Non-Periodic Phenomena in Variable Stars
IAU Colloquium, Budapest, 1968
VARIATIONS IN THE CONTINUOUS SPECTRUM OF GAMMA CAS
N. L. IVANOVA, I. D. KUPO and A. CH. MAMATKAZINA
Alma Ata Observatory, USSR
After an instability period late in the 30es some more or less long runs
of observations of gamma Cas, giving an idea of the behaviour of the continuous
and line spectrum of the star at different years (Vandekerkhove 1967, 1961;
Barber 1959) were accomplished. Unfortunately, when investigating the continuous
spectrum, little attention was paid to the ultraviolet spectrum, and as a
result all observations referred only to the photographic and visual regions.
The resulting curve for the ultraviolet spectrophotometric gradient
in the years 1948-1964 contains only two points (Chalonge, Divan 1952;
Dibai 1956). In about 1963 a new increase of the brightness of gamma Cas was
marked. About 1966 the continuous brightening stopped (Kalish 1966) but
the instability is still expressed in short-periodic oscillations in the star's
luminosity. It seemed interesting to compare the available observational data
and to investigate the question if the behaviour of the star in the
contemporary stage fits the conception of a variable shell, in agreement with
data on the previous flare of gamma Cas (Gorbatzkij, 1949).
We obtained photographic spectra of gamma Cas, beginning 1958 in Alma-Ata
(slit spectrograms with a dispersion of 140 A/mm at H gamma) and during the
years 1964-1967 in Byurakan (slitless spectrograms 175 A/mm at H gamma). In
1965 we began to observe gamma Cas in Alma-Ata with a diffraction spectrograph
(30 A/mm) attached to the 28" reflector AZT-8, and in autumn 1966 we started
to scan the star's spectrum on a 20 reflector. The observations were conducted
as far as possible parallel to perform simultaneous analysis of the continuous
and line spectrum. delta Cas was chosen as a comparison star for the investigation
of the continuous spectrum. We tried to conduct our observations on gamma Cas
for as long as possible, to find eventual short-periodic variations.
69 spectroelectric records, 22 slit spectrograms (1958-1964) and 23 slitless
spectra in the range 3200-6600 A were treated. Relative spectrophotometric
gradients were determined in the usual way for three spectral regions:
Phi_1 (lambda lambda 4000-4800 A), Phi_2 (lambda lambda < 3700 A) and
Phi_3 (lambda lambda > 4800 A). To get absolute gradients for gamma Cas,
average gradient values for the sepetral class of the comparison star were
used (Aller 1955).
Average absolute gradients of gamma Cas for each night are given in Table I.
We notice essential differences in the averaged gradient values from to date.
Even more astonishing are the considerable differences, occurring during
several hours, for example, on November 15 1966, which cannot be explained
by observational errors.
In Table I still another important fact should be noticed. For the whole
period covered by our observations the relation Phi_2 < Phi_1 < Phi_3 was
satisfied on the average. The UV-gradient Phi_2 was smaller than the blue one.
Formally it corresponds to an ultraviolet spectrophotometric temperature about
5000 deg higher than the temperature determined from the blue part of the
spectrum. The relation Phi_2 < Phi_1 contradicts to our recent knowledge on the
structure of stellar photospheres, requiring opposite correlation (Mustel 1941; 1944).
At the same time our results, based on a sufficient number of observations, and
obtained with different instruments and methods, seem to be reliable.
