Äîêóìåíò âçÿò èç êýøà ïîèñêîâîé ìàøèíû. Àäðåñ îðèãèíàëüíîãî äîêóìåíòà : http://star.arm.ac.uk/~ambn/euve_mm1.ps Äàòà èçìåíåíèÿ: Fri May 31 18:25:39 1996 Äàòà èíäåêñèðîâàíèÿ: Mon Oct 1 22:02:28 2012 Êîäèðîâêà:

A&A manuscript no.
(will be inserted by hand later)
Your thesaurus codes are:
07 (08.01.1; 08.01.2; 08.03.5; 08.06.1; 08.09.2 EUVEJ2056­17.1; 08.12.1)
ASTRONOMY
AND
ASTROPHYSICS
15.5.1995
Optical identification of EUV sources:
The secrets of EUVE J2056­17.1
M. Mathioudakis 1 , J.J. Drake 1 , N. Craig 1 , D. Kilkenny 2 , J.G. Doyle 3 , M.M. Sirk 1 , J. Dupuis 1 ,
A. Fruscione 1 , C.A. Christian 1 , and M.J. Abbott 1
1 Center for EUV Astrophysics, 2150 Kittredge Street, University of California, Berkeley, CA 94720--5030
2 South African Astronomical Observatory, PO Box 9, Observatory 7935, South Africa
3 Armagh Observatory, Armagh BT61 9DG, Northern Ireland, UK
Received date; accepted date
Abstract. We present optical, ultraviolet (UV), and ex­
treme ultraviolet (EUV) results for a new active late­type
dwarf star discovered by the Extreme Ultraviolet Explorer
(EUVE ). A large flare with an energy in excess of 10 35
erg and duration of ¸1.1 days was detected in the EUVE
Lex/B band (60--200 š A). The energetics of the event indi­
cate that radiative losses in the EUV are a significant con­
tributor to the energy budget in stellar flares. The ``qui­
escent'' EUV emission of the source is similar to the most
active flare stars with the ``quiescent'' optical spectrum
showing strong Hff, Ca II H & K, and Mg II h & k emis­
sion. A strong Li I 6707.8 š A absorption line is also present
in the spectrum. We have estimated a Li abundance of
log N(Li) = 2.5\Sigma0.4. Although the high Li abundance
suggests that EUVE J2056­17.1 is a young object hav­
ing recently arrived on the main sequence; the high flare
activity favours an interpretation where the enhanced Li
is sustained by spallation reactions.
Key words: stars: flare---stars: late­type---EUV emission
1. Introduction
With the great advances in the sizes and sensitivity of de­
tection systems over the past 20 years, the field of celestial
X­ray astronomy has seen several major breakthroughs.
The small satellite era of the 1970s was followed by large
dedicated missions such as HEAO 1, HEAO 2, EXOSAT,
GINGA, and ROSAT (Pallavicini 1989). Although there
have been several studies in X­rays over the past 20 years,
it was not until the 1990s that comprehensive studies of
the extreme ultraviolet (EUV) sky were to take place. The
higher opacity of the interstellar medium in EUV wave­
lengths restricts the number of objects detected indeed, to
Send offprint requests to: M. Mathioudakis
those located nearby. The Wide Field Camera, flown as
part of the ROSAT mission, surveyed the shortest EUV
wavelengths (60--200 š A) in 1990 (Pounds et al. 1993).
The Extreme Ultraviolet Explorer (EUVE ) is the
first mission fully dedicated to EUV (60--740 š A) astron­
omy. The first EUVE Catalog contains 410 EUV sources
(Bowyer et al. 1994). Optical counterparts has been found
for 90% of these sources. The unidentified sources may
constitute a potentially interesting class of objects since
they are bright in the EUV and faint in optical wave­
lengths. With the completion of the EUVE survey, an op­
tical spectroscopy campaign was initiated to identify EUV
sources without known optical counterparts (Christian et
al. 1993).
The numerous population of K and M dwarfs makes
them a very interesting subject, and the extensive studies
of stellar X­ray emission soon revealed strong ``quiescent''
X­ray emission from the dMe and dKe stars (where the e
signifies that the Hff line is in emission in their ``quiescent''
state). In certain cases, the X­ray emission can be as large
as 1% of the bolometric luminosity making these stars
the most efficient coronal emitters of all main­sequence
stars (Pallavicini 1989). This conclusion also extends to
the EUV (Mathioudakis et al. 1995). In 1975 the tele­
scope on the Apollo­Soyuz mission detected a significant
variation on the dMe star Prox Cen (Haisch et al. 1977).
