the line is blended with nearby photospheric lines of
and 
on the red
side and 
on the blue. It is possible to
"see'' the lines to the red but the line to the blue is inextricably
blended with H and impossible to deblend without recourse to either
synthetic spectra or inactive templates. However, it is apparent that the
H line is "filled in'' and is asymmetric to the blue, in the sense
that it is more "filled in'' on that side of line centre.
line were extracted and analyzed as were those for H
(Sect. 3.3 (click here)). The resultant spectra are shown in
Fig. 6 (click here) in the form of the overall mean spectrum derived from
the data on both nights of observation. The line is in net weak emission
with a peak intensity of 
of the continuum level.
Figure 6: The overall mean profile of II Peg's HeI D line derived
from three nights' data (upper curve). The dashed curve gives the
mean 1991 HeI D profile from Paper I which have been displaced
downward to aid visibility. Note that all spectra have been wavelength
corrected to the rest frame of the K-star primary and the vertical lines
give the rest wavelength of the HeI D doublet
These spectra were extracted in hardware in real-time at the telescope to give
1-dimensional data, which were later debiased, flat-fielded and wavelength
calibrated. The results of the low-resolution H monitoring will be
found in Fig. 7 (click here) as a plot of EW(H) against time.
Figure 7: The EW of the II Peg's H emission as a function of Julian
Date measured at the Univeristy of Birmingham's Wast Hills Observatory between
7-20 September 1992. Proposed flares are indicated by bracketts around
the flaring points
20% in all cases.
Figure 8: Blue region spectra of II Peg taken at the Isaac Newton Telescope
14-19 September 1992 shifted to the rest frame of the K star. The top
panel shows the five clear nights' spectra in the region of the CaII 
lines as the overall mean spectrum. The middle and lower panels
give the same data in the vicinity of the Balmer H
and the Balmer
H
lines. Also shown are spectra in these same spectral regions taken
in 1991 (Paper I) smoothed to match the spectral resolution of the present
data (dashed curves). The vertical lines indicate the rest
wavelengths of the Balmer and CaII lines
We note the following general characteristics. The CaII and Balmer
H
lines are strongly in emission. However, we have examined
carefully the spectra in the region of the Balmer H or H lines
(marked by vertical lines in Figs. 8 (click here)) and find that there is no
obvious evidence of either, whether in emission or absorption.
The IUE spectra were extracted from the spectral images and then wavelength and flux calibrated using the program IUEDR (Giddings 1983) which is available on the UK STARLINK astronomical computing network (Bromage 1984). Subsequent analysis was performed using the STARLINK program DIPSO (Howarth & Murray 1987).
). The line fluxes
for the most prominent SWP emission lines resulting from this procedure will
be found tabulated in Tables 9 (click here).
| JD | Phase | OI | CII | CIV | HeII | CI | AlII | SiII | |
|
mid-exp |  1302/5/6 Å | 1335/6 Å |
1548/51 Å | 1640 Å |  |
1671 Å | 1808/17/8 Å | ||
| 2440000.0+ | Line Flux at Earth ( ) | ||||||||
| 8871.327 | 0.183 | 6.9 | 12.1 | 37.2 | 15.4 | 5.3 | 2.7 | 1.8 | 3.4 |
| 8871.411 | 0.196 | 6.7 | 12.2 | 25.3 | 15.0 | 5.6 | 2.9 | 1.4 | 2.9 |
| 8872.380 | 0.340 | 4.6 | 4.8 | 5.3 | 4.6 | 3.1 | 1.1 | 1.4 | 2.2 |
| 8873.368 | 0.487 | 3.2 | 3.7 | 3.9 | 3.3 | 2.1 | 1.1 | 1.0 | 1.6 |
| 8873.441 | 0.498 | 2.6 | 2.5 | 4.7 | 2.9 | 1.6 | 1.1 | 0.7 | 1.2 |
| 8874.221 | 0.614 | 3.2 | 4.5 | 4.2 | 4.3 | 3.0 | 2.2 | 1.3 | 1.7 |
| 8874.328 | 0.630 | 2.4 | 2.4 | 3.7 | 4.0 | 2.8 | 1.1 | 1.0 | 1.2 |
| 8874.424 | 0.644 | 3.2 | 4.6 | 5.9 | 4.3 | 3.0 | 0.7 | 1.4 | 1.8 |
| 8875.398 | 0.789 | 3.3 | 4.8 | 9.8 | 6.6 | 2.5 | 1.2 | 1.7 | 2.5 |
| 8876.335 | 0.928 | 3.9 | 3.3 | 5.0 | 4.6 | 2.7 | 1.0 | 1.2 | 2.0 |
| 8876.440 | 0.944 | 3.4 | 2.8 | 4.9 | 3.7 | 1.9 | -- | 1.3 | 2.3 |
| 8877.340 | 0.078 | 3.5 | 4.8 | 4.8 | 4.1 | 3.0 | 0.6 | 1.4 | 2.2 |
| 8877.440 | 0.093 | 2.7 | 2.8 | 3.0 | 3.7 | 2.8 | 0.7 | 1.0 | 1.8 |
| 8878.344 | 0.227 | 3.7 | 4.3 | 5.6 | 3.5 | 3.1 | -- | 1.4 | 2.3 |
| 8878.431 | 0.240 | 4.8 | 6.2 | 9.8 | 5.1 | 3.3 | 1.4 | 1.5 | 1.9 |
| 8879.347 | 0.376 | 4.3 | 3.5 | 5.1 | 3.0 | 2.5 | 0.8 | 1.2 | 1.8 |
| 8879.433 | 0.389 | 3.9 | 4.8 | 5.8 | 4.2 | 2.7 | 0.7 | 1.4 | 2.2 |
| 8880.319 | 0.521 | 3.5 | 4.2 | 5.1 | 3.3 | 2.2 | 0.8 | 1.2 | 2.0 |
| 8880.418 | 0.536 | 3.6 | 3.8 | 4.2 | 3.5 | 2.6 | 0.7 | 1.3 | 1.7 |
| 8881.303 | 0.667 | 3.2 | 2.3 | 5.1 | 3.9 | 2.0 | 0.8 | 1.5 | 1.6 |
| 8881.430 | 0.686 | 3.7 | 3.4 | 5.7 | 3.7 | 2.7 | -- | 1.7 | 3.0 |
| 8882.343 | 0.822 | 1.3 | 1.3 | 2.9 | 1.9 | 1.4 | -- | 0.8 | 1.3 |
| 8882.430 | 0.835 | 3.2 | 4.0 | 6.3 | 4.3 | 2.9 | -- | 1.3 | 2.1 |
resonance doublet (
) and some
lines of the FeII UV1 multiplet near 
. In both
cases the emission lines were resolved. Furthermore, in the case of the
MgII resonance doublet there is appreciable interstellar (IS)
absorption superimposed on the stronger and broader stellar emission.
