Äîêóìåíò âçÿò èç êýøà ïîèñêîâîé ìàøèíû. Àäðåñ îðèãèíàëüíîãî äîêóìåíòà : http://star.arm.ac.uk/~ambn/334jgd.ps Äàòà èçìåíåíèÿ: Tue Apr 18 14:03:43 2000 Äàòà èíäåêñèðîâàíèÿ: Mon Oct 1 21:01:43 2012 Êîäèðîâêà: Ïîèñêîâûå ñëîâà: http www.astronomy.ru forum index.php topic 4644.0.html

A&A manuscript no.
(will be inserted by hand later)
Your thesaurus codes are:
08.01.2; 08.02.4; 08.03.3; 08.09.2 CC Eri; 08.12.1; 13.21.5
ASTRONOMY
AND
ASTROPHYSICS
14th April 2000
Rotational modulation and flares on RS CVn and BY Dra
stars.
XX. Photometry and spectroscopy of CC Eri in late 1989.
P.J. Amado 1? , J.G. Doyle 1 , P.B. Byrne ?? , G. Cutispoto 2 , D. Kilkenny 3 , M. Mathioudakis 4 , and
J.E. Neff 5
1 Armagh Observatory, College Hill, Armagh BT61 9DG, N. Ireland
2 Catania Astrophysical Observatory, V.le A. Doria, 6. I­95125 Catania, Italy
3 South African Astronomical Observatory, P.O. Box 7, Observatory 7935, South Africa
4 Department of Pure and Applied Physics. Queens University of Belfast, Belfast BT7 1NN, N. Ireland
5 College of Charleston, Dept. of Physics & Astronomy, Charleston, SC 29424, USA
Received date, accepted date
Abstract The active flaring binary CC Eri was studied
via multi­wavelength observations involving multi­based
photometry and ground­ and space­based spectroscopy.
Combining early spectroscopic data with the present im­
plies an orbital period of P = 1:5615 days. Further­
more, the spectroscopic data suggests spectral types of
K7 and M3 for the system. The optical photometry indi­
cated a small spot coverage in late 1989, consistent with
data taken a year later which showed CC Eri entering its
brightest­to­date phase.
Two flares were detected in the ultraviolet spectral
data. These flares radiated 2:7 10 31 erg and 1:6 10 31 erg in
the C iv line alone, each with a total estimated radiative
energy budget of ú 10 36 erg. For the higher­temperature
lines, such as C iv, there was no systematic variability with
phase. The lower­temperature lines show a slight indica­
tion of rotational modulation. However, there is a much
larger scatter in the individual measurements of the Mg ii
and C iv fluxes than would be expected from measure­
ment errors alone, consistent with an atmosphere showing
continual small­scale activity.
Key words: stars: activity -- binaries: spectroscopic --
stars: chromospheres -- stars: individual: CC Eri -- stars:
late­type -- ultraviolet: stars
1. Introduction
CC Eri (= HD16157 = CD\Gamma44 ffi 775 = GJ 103) is a
1.56 day spectroscopic binary with a mass ratio of ú2
Send offprint requests to: P.J. Amado
? Present address: Instituto de Astrof'isica de Andaluc'ia­
CSIC, Apartado 3004, 18080 Granada, Spain. E­mail:
pja@iaa.es
?? Deceased
(Evans 1959). The primary is a K7Ve and co­rotates with
the orbital motion, as shown by its light variations (Bopp
& Evans 1973). As such, it is one of the most rapidly
rotating late K solar neighbourhood dwarfs known. Oth­
ers include, YY Gem (= GJ 278c, 0.8 day) and HK Aqr
(= GJ 890, 0.4 day), although several ultra­fast rotators
are known to exist within the local association (e.g. see
Jefferies et al. 1994 and references therein).
CC Eri is a known flare star whose activity rate is at a
level of a \DeltaU =1 mag flare every ú12 hr (Busko & Torres
1976). CC Eri has been observed with the VLA, resulting
in 6cm flux measurements of 0:62 \Sigma 0:08 mJy and 1.34
mJy (Caillault et al. 1988) and Ÿ 0:23 mJy (White et al.
1994). It is also an infrared source with a luminosity of
¸2:2 10 30 erg s \Gamma1 at 12 ¯m (Tsikoudi 1988). In spite of
its many interesting properties, CC Eri is a relatively ne­
glected object and had only a single ultraviolet spectrum
taken prior to the present data. This spectrum, in the IUE
long wavelength spectral region (1900­3200 š A) yielded a
Mg ii resonance line luminosity that indicated a very high
level of chromospheric heating (Byrne et al. 1982). It is
also a powerful X­ray source, having been measured with
Einstein (Schmitt et al. 1987), exosat (Pallavicini et al.
1988, 1990) and rosat (Pan & Jordan 1995), yielding X­
ray luminosities between log LX = 29.2 and 29.8 erg s \Gamma1 .
In this paper we report on simultaneous optical and
ultraviolet spectroscopy plus photometry of CC Eri taken
in November 1989. Based on these data we examine the
orbital solution of the binary system and discuss the en­
ergy balance in the star's magnetically heated chromo­
sphere and transition region, in both quiescent and flaring
conditions. The ultraviolet data was supplemented with
more recent data from the extreme ultraviolet region. We
also examine the evidence for rotational modulation in the
optical and ultraviolet line emission. Preliminary results
were given by Byrne et al. (1992).