This was the first flare detected in EUV wavelengths on a
star other than the Sun. New active late­type stars could
therefore be found by searching for emission from their
flaring or ``quiescent'' magnetically heated corona.
In the present paper we present EUV, UV, and optical
data on the source EUVE J2056­17.1, which is one of the
brightest unidentified objects in the first EUVE Catalog.
We identify the object as a new very active flare star. An
enormous flare was detected during the EUVE observa­
tions. Follow­up optical and ultraviolet spectroscopy re­
vealed strong Hff, Ca II H & K and Mg II h & k emission.

2 Optical identification of EUV sources
A strong Li I 6707.8 š A absorption line is also present in
the spectrum.
Fig. 1. An HST Guide Star Catalogue finding chart for the
source EUVE J2056­17.1 centered on the EUVE position in
J2000. The 60 00 positional error circle of EUVE is also shown.
2. Observations and data reduction
2.1. The EUVE survey
EUVE was launched in June 1992 and using three co­
aligned scanning telescopes pointing at right angles to
the satellite spin axis conducted an all­sky survey in
four bands covering the wavelength ranges 55--200 š A
(Lexan/B), 160--240 š A (Al/Ti/C), 370--600 š A (Ti/Sb/Al),
and 500--740 š A (Sn/SiO) (Bowyer et al. 1994). A fourth
telescope, the Deep Survey (DS) telescope, was aligned
along the spin axis and while pointing to the anti­Sun
direction advanced at ¸1 ffi per day along the ecliptic.
The spacecraft rotated around its spin axis three times
per 96 minute orbit. Sources in the scan path of EUVE
were scanned once per orbit. The coverage varied as a
function of ecliptic latitude with sources on the ecliptic
equator being scanned for a minimum of 5 days while
sources at the poles were scanned throughout the 6 month
survey. In this mode the scanning telescopes covered the
entire sky in approximately 6 months. The DS carried out
a more sensitive survey in a 2 ffi \Theta 180 ffi strip of the sky
along the ecliptic in the Lexan/B (60--200 š A) and Al/C
(170--360 š A) bands. The filter of the DS telescope is di­
vided into three sections. The center section is the Lex/B
with two panels of Al/C on either side. As the spacecraft
rotates around its spin axis a source drifts into the field of
view of the DS telescope and is observed in the Lex/B and
Al/C filters (Haisch et al. 1993). The source makes a full
revolution through the two filters during an orbital night.
Useful data are collected only during orbital night which
lasts for about 1900 s. The combination of higher effective
areas and longer exposure times gave the DS Lex/B higher
sensitivity by approximately a factor of 10 compared with
the all­sky survey.
The sources included in the first EUVE catalog
(Bowyer et al. 1994) include late­type stars, white dwarfs,
early­type stars, cataclysmic variables, planetary nebulae,
X­ray binaries, novae, extragalactic sources, and several
unidentified objects (NOID). Almost 50% of the objects
detected during the all­sky survey are late­type stars. This
fraction is quite different from the DS where 70% of the
sources are late­type stars. This comparison shows that
late­type stars are exposure time limited as compared to
the remaining objects, which are further away and there­
fore limited by absorption from the interstellar medium.
These facts would suggest that a large fraction of the
NOID objects are probably late­type stars.
2.2. Analysis of optical and ultraviolet observations
In Fig. 1 we present a finding chart centered at the posi­
tion of EUVE J2056­17.1. The finding chart was extracted
from the digitized Schmidt plates of the sky stored at the
Space Telescope Science Institute (STScI, courtesy of Drs.
M. Shara & L. Bergeron). The 60 00 positional error circle of
EUVE is also given in the figure. The bright star near the
center of the field was observed spectroscopically at the
Kitt Peak National Observatory (KPNO) on the night of
1993 July 13--14. The observations were part of our cam­
paign to identify EUVE sources in optical wavelengths.
The observations were performed on the 2.1 m telescope
with the Gold Camera CCD spectrometer equipped with
the F3KC CCD and the 500 lines mm \Gamma1 grating blazed
at 5500 š A that provided a wavelength coverage in the
region of 3600--7200 š A with a spectral resolution of ap­
proximately 5 š A. A He­Ne­Ar arc spectrum was used for
wavelength calibration where flux calibration was achieved
using the standard star HZ 44. The spectra were analyzed
using standard reduction techniques available in IRAF
(Tody 1986).