Finally, the wavelength calibration of IUE in HIRES may be subject to
uncertainties of order 
(Byrne et al.
1989).
We have adopted the following procedure for deriving the flux of the MgII lines. We have assumed that the IS line is unresolved at the resolution of IUE HIRES and fitted it with a gaussian profile of fixed FWHM equal to that of the instrumental profile of the IUE spectrograph in HIRES mode, while simultaneously fitting the main emission profile with another gaussian whose central wavelength, FWHM and intensity are all free to vary. This procedure fits the IS line well but it is clear that there is excess flux in the wings of the main stellar emission line over and above a gaussian.
After this first round of fits the IS feature was used as a fiducial to bring the entire set of spectra to a common wavelength scale and the stellar emission fitted with the sum of two gaussians, one to represent the main body of the emission and the other to represent the wings. These fits have been used to estimate the total flux in the stellar emission line. The result will be found in Table10 (click here).
The FeII lines were measured by fitting a number of gaussians of fixed wavelength separation, corresponding to the laboratory separation of the UV1 lines. Their FWHM and intensities were, on the other hand, allowed to vary freely. The resulting measured line fluxes will also be found in Table 10 (click here).
| JD | Phase | MgII | Fe II | |||||
| mid-exp | 2795.5 Å | 2802.7 Å | 2620.7 Å | 2621.7 Å | 2625.7 Å | 2628.3 Å | 2631.1 Å | |
| 2440000.0+ |
 |
 | ||||||
| 8871.357 | 0.188 | 10.90 | 9.08 | 2.71 | 3.90 | 8.45 | 4.57 | 8.83 |
| 8872.328 | 0.332 | 7.13 | 5.05 | 3.13 | 1.78 | 4.31 | 2.33 | 2.98 |
| 8872.431 | 0.348 | 5.92 | 4.61 | 2.49 | 1.34 | 3.80 | 1.37 | 3.03 |
| 8873.316 | 0.479 | 5.85 | 5.47 | 3.41 | 2.56 | 3.20 | 1.88 | 2.25 |
| 8873.418 | 0.495 | 5.93 | 4.62 | 3.33 | 1.04 | 4.38 | 2.04 | 3.68 |
| 8874.171 | 0.606 | 5.40 | 4.68 | 1.85 | 1.66 | 3.08 | 2.03 | 2.76 |
| 8874.277 | 0.622 | 5.67 | 5.23 | 1.96 | 0.51 | 2.66 | 1.63 | 2.22 |
| 8874.381 | 0.638 | 5.45 | 5.67 | 2.46 | 1.84 | 3.69 | 2.63 | 1.64 |
| 8875.347 | 0.781 | 5.81 | 5.30 | 2.58 | 1.31 | 2.76 | 2.17 | 2.39 |
| 8875.443 | 0.796 | 5.68 | 5.29 | 2.55 | 2.17 | 3.18 | 2.94 | 4.57 |
| 8876.392 | 0.937 | 5.01 | 4.39 | 2.59 | 1.00 | 3.29 | 2.19 | 3.63 |
| 8877.390 | 0.085 | 5.90 | 4.92 | 1.75 | 1.30 | 3.64 | 2.19 | 3.64 |
| 8878.379 | 0.232 | 5.69 | 5.00 | 1.90 | 1.47 | 3.91 | 1.27 | 4.10 |
| 8879.390 | 0.383 | 5.86 | 5.02 | 1.93 | 0.97 | 3.87 | 1.64 | 3.17 |
| 8880.359 | 0.527 | 5.47 | 4.67 | 0.81 | 1.18 | 2.53 | 1.86 | 2.85 |
| 8881.383 | 0.679 | 4.75 | 4.17 | 1.83 | 1.14 | 3.04 | 1.66 | 3.42 |
| 8882.385 | 0.828 | 5.03 | 4.52 | 1.51 | 1.40 | 3.32 | 1.69 | 2.89 |
The data from each scan was vector integrated over the whole scan period and a plot of these scan averages will be found in Fig. 9 (click here). The star is detected at all times of observation and shows evidence of continuous variability in its flux at 5 GHz at all time scales examined by the data. This is true both on an hourly time scale and from night-to-night. A strong flare with a peak flux density of 15 mJy was observed on September 13.
Figure 9: The time sequence of BBI 5GHz observations of II Peg. The flaring
points discussed in the text are labelled with an "F"