2 P.J. Amado et al.: Rotational modulation and flares on RS CVn and BY Dra stars.
2. Observational data
2.1. Photometry
The optical photometry was carried out on the 1.0m tele­
scope at the South African Astronomical Observatory at
Sutherland during October/November 1989 and at the
European Southern Observatory's 0.5m telescope at La
Silla during December 1989.
2.1.1. Photometry--European Southern Observatory
The observations of CC Eri (v = variable star) at the
eso were carried out over the period 9­28 December
1989 using a 0.5m telescope equipped with a single­
channel photon­counting photometer, a thermoelectrically
cooled rca 31034 GaAs photomultiplier and standard eso
filters matching the UBV(RI) c system. HD 16371 and
SAO 215945 were used as comparison (c) and check (ck)
star, respectively. Each measurement of a star consisted
of 10­15 1­sec integrations in each filter, according to the
U--B--V--R c --I c colour sequence. A complete observation
consisted of sequential c­v­v­v­ck­c measurements. From
these data, after accurate sky subtraction, three v­c and
one ck­c differential magnitudes were computed; the three
v­c values were finally averaged to obtain one data point.
The observations were corrected for atmospheric extinc­
tion and transformed into the standard UBV(RI) c system.
The typical data point error for the differential photom­
etry is of the order of 0.005 magnitudes, with somewhat
larger values in the U­band due to the low photon counting
level. The accuracy of the absolute photometry is of the
order of 0.01 magnitudes (see Cutispoto 1995 for further
details).
2.1.2. Photometry--South African Astronomical
Observatory
The saao data were obtained between 3­10 November
1989 with the 1.0m telescope and the St. Andrews photo­
electric photometer. CC Eri was observed together with
local comparison stars HD 15567 and 16651, and reduc­
tion of the UBV(RI) c observations was carried out us­
ing E­region standard stars from the compilation given
by Menzies et al. (1989) and conventional saao reduc­
tion procedures. Mean values for the local comparisons
are given in Table 1 and were used to differentially cor­
rect the CC Eri data.
2.2. Optical spectroscopy
CC Eri was observed with the University College London
Echelle Spectrograph (ucles) on the Anglo­Australian
Telescope (aat). The ucles was equipped with a
31.6 grooves mm \Gamma1 echelle grating and a gec ccd
(256\Theta584 pxl) which give an effective resolution of R¸
55 000. The spectra covered a number of non­contiguous
Figure 1. An example of spectra taken with the ucles on the
Anglo Australian Telescope in the region of Ca ii H and Hffl. A
4­gaussian fit plus continuum is also shown, where the centres
of the gaussians are indicated by dashes.
spectral regions from ¸3860 š A to ¸4165 š A. Data were ob­
tained on three consecutive nights 2--4 November 1989 but
clouds affected all three nights at some time during the ob­
servations. An exposure time of two minutes was set for
all the spectra, which were obtained whenever sky condi­
tions allowed in a continuous cycling mode. This observ­
ing strategy was adopted to ensure good discrimination
against short­lived flare events, although it gave a low
signal­to­noise ratio, which made it difficult to measure
the photospheric absorption lines.
The data were reduced using the astronomical com­
puting environment, iraf (Tody et al. 1986), available on
the uk starlink astronomical computing network (Bro­
mage 1984). A sample spectrum in the region of the Ca ii
H/Hffl chromospheric emission lines is presented in Fig. 1,
where gaussians were used to produce the fit shown in the
Figure as a solid line. The binary nature of the star can
be clearly seen from the doubling of the emission lines.
2.3. UV spectroscopy
CC Eri was observed on the 2, 3 and 4 November 1989
using the UV spectrograph on board the International
Ultraviolet Explorer (iue) satellite with both the long
(lwp; 1900­3200 š A) and short (swp; 1150­1950 š A) wave­
length cameras. The star was moved between two positions
at either end of the spectrograph slit thus yielding two
spectra per image. This was done to maximise the amount
of data collected since the spacecraft readout overhead was

P.J. Amado et al.: Rotational modulation and flares on RS CVn and BY Dra stars. 3
Table 1. UBV(RI)c photometry from the South African Astronomical Observatory (saao) and the European Southern Obser­
vatory (eso). Mean magnitudes and colours for the local comparisons used are also given. Phases are according to the ephemeris
2 447 129:5293 + 1:56145E.
saao
HJD(2 440 000+) Phase V U --B B--V V --R V --I Comment
7834.2990 0.3559 8.756 0.969 1.362 0.894 1.819 Flare?