Additional spectroscopic data were obtained on 1994
June 9--10 and 11--12 using the Unit Spectrograph and
Reticon Photon Counting System (RPCS) on the 1.9 m
telescope of the South African Astronomical Observatory
(SAAO). These observations covered the wavelength re­
gions 3400--5300 š A and 5600--7200 š A with a spectral reso­
lution of approximately 4 š A in both spectral ranges. Before
and after each stellar observation a Cu­Ar arc was used for

Optical identification of EUV sources 3
Table 1. A log of the optical photometric observations of
EUVE J2056­17.1 obtained from SAAO in 1994 Septem­
ber--October.
JD V B--V U--B V--R V--I
2440000+
9621.346 10.40 1.260 1.092 0.792 1.575
9625.374 10.428 1.263 1.104 0.796 1.584
9626.384 10.501 1.275 1.113 0.808 1.613
9629.400 10.476 1.263 1.068 0.798 1.594
wavelength calibration; flux calibration was achieved us­
ing the white dwarf standards LTT 3864 and LTT 6248.
It should be noted that the RPCS is not a spectropho­
tometer, so the absolute flux calibration is unlikely to be
accurate in absolute terms. However, note that the con­
tinuum level at 6400 š A agrees within 15% with the R
magnitude of the object. The SAAO data were reduced
with the SKIP package currently in use at SAAO.
Photometric observations of EUVE J2056­17.1 in the
UBV (RI )c system were also made from SAAO in Septem­
ber 1994 using the modular photometer on the 0.5 m tele­
scope. The data were reduced to the Cousins' system using
observations of standards from Menzies et al. (1989). Typ­
ical errors in the photometry of a star of this brightness
measured with the 0.5 m telescope would be around 0.01--
0.015, although for a star of this color it is not surpris­
ing that errors in the (U--B) color are somewhat larger,
while errors in (V--R) and (V--I) are somewhat smaller,
simply as a result of the relative count rates. Evidence for
variability exists in the photometric data. The results are
listed in Table 1. In Fig. 2 we present the KPNO spectrum
of EUVE J2056­17.1 in the 3600--7200 š A range, whereas
Fig. 3 shows the SAAO spectrum centered in the Hff re­
gion. The spectrum is characterized by deep absorption
bands of TiO, MgH, and CaOH. A comparison with the
libraries of stellar spectra of Jacoby et al. (1984), Pettersen
& Hawley (1989), and Mathioudakis & Doyle (1991) clas­
sifies the object as a dwarf star of spectral type between
dK7e and dM0e. Because of their high dissociation energy,
the molecular bands in the spectra of late­type dwarfs are
temperature sensitive and therefore depend on the abso­
lute magnitude of the objects. We have used the depres­
sion of the TiO (4760 š A), MgH (4780 š A), and CaOH (6230
š A) bandheads relative to the local continuum and the cal­
ibration of Pettersen & Hawley (1989) to estimate the ab­
solute visual magnitude of the object. Using the absolute
visual magnitude we then find that EUVE J2056­17.1 is
located at a distance of 50\Sigma12 pc. The correlations be­
tween the absolute visual magnitude and the above bands
are well defined. The major source of uncertainty in the
distance determination arises from the uncertainty in the
measurement of these bands in the spectrum of EUVE
J2056­17.1.
The optical spectrum is also characterized by strong
Hff, Ca II H & K emission and strong Li I 6707.8 š A ab­
sorption. Emission line fluxes were determined by fitting
Gaussian profiles to the observed line profiles with the con­
tinuum contribution subtracted. The significance of the Li
line will be discussed in the following section. A compari­
son of the KPNO and SAAO spectra shows no significant
variations in the line equivalent widths between the two
epochs. The line fluxes and equivalent widths are listed in
Table 2.
Table 2. The observed line fluxes (in 10 \Gamma13 erg cm \Gamma2 s \Gamma1 ) and
equivalent widths (in š A) for EUVE J2056­17.1.
Hff Ca II Ca II Li I Mg II C IV
K H 6707.8 š A h &k
Flux 3.1 1.7 1.8 .. 3.3 Ÿ0.7
EW 1.3 7.6 5.4 ­0.4 .. ..
EUVE J2056­17.1 was observed with the International
Ultraviolet Explorer (IUE ) on 1994 September 18. The
IUE spectra were analyzed using the packages IUEDR
and DIPSO, which are available on the UK STARLINK
computer network. The most prominent feature in the
long­wavelength spectrum (1900--3200 š A) is the Mg II h &
k lines. Due to the limited exposure time no lines were de­
tected in the short­wavelength spectrum (1150--1950 š A).