Table I
Date D Phi_1 Phi_2 Phi_3 n
1958
October 21/22 - 0.70 - 4
November 6/7 - 0.59 - - 4
24/25 0.70 - 7
1964
September 2/3 - 0.63 - 4
October 12/13 -0.06 0.75 0.74 - 2
December 15/16 - 0.74 - 3
1966
October 28/29 -0.06 0.87 0.81 1.36 5
29/30 -0.06 1.07 0.80 1.77 4
November 13/14 -0.06 0.92 0.92 1.38 3
15/16 -0.07 0.83 0.98 0.84 5
15/16 - 0.88 0.34 1.26 5
23/24 -0.07 0.38 0.28 0.84 6
December 3/4 -0.07 1.28 0.75 0.72 2
1967
March 13/14 -0.10 0.76 0.74 - 6
14/15 -0.11 0.75 0.63 - 5
July 20/21 -0.05 0.98 0.77 1.12 2
21/22 -0.05 0.88 0.92 1.31 3
31/1 -0.06 1.07 0.64 1.25 3
September 12/13 - - 0.56 - 3
December 10/11 -0.08 0.91 0.83 1.43 3
10/11 -0.05 0.86 0.61 1.22 3
11/12 -0.06 0.96 0.57 1.34 12
12/13 -0.06 0.90 0.50 1.21 6
14/15 -0.06 1.17 0.48 1.44 3
1968
January 5/6 -0.07 0.57 0.32 1.65 2
Fig. 1
From Gorbatzkij's (1949) research it follows, that at the late 30es
the gradient ratio was in agreement with theory. It was of interest to confront
all available observations executed during a longer period, for example during
the last 20 years, to reveal the character of changes, which the gradients have
undergone, to see at which time the UV and blue gradients obtained their recent
abnormal ratio and to test the reality of our results. In Fig. 1 averaged
half-year values of the spectrophotometric gradients, based on available
data are represented. The diameters of the marks accord with the number
of observations used, the scattering, as usual, is marked by a vertical line.
(Vandekerkhove 1957, Barber 1959, Zilevicute 1965, Taffara 1957, Chalonge
and Divan 1952, Dibai 1956, Boyarchuk 1958.)
From Fig. 1 the following conclusions can be drawn:
l. During the last 20 years the spectrophotometric gradients of gamma Cas
showed a smooth change at all investigated parts of the spectrum.
a) The ultraviolet gradient showed a distinct decrease, in accordance with
the increase of the spectrophotometric temperature of gamma Cas by 8000 deg.
b) On the other hand, the blue gradient increased from ~ 0.6 in 1948
up to an average 0.9 in 1967. That corresponds formally to a cooling by more
than 15000 deg.
c) In the first half of the 50es the red gradient showed a distinct wave
with an amplitude larger than 0.3. Since 1956 a stable increase-tendency was
marked, corresponding to a "cooling" by more than 20000 deg.
We want to draw special attention to the fact, that the gradient-changes
have a systematic character. It seems that a smooth process is proceeding in
the inner parts of the star and nobody can yet say to what results it will lead
in the near future.
2. Comparing the run of the gradients Phi_1 and Phi_2, we see, that till the
middle of the 50es a normal ratio Phi_2 > Phi_1 was observed. Apparently in the
time interval 1956 to 1958 the numerical values of the gradients became equal
and since then the anomaly Phi_2 < Phi_1, in a continuously increasing degree
(due to the steady decrease of Phi_2 and especially to the growth of Phi_1) has set
in. It was already marked that at the same time, about 1956, the wavy changes
in the visual spectrum described by the gradient Phi_3, passed on to a
continuous growth.
3. Our results for the blue and red spectra are in good agreement with
those obtained by Vilnius observers at nearly the same time.
To make it obvious what spectral regions are responsible for the abnormal
structure of the continuous spectrum of gamma Cas and what the appearance of
the anomaly, is, we have constructed curves of absolute energy distribution
for different years, using our observational data. Curves of the relative energy
distribution of gamma Cas were averaged for the seasons of 1958, 1966, and for
the summer and winter of 1967. Then, using the absolute energy distribution
tabulated by A. V. Kharitonov for the comparison star delta Cas, we calculated
in absolute units the energy distribution for the averaged spectra of gamma Cas.
These results were compared with an energy distribution curve for an average
B0 spectrum, constructed by using data for B0-B1 stars, included in Kharitonov's
list (1964). No correction for interstellar extinction was necessary, because
its differential effect on gamma Cas and the comparison B0 star was negligible.
In Fig. 2 absolute distribution curves for an average B0 star and gamma Cas
are plotted for different seasons.
By examining this figure we notice that in the blue part of the spectrum
a deep depression, which mainly causes the anomalous gradient-ratios, is
conspicuous. It seems that beginning in 1958, when the blue spectrum was
practically normal, spectral changes developed, deepening the depression,
more pronounced in the years 1966-1967. The energy distribution in the
UV spectrum seemed practically normal, in 1966 it completely corresponded
to the distribution of a B0-star, in 1967 it became somewhat steeper, "hotter".