7834.4640 0.4616 8.775 1.098 1.372 0.898 1.829
7834.5780 0.5346 8.776 1.097 1.370 0.897 1.826
7835.3000 0.9970 8.751 1.121 1.375 0.892 1.814
7836.3010 0.6380 8.768 1.110 1.377 0.895 1.816
7836.4360 0.7245 8.764 1.116 1.380 0.896 1.820
7837.2890 0.2708 8.780 1.100 1.380 0.893 1.820
7849.2970 0.9611 8.746 1.078 1.373 0.893 1.813
7849.5380 0.1154 8.764 1.130 1.371 0.894 1.822
7850.4880 0.7238 8.752 1.104 1.372 0.892 1.813
7851.3000 0.2439 8.769 1.070 1.376 0.893 1.820
HD15567: V =8.695, (B--V )=1.597, (U --B)=1.952, (V --R)c=0.934, (V --I)c =1.963
HD16651: V =8.132, (B--V )=1.167, (U --B)=1.064, (V --R)c=0.596, (V --I)c =1.144
eso
HJD(2 440 000+) Phase V U --B B--V V --R V --I Comment
7869.6730 0.0105 8.741 1.128 1.372 0.870 1.805
7870.6610 0.6432 8.748 1.116 1.373 0.868 1.814
7872.6280 0.9029 8.735 1.117 1.364 0.870 1.810
7873.6510 0.5581 8.762 1.144 1.370 0.867 1.816
7874.6501 0.1980 8.761 1.131 1.380 0.867 1.811
7876.6824 0.4995 8.774 1.132 1.390 0.872 1.823
7878.6744 0.7752 8.745 1.115 1.375 0.865 1.815
7879.5957 0.3653 8.767 ----- 1.384 0.873 1.819 No U­band mesured
7882.6313 0.3094 8.762 1.132 1.376 0.873 1.811
7885.6437 0.2386 8.759 1.127 1.378 0.866 1.811
7888.5520 0.1012 8.759 1.136 1.372 0.879 1.818
7888.6186 0.1438 8.764 1.136 1.373 0.876 1.814
7888.6603 0.1705 8.764 1.132 1.374 0.871 1.813
HD16371: V =8.09, (B--V )=0.90, (U --B)=0.57, (V --R)c =0.47, (V --I)c =0.92
SAO215945: V =9.73, (B--V )=0.39, (U --B)=\Gamma0:01, (V --R)c=0.22, (V --I)c =0.44
comparable to the exposure time. A further efficiency was
gained by alternating long and short wavelength spectra,
in such a way that one camera was exposing while the
other was being read. A full log of the iue spectra will
be found in Tables 2 and 3. The spectra were extracted
and flux calibrated from the iue images using the pro­
gram iuedr (Giddings 1983). They were then subjected
to a light gaussian smoothing procedure using the pro­
gram dipso (Howarth & Murray 1988). This smoothing
was such as to be consistent with the known resolution of
the spectrograph. An example of the smoothed spectra is
shown in Fig. 2, where emission lines are identified. Their
fluxes were determined by fitting gaussian profiles, again
using dipso, by a least­squares measure of the goodness
of the fit. Where lines were blended, multiple gaussians
were used. The resulting line fluxes are found in Tables 2
and 3.
The spectra in Fig. 2 show a number of emission
lines typical of active stars, including, in the lwp re­
gion, the Mg ii h and k (–– 2796/2803 š A) and the strong
Fe ii lines near –2600 š A and, in the swp region, the C iv
(––1548=1551 š A), C ii (––1335=1336 š A) and He ii (–1640 š A)
lines. As can be seen in the lower panel of Fig. 2, in ad­
dition to a greatly increased C iv line emission, there is a
sizeable continuum flux in the swp spectra SWP37513B,
indicating a large flaring event. Details on the line fluxes
of this flare and another two are given in Table 4, where
the fluxes have been computed by subtracting the mean
flux for each day (without the flares, in Table 2) from the
total flux measured in the lines.

4 P.J. Amado et al.: Rotational modulation and flares on RS CVn and BY Dra stars.
Table 2. iue swp emission line fluxes at Earth for CC Eri between 2--4 November 1989. A mean flux and associated variances
are given for each line on each date. A second mean is given in brackets, where appropriate, which excludes those points flagged
as flares in the Comments column. Phases are according to the ephemeris 2 447 129:5293 + 1:56145E. The notation jd2, jd3
and jd4 refer to data taken on 2 Nov., 3 Nov. and 4 Nov. respectively.
iue Image JD Phase Line Flux (\Theta10 \Gamma13 erg s \Gamma1 cm \Gamma2 ) Comment
Number 2 447 830.0+ N v C iv He ii C ii Si ii
1238/1242 š A 1548/1551 š A 1640 š A 1335/1336 š A 1808/1820 š A
SWP37500A 2.6913 0.326 2.97 5.15 1.71 4.31 1.70
SWP37500B 2.7188 0.344 2.42 6.20 1.30 5.30 1.93
SWP37501A 2.7810 0.384 3.19 9.45 1.82 5.83 2.21 Flare?