The Mg II h & k line flux and the C IV upper limit are
also listed in Table 2. A comparison of the chromospheric
parameters of EUVE J2056­17.1 with the compilation of
spectroscopic and spectrophotometric data of Panagi &
Mathioudakis (1993) shows that this object is one of the
most active late­type dwarfs. Its activity levels are com­
parable with stars such as CC Eri, BY Dra, AU Mic, Gl
867AB, Gl 890, and the others that are located at the
saturation limit of atmospheric emission.
3. Results & Discussion
3.1. The EUV flare
EUVE J2056­17.1 was observed in survey mode with the
EUVE DS instrument from approximately 1992 11:00 UT
August 2 to 22:00 UT August 4. The orientation of the
EUVE telescopes and the geometry of the survey was
such that the same position in the sky was observed by
the three scanning telescopes approximately three months
later. One of the initial puzzling results was that although
EUVE J2056­17.1 is one of the brightest EUV DS sources,
it was not detected during the EUVE all­sky survey. As
we mentioned in Section 2.1 the DS is coaligned with the

4 Optical identification of EUV sources
Fig. 2. The optical spectrum of EUVE J2056­17.1 in the 3600--7200 š A range with a spectral resolution of ¸5 š A
satellite spin axis. Therefore when a source enters the DS
field of view it traces a circle passing from the Lex/B into
the Al/C filter. Analysis of the DS Lexan/B count rate as
a function of time shows that the source exhibited a very
large, flare­like outburst during the DS observations. The
source was not detected in Al/C. Note that the DS Al/C
has a broader wavelength coverage than the all­sky survey
Al/Ti/C filter and includes a high background contribu­
tion from the He II 304 š A geocoronal emission. Thus it
is not surprising that the source was not detected in the
Al/C. The Lex/B count rates were determined by using
standard aperture photometry. The source region is de­
fined by a circle of 3.0 0 in radius, which is large enough
to encompass the instrument point­spread function. The
background is estimated in an annulus with inner and
outer radii of 8.0 0 and 9.1 0 , respectively. The lightcurve
of the flare is presented in Fig. 4. The data have been
binned into one time bin per EUVE orbit. Because of the
geometry of the EUVE all­sky survey, each source spends
an average of 17 s per orbit in the scanners' 5 ffi field of
view. However, a source remains continuously in the DS
2.2 ffi field of view over a ¸2.2 day period. Given the typi­
cal durations of flares from late­type stars, it is therefore
much more likely for a flare to be detected during the
EUVE DS survey than in the all­sky survey. The present
flare is characterized by a sharp rise lasting less than 4,000
s reaching a peak count rate of 0.42\Sigma0.02 counts s \Gamma1 . This
is about 1 order of magnitude higher than the ``quiescent''
emission. Since the rise to the peak occurred during a
day time orbit and a SAA (South Atlantic Anomaly) pass
where no data were collected, we are unable to constrain
the rise time of the event to better than 4,000 s. The main
characteristics of the decay phase are an initial fast decay
followed by a much longer tail. The duration of the flare

Optical identification of EUV sources 5
Fig. 3. A spectrum of EUVE J2056­17.1 centered at the Hff with a resolution of ¸4 š A; note the presence of Li I 6707.8 š A.
was ¸1.1 days. The ``quiescent'' count rate of the object
in the DS Lexan/B band is approximately 0.035 counts
s \Gamma1 . This count rate is consistent with a 3 oe upper limit
of 0.040 counts s \Gamma1 in the all­sky survey Lex/B band.
EUVE spectroscopic observations of ``quiescent'' emis­
sion as well as flares from active stars such as HR 1099 and
AU Mic have shown that the wavelength range covered
by the Lex/B band is dominated by lines of Fe XVIII--
Fe XXIII formed in the temperature range 10 6:7 \Gamma 10 7:2
K (Brown 1994; Cully et al. 1994). We have determined
fluxes from the count rates using the Monsignori­Fossi
& Landini (1994) line emissivities for an average coro­
nal temperature of 10 7 K. We examined the temperature
dependence of the flux and found that in the above tem­
perature range the resultant flux is within a factor of 2.
The interstellar medium column density to the object was
computed using a model of the interstellar medium (Jelin­
sky & Fruscione 1994) that employs a three­dimensional
interpolation method on a large database of column densi­
ties (Fruscione et al. 1994) in order to estimate the amount
of hydrogen in any given direction and distance. Apply­
ing this method to EUVE J2056­17.1 we have estimated a
value of N (HI) = 3 \Theta 10 19 cm \Gamma2 . The interstellar medium
attenuation was calculated using the hydrogen and helium
photoionization cross sections compiled by Rumph et al.