The intensity in the red spectrum is larger than in the case of a normal star
and the discrepancy is growing with the wavelength. It was desirable to get
at least a rough idea of the infrared spectral part, inaccessible for our
observations. For this purpose the results of multicolour photometry were used.
In Fig. 3 energy distribution curves based on 8-colour photometric data (Johnson
1964, Johnson et al. 1966) for gamma Cas in 1963 and 1966 and for the average
B0-star, obtained by averaging the colour-indices of the listed B0 and B1
stars (Johnson 1966) of the luminosity classes IV-V, are plotted. Examining
Figure 3 we draw the following conclusions.
1) The intensity increase with increasing wavelength, noticed in the visual
range, is continuing in the infrared part of the spectrum as far as observations
exist. Reddening of gamma Cas in comparison with other early B-stars has already
been marked (Lunel, 1954).
2) The reddening of gamma Cas is intrinsic. This becomes obvious when analysing
the experimental law of interstellar extinction and especially from the variability
of the energy excess in the red spectral region, clearly seen in Fig. 3.
3) The reddening increased considerably in 1966 as compared with 1963.
This agrees with the detected tendency of intensity increase in the red spectrum.
H beta emission intensity variations as well as changes of gradients and
Balmer discontinuities, occurring during one night, were observed several times.
V. Liutij found brightness oscillations of gamma Cas, which reached 0.1 mag in
U during several hours. But the question of short-periodic changes in gamma Cas
requires a more detailed investigation.
Fig. 2
Fig. 3
An unsuccessful attempt was made to explain the unusual structure of
the continuous spectrum of the star by a superposition of a shell on a B0-star.
Thus the spectrum of gamma Cas cannot be interpreted in terms of the usual
Be-star model of a hot star and a variable shell, as it was possible in
the 30es (Gorbatzkij, 1949). The contemporary state of the star requires
an additional mechanism.
In conclusion we want to emphasize once more that the noticed accumulation of
some continuous process in the star's interior makes an appreciable change of
its physical state highly probable. Interesting spectral changes can be awaited.
This makes systematic detailed observations of gamma Cas in the nearest future
urgently necessary.
REFERENCES
Alley, L., 1955, Astrophysics 1, 183.
Barber, D., 1959, Mont. Not. astr. Soc. 119, 381, 534.
Boyarchuk, A. A., 1958, Izv. Krym. astrofiz. Obs. 18,. 3.
Chalonge, D., and Divan, L., 1952, Ann. Astrophys. 15, 201.
Dibai, O., 1956, Astr. Zu. 33, 506.
Gorbatzkij, V. G., 1949, Astr. Zu. 26, 307.
Johnson, H. L., 1964, Bol. Obs. Tonantzintla Tacubaya, 3, 305.
Johnson, H., 1966. Mitchell, L. L, Iriarte, B., Wisniewski, W. Z.,
Commun. Lunar planet. Lab. 99.
Kharitonov, A. V., 1964, Izv. AN Kaz. SSR, Astrofysika, 17, 28.
Kharitonov, A. V., 1964, Kariagina, Z. V., 1964, Astrofysika 17, 10.
Kopilov, I. M., 1954, Izv. Krym. astrofiz. Obs. 12, 162.
Kopilov, I. M., Boyarchuk, A. A., 1955, Izv. Krym. astrofiz. Obs. 15, 190.
Kalish, L., 1966, Sky Tel. 32, 265.
Lunel, M., 1954, Ann. Astrophys. 17, 153.
Mamatkazina, A. Ch., 1967, Astr. Cirk. Izdav. bjuro astr. Soobsc. Kazah, No. 399, 2.
Mustel, E. R., 1944, Astr. Zu. 18, 297, 1941. 21, 133.
Taffara, S., 1957, Contr. Oss. astrofis. Univ. Padova No. 86, 3.
Vandekerkhove, E., 1957, Belgique Com. No. 126.
Vandekerkhove, E., 1961, Observatory, 81, 144.
Zilevicute, Z., and Straizis, V., 1965, Bjul. Vilnius Obs. No. 15, 49.