SWP37501B 2.8100 0.402 2.58 7.60 1.30 5.52 1.93
SWP37502A 2.8685 0.440 2.37 7.14 2.41 5.51 2.43
SWP37502B 2.8965 0.458 1.79 5.87 1.74 4.93 1.20
SWP37503A 2.9576 0.497 2.01 5.59 1.48 4.57 2.37
Mean jd2 2.48(2.47) 6.71(6.26) 1.68(1.56) 5.14(5.02) 1.97(1.93)
\Sigma0.49(0.49) 1.47(0.94) 0.38(0.42) 0.55(0.51) 0.43(0.45)
SWP37506A 3.4511 0.813 1.62 5.65 1.69 4.37 1.36
SWP37506B 3.4789 0.831 2.15 11.92 1.58 6.50 2.30 Flare
SWP37509A 3.7313 0.992 3.07 7.67 1.90 4.23 2.28
SWP37509B 3.7654 0.014 1.54 6.46 1.46 3.91 2.17
SWP37511A 3.9389 0.125 1.54 3.49 0.98 1.78 0.97
SWP37511B 3.9843 0.154 1.18 3.99 0.86 2.75 1.17
Mean jd3 1.85(1.79) 6.53(5.45) 1.41(1.38) 3.92(3.41) 1.71(1.59)
\Sigma0.67(74) 3.06(1.73) 0.41(0.45) 1.61(1.11) 0.61(0.60)
SWP37513A 4.2549 0.328 1.87 5.47 1.39 4.07 1.91
SWP37513B 4.2858 0.347 2.93 13.47 3.63 7.36 4.24 Flare
SWP37514A 4.3355 0.379 2.09 5.35 2.00 4.86 1.96
SWP37514B 4.3780 0.407 2.33 8.77 1.91 6.68 3.06
SWP37515A 4.4433 0.448 2.54 6.41 1.47 4.83 1.78
SWP37515B 4.4767 0.470 2.10 7.85 2.09 6.35 2.25
SWP37516A 4.5457 0.514 1.96 5.96 1.23 4.64 1.91
SWP37516B 4.5770 0.534 1.52 5.24 1.31 4.35 2.14
Mean jd4 2.17(2.06) 7.32(6.44) 1.88(1.63) 5.39(5.11) 2.41(2.14)
\Sigma0.43(0.33) 2.79(1.37) 0.78(0.36) 1.22(1.00) 0.84(0.43)
Overall Mean 2.18(2.08) 6.89(6.10) 1.68(1.57) 4.89(4.61) 2.06(1.92)
\Sigma0.56(0.52) 2.42(1.34) 0.58(0.40) 1.29(1.14) 0.70(0.51)
2.4. euve Fluxes
CC Eri was observed by the Extreme Ultraviolet Ex­
plorer (euve) in September 1995 (Bowyer & Malina 1991).
The observations were carried out with the Deep Sur­
vey/Spectrometer (DS/S) assembly for ú 160 ksec. The
euve data were analyzed and corrected for instrumental
effects using the iraf­based euv software egocs 1.6 and
egodata 1.11 (Miller & Abbott 1995). Further analysis
of the extracted spectra was carried out using the dipso
package. Emission line fluxes were determined by fitting
gaussians to the observed line profiles. The fluxes were cor­
rected for interstellar attenuation assuming a mean hydro­
gen density of 0.05 cm \Gamma3 resulting to a column density of
¸ 2\Theta10 18 atoms cm \Gamma2 at the distance of CC Eri (Zombeck
1990). The choice of this column density for CC Eri is in
agreement with the database of column densities of Frus­
cione et al. (1994). Given the proximity of the source, a
factor of two uncertainty in the hydrogen column density
will only produce a 16% uncertainty in the line fluxes.

P.J. Amado et al.: Rotational modulation and flares on RS CVn and BY Dra stars. 5
Table 3. iue lwp emission line fluxes at Earth for CC Eri between 2--4 November 1989. A mean flux and associated variances
are given for each line on each date. Phases are according to the ephemeris 2 447 129:5293 + 1:56145E. The notation jd2, jd3
and jd4 refer to data taken on 2 Nov., 3 Nov. and 4 Nov. respectively.
iue Image JD Phase Line Flux (\Theta10 \Gamma13 erg s \Gamma1 cm \Gamma2 ) Comment
Number 2 447 830.0+ Mg ii Fe ii
2796/2803 š A 2600 š A
LWP16719A 2.6628 0.3080 42.00 1.46
LWP16719B 2.6779 0.3177 41.38 1.40
LWP16720A 2.7612 0.3710 40.69 1.46
LWP16720B 2.7690 0.3760 43.44 1.27
LWP16721A 2.8529 0.4298 45.49 1.67
LWP16721B 2.8590 0.4337 46.61 0.88
LWP16722A 2.9417 0.4866 42.51 1.01
LWP16722B 2.9479 0.4906 44.33 1.56
Mean jd2 43.31 1.34
\Sigma 1.92 0.25
LWP16727A 3.5216 0.8580 31.16 1.27
LWP16727B 3.5282 0.8622 30.40 1.04
LWP16730A 3.8163 0.0468 34.29 1.15
LWP16730B 3.8238 0.0516 32.34 1.01
Mean jd3 32.05 1.18
\Sigma 1.47 0.10
LWP16734A 4.3187 0.3685 49.74 1.38 Post­flare?
LWP16734B 4.3271 0.3739 46.62 1.61
LWP16735A 4.4054 0.4240 47.37 1.49
LWP16735B 4.4135 0.4292 48.38 1.13
LWP16735A 4.5103 0.4912 53.27 1.20
LWP16736B 4.5179 0.4961 49.37 1.88
LWP16737 4.6055 0.5522 44.99 1.47
Mean jd4 48.53 1.45
\Sigma 2.54 0.23
Overall mean 42.86 1.33
\Sigma 6.39 0.25
3. Results
3.1. Contemporaneous optical photometry
The saao and eso data are given in Table 1 while the
resulting light and colour curves are shown in Fig. 3.