(1994). A He i /H i of 0.1 was used. Using these parame­
ters we derive a luminosity of 1:3 \Theta 10 31 erg s \Gamma1 at the flare
peak where the energy integrated over the total duration
of the event is estimated to be in excess of 10 35 erg. The
``quiescent'' luminosity of the source in the Lex/B band is
L Lex=B = 10 30 erg s \Gamma1 . The observed parameters of the
EUV flare are listed in Table 3.
Several flares have been observed on dMe and dKe
stars in X­rays by HEAO 2 and EXOSAT. The EXOSAT
low energy (LE) experiment had a spectral coverage of
6--250 š A and therefore extended into EUV wavelengths.
However, soft X­rays had a substantial contribution to
the observed fluxes, therefore the EUV component could

6 Optical identification of EUV sources
Table 3. The observed parameters for the flare in the EUVE DS Lex/B band (60--200 š A)
Date UT (max) t rise t decay L peak E total L ``quiescent''
(hours) (hours) erg s \Gamma1 erg erg s \Gamma1
1992 August 3 07 h 16 m Ÿ 1.1 23.6 1:3 \Theta 10 31 2:0 \Theta 10 35 1:0 \Theta 10 30
not be estimated (Pallavicini et al. 1988, 1990). The num­
ber of purely EUV flares reported so far is very limited. A
flare on BY Dra was detected with the Wide Field Camera
during the ROSAT all­sky survey (Barstow et al. 1991).
The BY Dra event lasted for about 4.5 hours with a total
integrated energy of 7 \Theta 10 32 erg in the 70--150 š A range. A
spectacular flare detected in the EUVE Lex/B band on the
dMe star AU Mic has been reported by Cully et al. (1994).
The AU Mic flare was characterized by an impulsive rise to
the peak followed first by a fast decay and then by a much
longer tail. The flare lasted for 1.5 days, reaching a peak
luminosity of 1 \Theta 10 30 erg s \Gamma1 and a total energy of 3 \Theta 10 34
erg. The observed time scales were much longer than the
radiative and conductive cooling time scales; this led Cully
et al. (1994) to the conclusion that the long tail of the
event could by explained by a model of rapid expansion,
causing the plasma to become tenuous sufficiently quickly
that it avoids catastrophic radiative cooling. However, in
the Cully et al. (1994) model, no additional heating is as­
sumed during the decay phase of the flare. It remains to
be investigated whether the long lightcurve could be fit
by a model where the plasma cools by radiation and/or
conduction with continual heating taking place through­
out the duration of the event, similar, say, to the solar
two ribbon flares (Doyle & Widing 1990; Widing & Doyle
1990; Cargill & Priest 1983).
The flare on EUVE J2056­17.1 is one of the most en­
ergetic flares reported on a late­type dwarf. Its lightcurve
is very similar to the AU Mic flare although it should
be noted that the AU Mic flare was observed in pointed
mode, and therefore achieved a better time resolution. In
a recent coordinated EUVE and ground­based observa­
tional campaign of AD Leo, a large flare was detected and
the flare energy emitted in the EUV was 3 times larger
than the energy emitted in the Johnson U band (Haw­
ley 1995). The EUV flare events observed to date indicate
that the extreme ultraviolet radiation carries a substantial
fraction of flare energy losses. The high radiative energy
and long duration of this particular event being similar to
the largest flares seen on RS CVn binaries (Doyle et al.
1992a,b).
3.2. Detection of lithium
One of the most conspicuous lines in the spectrum is an
absorption feature centered at 6707.8 š A with a FWHM ap­
proximately equal to the instrumental width (Fig. 3). Be­
cause of the limited spectral resolution, this feature could
be due to a blend of Fe I 6705.4 š A, Fe I 6707.4 š A, the Li
I doublet at 6707.8 š A and Fe I 6710.6 š A. However, since
the observed line profile can be fitted quite well with a
single Gaussian centered at 6707.8 š A and a width equal
to the instrumental resolution, we therefore believe that
the absorption feature is mainly due to Fe I 6707.4 š A and
Li I 6707.8 š A. A comparison with the spectra of late­type
dwarfs presented by Pettersen (1989), Cayrel de Strobel &
Cayrel (1989) and Pallavicini et al. (1992) shows that the
equivalent width of Fe I 6707.4 š A is expected to be around
0.03 š A. Thus we conclude that the absorption feature is
dominated by the Li I line, with the contribution of Fe I
6707.4 š A to the observed equivalent width estimated to
be ¸10%.
