T. Tsuji , W. Aoki , & K. Ohnaka
Institute of Astronomy, The University of Tokyo, Mitaka, Tokyo, Japan
``Water everywhere in the Universe!'' is certainly
one of the great discoveries of ISO, and
star is not an exception. Although it has been
known that water exists in very cool stars
such as Miras and OH/IR stars (e.g., H2O maser has
been discovered some 30 years ago in such stars),
water has not been expected in plain red giants
and supergiants. However, ISO revealed the presence of water
in M giants and supergiants as early as M1-2, first by the low
resolution mode of SWS (Tsuji et al. A&A, 320, L1,1997;
Ap&SS, in press). We now confirmed
this finding by the highest resolution grating mode of SWS
on
Cet (M2III) and KK Per (M2Iab). We also observed M giants
stars through M4 to M8 by the high resolution of SWS06 and detected
H2O in all the cases including M4-6 giants (and also in
M4 supergiant S Per).
It is somewhat surprising that water is found not only in
late M giant stars but also in early and middle M (super)giants,
and the problem is what this fact means. Our attempt to understand
water in these early and middle M (super)giants by the present
photospheric models could never be successful. The excitation
temperatures of these H2O lines appear to be rather high,
above 1000K. Also, ISO spectra of late M giants (e.g. SW Vir)
show CO2
band at 4.3 m in absorption
while
band at 15 m in emission. Again, these CO2 features cannot be explained by the present photospheric models.
The shapes of the observed absorption band and the emission
features suggest excitation temperature of several hundred Kelvin,
and this fact suggests a presence of an extra CO2 envelope
outside the photosphere.
We have confirmed the ISO detection of H2O in early and middle
M giants by the high resolution infrared spectra observed by KPNO FTS.
Despite the disturbing atmospheric absorption fatal to the ground-based
observations, some H2O lines of the hot bands can be confirmed in
early through late M giants. With the spectral resolution as high as
105, the radial velocities of the H2O lines can now be measured
with high precision, and it turned out that they are not so different
from other photospheric lines. This fact suggests that the H2O is
not expanding and thus the H2O forming region will be quasi-static.
We conclude that there should be a molecule forming region effectively producing H2O and CO2 around M giants including early through late types. This molecule forming region may be not expanding yet and will serve as a ``reservoir'' for supplying matter to the expanding stellar wind observed through early to late M (super)giants. So far, presence of such ``reservoir'' has been suggested only for Mira variables in which matter can be levitated to the outer layer by the shock wave due to stellar pulsation. Now, ISO revealed that the presence of ``reservoir'' is a more general phenomenon in red (super)giants and this fact provides a new key in understanding mass-loss phenomena of red (super)giant stars in general.