From the data it was obvious that there were some
small differences in the transformations and zero­point
corrections between the two datasets, which could have
been made worse by the use of two different comparison
stars. Therefore, and since it was not possible to directly
compare the mean colours of the star at these two observ­
ing runs, we shifted the saao data by 0.01 magnitudes
in (U --B), 0.025 in (V --R) c and 0.005 in (V --I) c , with no
correction in (B--V ). These shifts result when we compare
the V light curves, which in both datasets have the same
values either at phase 0.15 or phase 0.50, and we require
the star to have at these phases for both datasets the same
(U --B), (V --R) c and (V --I) c colours respectively.
There is a slight evolution between the saao data (and
probably also within it) and the eso data taken almost a
month later. A low amplitude modulation is visible in V
with a peak­to­peak variation of ¸ 0:035 and a mean value
of 8.764 for the saao curve. These data show, within the
errors of the photometry, no modulation in (B--V ) and
(V --R) c and only small variations in (V --I) c weakly cor­
related with the light curve. The largest variation, larger
than the intrinsic errors of the photometry, is present in
the (U --B) colour with an amplitude of ¸ 0:05 mag. The

6 P.J. Amado et al.: Rotational modulation and flares on RS CVn and BY Dra stars.
Table 4. The CC Eri flare line fluxes at Earth
iue Image JD Phase Line Flux (\Theta10 \Gamma13 erg s \Gamma1 cm \Gamma2 )
Number 2 447 830.0+ Nv C iv He ii C ii Si ii
1238/1242 š A 1548/1551 š A 1640 š A 1335/1336 š A 1808/1820 š A
SWP37501A 2.7810 0.384 0.72 3.19 0.26 0.61 0.28
SWP37506B 3.4789 0.831 0.36 6.47 0.20 3.09 0.71
SWP37513B 4.2858 0.347 0.87 7.03 2.00 2.25 2.10
Figure 2. (a) Mean lwp and (b) swp iue spectrum of CC Eri
outside of flaring and (c) swp spectrum of the flare event on
4 November.
eso data, taken one month later, appear brighter in V
than the saao data. The double­peaked V light curve
shows an amplitude of ¸ 0:040 with a mean value of 8.757
and the colour curves are clearly correlated with it, al­
though their amplitudes are very small.
3.2. Orbit and system parameters
Radial velocities have been determined from Ca ii H and K
for the primary and secondary components of the binary
using the task fxcor within iraf. The spectra were cross­
Figure 3. Light and colour curves for CC Eri based on saao
data taken between 3--10 November 1989 (triangles) and eso
data taken 9--28 December 1989 (squares). The filled circles in­
dicate the point which records the flare of 4 November. Phases
are those from Table 1.
correlated with one of the spectra taken on November 4,
which had a high signal­to­noise ratio, as no radial velocity
standards were observed. The velocity residuals for the
primary measurements are of the order of ¸ 2 km s \Gamma1
and up to 8 km s \Gamma1 for the secondary. These residuals
were used as standard errors in the weighting of the least­
square orbital fit.
We obtained a preliminary orbital solution, using
Evans (1959) orbit as a first guess. Fixing his value for the

P.J. Amado et al.: Rotational modulation and flares on RS CVn and BY Dra stars. 7
Figure 4. Measured radial velocities for CC Eri and its com­
panion along with the orbital solution mentioned in the text
(triangles indicate the secondary while square indicate the pri­
mary).
Table 5. The orbital solution for CC Eri of Evans (1959) com­
pared with the present results.
Evans Present work
Period (d) 1.56145 1.5615
fl (km s \Gamma1 ) +41.94 --
Kp (km s \Gamma1 ) 37.77 37.185
K s (km s \Gamma1 ) -- 69.289
! 63: ffi 8 --
e 0.0461 0.0
ap sin i (km) 8:100 10 5 7:984 10 5
a s sin i (km) 15:71 10 5 14:878 10 5
mp/m s 1.94 1.863
mp sin 3 i 0.1462 0.1271
m s sin 3 i 0.0753 0.0682
velocity of the centre of mass of the system fl, resulted in a
value for the orbital period of P = 1:54539 \Sigma 0:00171 days
(the eccentricity was fixed equal to 0). If we add Evan's
data to enlarge the time baseline and, thus, improve the
fit, we obtain a period of P = 1:5615 days. This is the
value for the period we have used in the calculations of
the minimum masses (M sin 3 i) and separations given in
Table 5, where we compare Evans' solution with our own.
Figure 5. iue swp emission line fluxes versus phase. From top
to bottom, Nv, C iv, He ii, C ii and Si ii line fluxes plotted
against phase. Triangles represent the data taken on 2 Nov,
squares on 3 Nov and pentagons on 4 Nov.
The radial velocity data are plotted in Fig. 4 with our
orbital solution. The mass function of the system is
m 3
s sin 3 i
(m p + m s ) = 8:319 10 \Gamma3 ;
where m p and m s are the masses of the primary and sec­
ondary respectively.
3.3. Emission line fluxes
Line fluxes from Tables 3 and 4 are plotted in Figs. 5
and 6 against phase, where this has been computed from
the ephemeris JD= 2 447 129:5293 + 1:5615E. It is readily
apparent that these line fluxes are variable. The data cover
phase intervals of \DeltaOE ú 0:179 on jd2 (2 Nov), 0.193 on
jd3 (3 Nov) and 0.184 on jd4 (4 Nov). Due to the almost
day and a half orbital period of the system, the third data
set coincided in phase with the first one.