The Li I 6707.8 š A resonance line has only been detected
in a small number of late­type dwarfs. These include the
dK stars HD 17925 (Cayrel de Strobel & Cayrel 1989),
HD 197890 (Anders et al. 1993), and HD 98800 (Fekel &
Bopp 1993). The coolest dwarf in which Li has been de­
tected is the dM0.5e star Gl 182 (Bopp 1974). The fact
that lithium is rarely seen in late­type dwarfs with spec­
tral types later than G5 is attributed to the fact that deep
convection transports Li I to temperatures higher than
2 \Theta 10 6 K where it is destroyed. We have estimated the
Li abundance of EUVE J2056­17.1 by spectrum synthe­
sis. All known lines in a 10 š A region centered on the Li
resonance line were included in the calculation. We used
the spectrum synthesis program MOOG, together with a
model atmosphere generated using the MARCS program
(Gustafsson et al. 1975) for solar metallicity. An effective
temperature of 4000 K and surface gravity of log g = 4.6
were used in the computations. These parameters were
based on estimates of the spectral type from the optical
spectra. The Li abundance derived in this way is log N
(Li) = 2.5\Sigma0.4.
Could EUVE J2056­17.1 be a weak T­Tauri star
(WTTS) ? Most of the WTTS have a Li abundance higher
than EUVE J2056­17.1, however a small number of WTTS
with a similar Li abundance do exist (Magazz`u et al.
1992). The chromospheric emission of pre--main sequence
stars (PMS) is an extrapolation to higher activity levels
from main sequence dwarfs (Calvet et al. 1985; Finken­
zeller & Basri 1987). The chromospheric activity levels
of EUVE J2056­17.1 are considerably lower than those
of PMS stars, they are in fact, as we already mentioned,
consistent with those of very active late­type dwarfs. Al­

Optical identification of EUV sources 7
Fig. 4. The lightcurve of the EUV flare in the DS Lex/B band (60--200 š A). The time is in Julian Dates since JD 2440000.5.
We have used one time bin per EUVE orbit for the flare. The ``quiescent'' data have been averaged over two orbits.
though EUVE J2056­17.1 has a high Li abundance, we
have no additional information to support the suggestion
that this source is a WTTS.
The Li abundance of EUVEJ2056­17.1 is below the
value derived for (PMS) which implies significant Li deple­
tion compared with PMS levels (Strom 1994). A compari­
son with the observations of stellar clusters would suggest
an age similar or smaller to that of the Pleiades (80--100
Myr). Note however, that the spread of the Li abundance
at a given spectral type is large and in the K dwarfs in par­
ticular can be 2 orders of magnitude or larger (Soderblom
et al. 1993). The abundance of Li can, however, be ef­
fected by other parameters. In a recent paper Houdebine
& Doyle (1995) solved the NLTE radiation transfer prob­
lem for a grid of model atmospheres that mimic the ef­
fect of magnetic activity for an M1 dwarf. The authors
showed that the Li I 6707.8 š A line is sensitive to activity,
with this dependence commencing at high levels of activ­
ity, i.e., where the Hff line is driven into emission. Thus
in M dwarfs the Li lines are weaker in plage­like chromo­
spheres than in ``quiescent'' regions. Changes in the UV
radiation field are responsible for a decrease in Li I equiv­
alent widths with increasing activity levels. UV photoion­
ization from the ground state has the largest effect on the
line formation. These results indicate that great care must
be exercised in the use of the Li I line in abundance and
age calculations for active stars.

8 Optical identification of EUV sources
A number of interpretations could be used to explain
the excess Li in EUVE J2056­17.1. For example, enhanced
Li is observed in solar spots as compared to plages (Gi­
ampapa 1984). This finding led to the suggestion that stars
showing large rotational modulation in white light, pre­
sumably from spots, should show significant variations in
Li I. However, an extensive survey of RS CVn binaries by
Pallavicini et al. (1992) did not detect any significant vari­
ations in the Li I equivalent width as a function of phase.
An alternative and possibly more likely explanation could
be that the high flare activity of the object is responsi­
ble for the excess Li. For example, Li can be produced
during energetic events such as flares by spallation reac­
tions (Canal et al. 1980). For a Li abundance similar to
that of EUVE J2056­17.1, Schramm et al. (1990) predict
that the energy required for producing Li by spallation is
¸ 10 46 \Gamma 10 47 erg. Since we do not have a statistically
significant sample of flares for this star, we do not have a
value for the time­average flare energy directly. However, a
well established relation exists between the time­averaged
flare energy in the Johnson U­band and the ``quiescent''
X­ray emission of late­type dwarfs (Doyle & Butler 1985).
We assume that the ``quiescent'' X­ray luminosity of the
source is equal to its Lex/B luminosity, L Lex=B = 10 30 erg
s \Gamma1 . The flux­flux relations of Doyle (1989) and Young et
al. (1989) show that this estimate is consistent with the
star's ``quiescent'' Hff emission. Mathioudakis et al. (1995)
have also shown that the EUV fluxes in late­type dwarfs
are comparable with the X­ray fluxes. The X­ray lumi­
nosity would imply a total time­averaged flare energy of
L \Lambda
tot ¸ 10 30 erg s \Gamma1 . Assuming that the age of the star is
similar to other nearby flare stars, ¸ 10 9 years, the energy
produced is ¸ 3 \Theta 10 46 erg. If the above assumption is cor­
rect, the energy released by flares is sufficient to sustain
the high Li abundance on this object. We would like to
emphasize that Schramm et al. (1990) concluded that the
flare process does not have enough energy to produce sig­
nificant Li in the surface of main--sequence stars. However,
the time­averaged flare energy used in their calculation is
3 orders of magnitude lower than that of EUVE J2056­
17.1. It is the difference in the time­averaged flare energy
that led us to the opposite conclusion. Both the high Li
abundance and strong coronal emission would support the
suggestion that EUVE J2056­17.1 is a young object that
recently arrived on the main sequence.
4. Conclusions
EUVE J2056­17.1 is the brightest unidentified EUV
Lex/B source in the first EUVE catalog. We presented op­
tical, ultraviolet, and extreme ultraviolet results showing
that the source is presently identified as a late­type star
with spectral­type in the range dK7e--dM0e. The ``qui­
escent'' chromospheric and coronal emission measured in
the Hff, Ca II H & K, Mg II h & k lines, and EUV, show
that this is one of the most active late­type dwarfs. Dur­
ing the EUVE observations the source gave a very large
flare­like outburst with an energy in excess of 10 35 ergs
measured in the EUVE DS Lex/B band (60--200 š A). The
duration of the event was ¸1.1 days. The optical spectra
show a strong Li I 6707.8 š A line. The estimated Li abun­
dance of log N(Li) = 2.5\Sigma0.4 is high compared with other
stars of similar spectral type. We discuss the possibility of
Li production in the atmosphere of EUVE J2056­17.1 by
spallation reactions during flares. Production of Li by such
a process could delay the fast destruction occurring in the
convection zone and therefore sustain a high abundance.
EUVE J2056­17.1 appears to be an extremely interesting
source, and future observations are therefore strongly en­
couraged.
Acknowledgements. We would like to thank the EUVE princi­
pal investigators, S. Bowyer and R.F. Malina, and the EUVE
science team for their advice and support. We would like to
thank Drs. S. Vennes, J. Ferneley, & K. Mitrou for obtaining
the IUE data and Drs. M. Shara & L. Bergeron for provid­
ing us with the finding chart of the source. We extend our
thanks to F. Marang & F. van Wyk for obtaining the UBVRI
observations. MM would like to thank S. Cully for useful dis­
cussions and an anonymous referee for useful comments and
suggestions. This work has been supported by NASA Contract
NAS5­30180 and has benefited from the use of the SIMBAD
database operated at CDS, Strasbourg, France. Research at
Armagh Observatory is grant­aided by the Dept. of Education
for N. Ireland with support provided in terms of both software
and hardware by the STARLINK Project funded by the UK
PARC.