It will also be noted from Tables 2 and 3 that the flux
of all the emission lines is systematically lower on jd3 than
on either of the other dates. This is most obvious in the
case of the Mg ii lines but it can also be seen throughout.

8 P.J. Amado et al.: Rotational modulation and flares on RS CVn and BY Dra stars.
Figure 6. iue lwp emission line fluxes versus phase. The Mg ii
(upper panel) and Fe ii line fluxes (lower panel) are plotted
against phase. Symbols are as in Fig. 5
Even at the same phases, there is evidence, at least in the
Mg ii fluxes, of a change in the mean level after only two
rotations (from jd2 to jd4). The mean Mg ii line flux at
Earth is approximately 4:3 10 \Gamma12 erg s \Gamma1 cm \Gamma2 , compara­
ble to that found in 1979 by Byrne et al. (1982) and that
of the C iv line is ú 6:1 10 \Gamma13 erg s \Gamma1 cm \Gamma2 , excluding the
two flare points.
There is no clear evidence of a systematic behaviour of
the line fluxes with phase, partly due to the gaps in cov­
erage, but there is a much larger scatter in the individual
measurements of the Mg ii flux than would be expected
from measurement errors alone (ú 10%). The only fea­
ture that seems to repeat in the flux curve is a dip in the
flux of the Fe ii lines, which can be seen at OE ú 0:45 for
both the jd2 and jd4 data.
3.4. Flares
The C iv (–1548/1551 š A) line, the strongest in the swp
spectra (after Ly ff), has two widely discrepant points. The
first of these is at phase 0.347 on jd4, while the second is
at phase 0.831 on jd3. This C iv flux increase on jd4 is
also seen in virtually all of the other prominent lines. That
on jd3 is only obvious in C ii. These were two discrete
flares, marked as such in the C iv line flux plot (Fig. 5).
The corresponding points in the C ii and Si ii line fluxes
are also marked.
4. Discussion & Conclusions
4.1. Physical parameters of the binary system from the
orbital solution
Before we can determine surface line fluxes, we must
determine a value for the radii of the two components
of the system. If we take the spectral type of the pri­
mary as K7.5Ve and its mass (Schmidt­Kaler 1982) as
0.57 M fi , then the companion is of mass 0.306 M fi or
spectral type M3.5Ve. The radii are (Schmidt­Kaler 1982)
RP = 0:645 R fi and R S = 0:41 R fi . Using the effec­
tive temperature, T eff and the visual absolute magnitude,
M V , G/l¸ebocki & Stawikowski (1995) inferred a radius of
R = 0:61 R fi . The spectral types match those derived from
the colours for another epoch (K7V + M3V) by Cutispoto
(1998).
Pettersen (1983) gave a v rot of 19.8 km s \Gamma1 , which
yields a radius of RP = 0:61 R fi . Bopp & Evans (1973)
assumed a v sin i of ¸ 15, which yielded a minimum radius
of RP sin i = 0:46 R fi .
The mean flux ratio at Earth FP(Ca H)/F S (Ca H) on
2 Nov was measured as 2.6. Assuming the radii as derived
above, the ratio of the areas, (RP=R S ) 2 = 2:47. Thus,
within the errors, the CaH brightness of the components
is similar.
4.2. Spot distribution
Comparing the photometric data of CC Eri taken be­
fore and after this observing run (Cutispoto 1991, 1992;
Cutispoto & Leto 1997), it can be readily seen that the
star was in the process of becoming brighter; in fact,
almost a year later (in September 1990) it reached the
brightest maximumever observed, namely, V = 8:70 mag­
nitudes (Cutispoto & Leto 1997), which is even brighter
than the V = 8:71 observed in late 1958 by Evans (1959).
Phillips & Hartmann (1978) reported photographic mag­
nitudes from the Harvard plate collection of CC Eri and
found a value for the B filter at maximum brightness of
Bmax = 10:03. If we calculate the corresponding Bmax
for the 1990 epoch (Cutispoto & Leto 1997), we obtain
Bmax = 8:70 + 1:35 = 10:05, which is, within the er­
rors of the photographic measurements, the same as the
maximum given by Phillips & Hartmann (1978). That the
star was near its unspotted magnitude together with the
very small amplitudes we measured in the light and colour
curves, point towards the idea of a reduced coverage of
spots at the 1989 epoch.
The fact that the mean V magnitude was fainter in the
saao data than for the eso data (taken one month later)
could be explained by the star becoming brighter towards
its historical maximum.The large intrinsic variation of the
(U --B) saao data could be explained by the appearance
of hotter, brighter material on the atmosphere of one of
the stars (e.g. a flare or plages), which could produce tem­
porary bluer ultraviolet colours (Amado & Byrne 1997).

P.J. Amado et al.: Rotational modulation and flares on RS CVn and BY Dra stars. 9
Table 6. euve line fluxes in erg s \Gamma1 cm \Gamma2
Spectral line Flux
Fe xviii 93.92 š A 1:1 10 \Gamma13
Fe xix 108.37 š A 1:0 10 \Gamma13
Fe xx/xxiii 132.85 š A 1:9 10 \Gamma13
He ii 304 š A 8:5 10 \Gamma13
This scenario coincides with the flaring state at which the
star was during the saao observations.