References
Anders G.J., Jeffries R.D., Kellett B.J., Coates D.W., 1993,
MNRAS 265, 941
Barstow M.A., Bromage G.E., Pankiewicz G.S., Gonzalez­
Riestra R., Denby M., Pye J.P., 1991, Nature 353, 635
Bopp B.W., 1974, PASP 86, 281
Bowyer S., Lieu R., Lampton M., Lewis J., Wu X., Drake J.J.,
Malina R.F., 1994, ApJS 93, 569
Brown A., 1994. In: Caillault J.P. (ed.) Proc. Eighth Cam­
bridge Workshop on Cool Stars, Stellar Systems and the
Sun, Astronomical Society of the Pacific Conference Series,
Vol. 64, p. 23
Calvet N., Basri G., Imhoff C.L., Giampapa M.S., 1985, ApJ
293, 575
Canal R., Isern J., Sanahuja B., 1980, ApJ 235, 504
Cargill P.J., Priest E.R., 1983, ApJ 266, 383
Cayrel de Strobel G., Cayrel R., 1989, A&A 218, L9
Christian C.A., Abbott M., Carone T., Drake J.J., Dupuis J.,
Finley D., Fruscione A., Hawkins I., Malina R.F., Marshall
H.L., Vallerga J., Dopita M., Boroson T., Shara M., Berg­
eron L., 1993, BAAS 25, 863
Cully S.L., Fisher G.H., Abbott M.J., Siegmund O.H.W., 1994,
ApJ 435, 449
Doyle J.G., Butler C.J., 1985, Nature 313, 378
Doyle J.G., 1989, A&A 218, 195
Doyle J.G., Widing K.G., 1990, ApJ 352, 754
Doyle J.G., et al. 1992a, A&AS 96, 351

Optical identification of EUV sources 9
Doyle J.G., van den Oord G.H.J., Kellett B.J., 1992b, A&A
262, 533
Fekel F.C., Bopp B.W., 1993 ApJL 419, L89
Finkenzeller U., Basri G., 1987 ApJ 318, 823
Fruscione A., Hawkins I., Jelinsky P., Wiercigroch A., 1994,
ApJS 94, 127
Giampapa M.S., 1984, ApJ 277, 235
Gustafsson B., Bell R.A., Eriksson K., Nordlund A., 1975,
A&A 42, 407
Haisch B.M., Linsky J.L., Lampton M., Paresce F., Margon
B., Stern R., 1977, ApJ 213, L119
Haisch B.M., Bowyer S., Malina R.F., 1993, JBIS 46, 331
Hawley S.L., 1995 (in preparation)
Houdebine E.R., Doyle J.G., 1995, A&A (submitted)
Jacoby G.H., Hunter D.A., Christian C.A., 1984 ApJS 56, 257
Jelinsky P., Fruscione A., 1994 (in preparation)
Magazz`u A., Rebolo R., Pavleko Y.V., 1992, ApJ 392, 159
Mathioudakis M., Doyle J.G., 1991, A&A 244, 409
Mathioudakis M., Fruscione A., Drake J.J., McDonald K.,
Bowyer S., Malina R.F., 1995, A&A (in press)
Menzies J.W., Cousins A.W.J., Banfield R.M., Laing J.D.,
1989, SAAO Circ. 13, 1
Monsignori­Fossi B., Landini M., 1994, Solar Physics 152, 81
Pallavicini R., Monsignori­Fossi B.C., Landini M., Schmitt
J.H.M.M., 1988, A&A 191, 109
Pallavicini R., 1989, A&AR 1, 177
Pallavicini R., Tagliaferri G., Stella L., 1990, A&A 228, 403
Pallavicini R., Randich S., Giampapa M.S., 1992, A&A 253,
185
Panagi P.M., Mathioudakis M., 1993, ApJS 100, 343
Pettersen B.R., 1989, A&A 209, 279
Pettersen B.R., Hawley S.L., 1989, A&A 217, 187
Pounds K.A., et al., 1993, MNRAS 260, 77
Rumph T., Bowyer S., Vennes S., 1994, AJ 107, 2108
Schramm D.N., Steigman G., Dearborn D.S.P., 1990, ApJ 359,
L55
Soderblom D.R., Jones B.F., Balachandran S., Stauffer J.R.,
Duncan D.K., Fedele S.B., Hudon J.D., 1993, AJ 106, 1059
Strom S.E., In: Caillault J.P. (ed.) Proc. Eighth Cambridge
Workshop on Cool Stars, Stellar Systems and the Sun, As­
tronomical Society of the Pacific Conference Series, Vol. 64,
211
Tody D., 1986, Proc. SPIE 627, 733
Widing K.G., Doyle J.G., 1990, ApJ 352, 766
Young A., Skumanich A., Stauffer J.R., Bopp B.W., Harlan E.,
1989, ApJ 344, 427
This article was processed by the author using Springer­Verlag
L A T E X A&A style file L­AA version 3.
E­mail addresses
M. Mathioudakis Internet mihalis@cea.berkeley.edu
J.J. Drake Internet jdrake@cea.berkeley.edu
N. Craig Internet ncraig@cea.berkeley.edu
D. Kilkenny Internet dmk@mv.saao.ac.za
J.G. Doyle Internet jgd@star.arm.ac.uk
M. Sirk Internet sirk@cea.berkeley.edu
J. Dupuis Internet jdupuis@cea.berkeley.edu
A. Fruscione Internet antonell@cea.berkeley.edu
C.A. Christian Internet carolc@cea.berkeley.edu
M.J. Abbott Internet mabbott@cea.berkeley.edu