4.3. Ultraviolet line fluxes and evidence for rotational
modulation
The individual flux measurements of the Mg ii resonance
lines give some small evidence of a modulation in anti­
phase with the optical variation. This would perhaps con­
firm earlier results that chromospheric emission is corre­
lated in a general sense with optical spots. However, the
lesser modulation of the broadband optical illustrates that
the area coverage in the two are not related on a one­to­
one basis.
Comparing the line flux variations with the broadband
V light curves, we note that the contrast in the Mg ii (max­
to­min) is much larger, ú 40%, than in V, ú 5%. The evi­
dence for modulation of the higher temperature lines, i.e.,
Si ii, C ii and C iv resonance lines, is much weaker. Any
variation here is dominated by intrinsic variations, which
are ú 50%, much greater than typical errors of measure­
ment (ú 10 \Gamma 20%).
The lack of detectable modulation in C iv due to the
presence of a large scatter, possibly due to low level flaring,
is in keeping with previous results for the stars BY Dra
and AU Mic (Butler et al. 1987).
4.4. Emission measure and quiescent radiative losses
Based on the iue and euve lines fluxes of Table 2 and Ta­
ble 6 respectively, an emission measure (EM) curve (see
Fig. 7) can be constructed using the atomic data as given
in Doyle & Keenan (1992a) and Brickhouse, Raymond &
Smith (1995). Then from the adopted EM curve (this is
actually a set of loci which give upper limits to the ac­
tual EM distribution), we can estimate the total radiative
losses from the upper chromosphere to the corona (i.e. over
the temperature range 4:1 ! log T e ! 7:1) by multiplying
the EM by the radiative loss function. Here, we used the
radiative cooling curves of Cook et al. (1989). Brickhouse
et al. give the split of the line at 133 š A as 33% Fe xx,
67% Fe xxiii assuming a constant EM. However, such a
divide implies the Fe xx point significantly below that of
Fe xix. For the EM points as given in Fig. 7 we adopted
Figure 7. The emission measure curve for CC Eri based on
the iue and euve data in Tables 2 and 6 respectively. Note the
data gap between log Te = 5:5 and 6:5 .. see text.
Figure 8. The radiative losses (in erg s \Gamma1 ) from the chromo­
sphere, transition region and corona of CC Eri as a function
of log Te . The region between log Te = 5:5 and 6:5 is uncertain
due to the lack of spectral lines and its shape here is dominated
by the shape of the radiative loss function.
a 50:50 split. Due to the lack of lines, particularly in the
mid­to­upper transition region, we simply fitted a series
of linear fits for the outline of the assumed EM curve. The
correct shape of the EM curve in this temperature region
is therefore unknown. From these fits we estimate surface
radiative losses of ¸ 5:6 10 7 erg s \Gamma1 cm \Gamma2 in quiescent, giv­

10 P.J. Amado et al.: Rotational modulation and flares on RS CVn and BY Dra stars.
ing total losses over the whole surface of 2:9 10 30 erg s \Gamma1 .
The total radiative losses as a function of temperature is
shown in Fig. 8, the shape of this figure is largely domi­
nated by the radiative loss function, although the second
high temperature peak reflects the 10 7 K coronal temper­
ature of CC Eri. In the construction of the EM curve we
used the hipparcos distance of 11.51 pc and a stellar ra­
dius of 0:65R fi , i.e. we assumed the emission was from the
K star.
An alternative method of obtaining an estimate of
the total radiative losses (chromospheric, transition region
and coronal) is to use a relationship derived by Bruner
& McWhirter (1988) between the total radiated power
in a hot plasma and the power radiated in specific spec­
tral lines, e.g. C iv 1550 š A or Nv 1240 š A. For the quies­
cent state we get 2:1 10 30 erg s \Gamma1 (from C iv), in excellent
agreement with that estimated via the EM technique.
4.5. Flares
The two flares radiated 2:7 10 31 erg and 1:6 10 31 erg in the
C iv 1550 lines alone. These are large flares by the stan­
dards of iue observations of dMe stars and are several
orders of magnitude larger than solar two­ribbon flares.
Using the scaling between the C iv radiative output and
the total flare radiative output derived for solar flares by
Bruner & McWhirter (1988) we estimate the energy radi­
ated over the entire outer atmosphere (4:1 Ÿ log T Ÿ 8:0)
of these flares to be each ú 10 36 erg.
Van den Oord (1988) showed that for a two­ribbon
flare involving a filament of length l, the maximumamount
of energy stored is given by
W = 1:6 10 37
`
l
R fi
'`
R \Lambda
R fi
' 2 `
B surf
1000 G
' 2
erg (1)
where B surf is the surface magnetic field strength at the
star. Taking R \Lambda = 0:65 R fi and a filament length of
¸ 0:2 R fi (i.e a quarter of the stellar radii) implies kilo­
gauss fields on CC Eri. Since this however assumes a very
high magnetic energy conversion rate, these larger super­
flares could therefore be the result of the magnetic fields
in the two stars interacting. For example, van den Oord
(1988) showed that for a binary the maximal storage of
energy is obtained when the filament is located between
the two components. In this case the binary nature allows
storage of a factor of (1:6 a=R \Lambda ) 2 more energy (where a is
the binary separation), implying at least an order of mag­
nitude more energy available than for a single star. Such
an interpretation has being applied to flares on RS CVn
binaries (Doyle et al. 1992b).
Acknowledgements. Research at Armagh Observatory is grant­
aided by the Dept. of Education for N. Ireland. This work used
computer hardware and software provided by the UK Starlink
Project which is funded by the UK PPARC. PJA acknowledges
financial support from Armagh Observatory, and Catania As­
trophysical Observatory, and computing and technical support
from the Instituto de Astrof'isica de Andaluc'ia. Our work was
based on observations made with the International Ultravio­
let Explorer satellite at the ESA Satellite Tracking Station,
Vilspa, Spain, and at the NASA Goddard Space Flight Cen­
ter, Maryland, USA and on observations from the Anglo Aus­
tralian, South African Astronomical Observatory and Euro­
pean Southern Observatory. We acknowledge the contribution
of Dr. P.M. Panagi to this paper. A lot of this work could not
have been achieved without the effort and determination of our
late colleague Brendan Byrne who's untimely death occurred
before this paper was completed.
References
Amado P.J. & Byrne P.B., 1997, A&A 319, 967
Bopp B.W. & Evans D.S., 1973, MNRAS 164, 343
Bowyer S. & Malina R.F., 1991, In Extreme Ultraviolet As­
tronomy, eds. R.F. Malina & S. Bowyer, Pergamon Press,
P.397.
Brickhouse N.S., Raymond J.C. & Smith B.W., 1995, ApJS
97, 551
Bromage G.E., 1984, in European IUE Conf., p. 473
Bruner M.E. & McWhirter R.W.P., 1988, ApJ 326, 1002
Busko I.C. & Torres C.A.O., 1976, IAU Comm. 27 Inf. Bull.
Var. Stars No. 1186
Butler C.J. et al., 1987, A&A 174, 139
Byrne P.B., Butler C.J. & Andrews A.D., 1982, Irish Astron.
J. 14, 219
Byrne P.B., Cutispoto C., Kilkenny D.W., Neff J.E. & Panagi
P.M., 1992, Surface Inhomogeneities on Late­Type Stars,
ed. P.B. Byrne & D.J. Mullan, (Springer), p. 255.
Caillault J.­P., Drake S. & Florkowski D., 1988, AJ 95, 887
Cook J.W., Cheng C.­C., Jacobs V.L. & Antiochos S.K., 1989,
ApJ 338, 1176
Cutispoto G., 1991, A&AS 89, 435
Cutispoto G., 1992, A&AS 95, 397
Cutispoto G., 1995, A&AS 111, 507
Cutispoto G. & Leto G., 1997, A&AS 121, 369
Cutispoto G., 1998, A&AS 127, 207
Doyle J.G. & Keenan F.P., 1992a, A&A 264, 173
Doyle J.G., van den Oord G.H.J. & Kellett B.J., 1992b, A&A
262, 533
Doyle J.G., 1996, A&A 307, L45
Evans D.S., 1959, MNRAS 119, 526
Fruscione A., Hawkins I., Jelinsky P. & Wiercigroch A., 1994,
ApJS, 94, 127
Giddings J., 1983, IUE ESA Newsletter 17, 53
G/l¸ebocki R. & Stawikowski A., 1995, Acta Astron. 45, 725
Howarth I.D., Murray J., 1988, STARLINK User Note 50, 10
Jefferies R.D., Byrne P.B., Doyle J.G., Anders G.J., James D.J.
& Lanzafame A.C., 1994, MNRAS 270, 153
Menzies J.W., Cousins A. Banfield R.M. & Laing J.D., 1989,
S. Afr. Astr. Obs. Circ. 13, 1
Miller A. & Abbott, M., 1995, EUVE Guest Observer Software,
Version 1.5
Pallavicini R., Monsignori­Fossi B.C., Landini M. & Schmitt
J.H.M.M., 1988, A&A 191, 109
Pallavicini R., Tagliaferri G. & Stella L., 1990, A&A 228, 403
Pan H.C. & Jordan C., 1995, MNRAS 272, 11

P.J. Amado et al.: Rotational modulation and flares on RS CVn and BY Dra stars. 11
Pettersen B.R., 1983, in P.B. Byrne & M. Rodon`o, eds, Activ­
ity in red­dwarf stars, Dordrecht, D. Reidel Publishing Co.
p. 17
Phillips M.J. & Hartmann L., 1978, ApJ 224, 182
Schmidt­Kaler T., 1982, 2b, Landolt­B¨ornstein, Springer Hei­
delberg
Schmitt J.H.M.M., Pallavicini R., Monsignori­Fossi B.C. &
Harnden F.R.J., 1987, A&A 179, 193
Tody D., 1986, in D. Crawford, ed, Instrumentation in Astron­
omy VI., Proc. SPIE p. 733
Tsikoudi V., 1988, AJ 95, 1797
van den Oord G.H.J., 1988, A&A 205,167
White S.M., Lim J. & Kundu M.R., 1994, ApJ 422, 293
Zombeck M.V., 1990, In Handbook of Space Astronomy & As­
trophysics, Cambridge University Press