An investigation of GSC 02038-00293, a suspected RS CVn star, using CCD photometry.
We have undertaken a differential photometric study of GSC 0203800293, known to display short period modulations in optical magnitude and to emit X-rays, with a view to characterising the physical properties of this source.
The target is located at RA 16h 02m 48.22s and Dec +25° 20' 38.2" in the constellation of Serpens Caput.
The star was identified as an object warranting further investigation by Norton et al. (2007), (1) with the suggestion it was of the type RS CVn (RS Canum Venaticorum). That paper presents optical light-curves obtained during the SuperWASP13 photometric survey for 428 periodic variable stars coincident with ROSAT14 X-ray sources. (It should be noted that GSC 02038-00293 is known in the Norton paper (1) by its SuperWASP name of 1SWASP J160248.22 +252038.2).
RS CVn stars traditionally represent a class of close detached binaries with the more massive primary component being a G-K giant or sub-giant and the secondary a sub-giant or dwarf of spectral classes G to M with a temperature difference of ~1000K, though Pandey (2005) (2) also states that there are no implicit restrictions on the class of the secondary. RS CVn type stars show optical variability (outside eclipses) which is characterised by an amplitude up to 0.6 mag in the V band and interpreted as the rotationally modulated effect of cool spots on their surfaces, a result of increased chromospheric activity. The X-ray emission is an indicator of an active coronal region believed to be the result of rapid rotation and enhanced magnetic fields.
Fast rotation, Ca II, H and K emission lines and a sub-giant component well within its Roche lobe are usually taken to be necessary features. Short-period RS CVns have a period <1 day, according to Hall et al. (1976). (3)
At the time of our observations, there were few previous references to GSC 02038-00293 to be found. Bernhard & Frank (2006) (4) and Frank & Bernhard (2007) (6) first suggested it as an eclipsing RS CVn binary with an orbital period of 0.495410 days. This is similar to the period indicated by SuperWASP observations (Norton et al. 2007). (1) Bernhard & Frank (2006) (4) also estimated a long starspot activity cycle generating a 6 to 8 year cycle.
Several weeks after our observations, Korhonen et al. (2010) (16) obtained intermediate Ha and low resolution spectroscopic data showing GSC 02038-00293 to be a mid-type K star with [T.sub.eff] 4750 and v sini=90 km/s, indicating a rapidly rotating star.
We observed GSC 02038-00293 regularly on 10 nights over a period of 34 days in 2010 May and June.
At the time of our observations, the Physics Innovations Robotic Astronomical Telescope Explorer (PIRATE) facility comprised a remotely controllable 35cm, f/10 Schmidt-Cassegrain telescope (Celestron-14) equipped with an SBIG STL 1001E CCD camera with 1024x1024 24ujn pixels, resulting in a field of view of 21 arcmin and a pixel scale of 1.21 arcsec per pixel. (For reference, in 2010 August the OTA was replaced with a Planewave CDK-17 f6.8 astrograph. Refer to Lucas & Kolb (2011). (18))
The CCD camera is equipped with an 8 position filter wheel and has 5 broadband filters of which only R, V and B were used for this study.
The OTA is mounted on a Paramount ME, a robotic German Equatorial Mount manufactured by Software Bisque, all housed in a remotely controllable 3.5m dome manufactured by Baader Planetarium, which is on top of the main observatory building at the Observatori Astronomic de Mallorca (OAM), Longitude E 2° 57' 06"; Latitude N 39° 38' 38"; Altitude 203m. (See PIRATE website http://pirate.open.ac.uk/).
The main user control interface is the software program ACP Observatory Control (by DC-3 Dreams) with TheSky driving the Paramount mount. Data reduction and processing, and display were done using MaximDL version 5.08 from Diffraction Limited.
The observers' log is summarised in Table 1. It shows the dates observations were made; the start time of each automated run; the run reference letter allocated to the images; the filter used; the exposure length in seconds; and finally, the number of exposures in the run. Where multiple filters were used in a single automated run the bracketed numbers show the number of exposures per filter. So 10x(3,2,3) gives 30 V filter, 20 B and 30 R.
The finder chart showing the location of GSC 02038-00293 (shown by its SuperWASP name 1SWASPJ160248.22+252038.2) and annotated with reference and check stars for photometry is reproduced in Figure 1.
The field of view was deliberately offset for GSC 02038-00293 in order to provide a more appropriate reference star. The image was centred at 16:03:12.22 +25:20:38.2 (J2000) from May 13 onwards. It should be noted that for the evening of May 10, the reference star lay nearer the edge of the frame.
Table 2 details the reference and check stars used in the analysis of frames from all nights.
Data reduction and photometry
All data were bias and dark frame subtracted and then flat fielded with twilight flats using Maxim DL. Aperture settings of 5 pixels, gap 8 pixels, and annulus 3 pixels were used.
Using the period analysis software package PERANSO17 and its Analysis of Variance (ANOVA) fitting method, the best estimate of the orbital period obtained is 0.4955 ± 0.0001 days.
The graphs in Figures 2 and 3 have been phase folded using that orbital period with zero point at JD 2455327.614. Figure 2 shows two complete phases of V-band data. Figure 3 shows two phases of data from evenings where observations were performed in B, V and R-band.
It can be seen from both Figures 2 and 3 that the primary eclipse, shown at phase 1, is very well defined. The secondary eclipse, suspected at phase 0.5 is less well defined due in part to the increased scatter around phase 0.5-0.8 and the incompleteness of our B and R-band curves and the asymmetry of the two maxima. The primary eclipse itself is slightly asymmetric with indications of a bell-shaped curve at the minimum as opposed to a sharp change.
Both the R and B-band curves show broadly the same features as in V. One notable difference is the sharp upturn in brightness at phase 0.6 following the secondary minimum, most notably in B. There is also a less pronounced shelf coming out of the primary minimum in B.
Figure 4 shows a magnified section of the V-band curve between phase 0.4 and 0.8 containing the increased scatter mentioned above. The data here are displayed according to the date the observations were made to better illustrate the observed changes. The evenings of May 10, May 25 and June 13 show a marked increase in scatter and brightness around phases ~0.55 and ~0.78. Over the period from May 13 to May 22 there seems to be a slight increase in overall brightness, perhaps a prelude to the subsequent high-scatter data on May 25 but this is unclear.
B-V changes during eclipse
We observed a reddening at the primary minimum, see Table 3. The calculated B-V figures point towards a spectral class G which is in line with RS CVn class systems. This assumes a negligible contribution from the non-active binary component. Korhonen et al. (2010) (16) obtained spectroscopic data showing GSC 02038-00293 to be a mid-type K star.
The spread in our V data meant we were unable to confirm any reddening at the secondary minimum that would indicate starspot activity as observed by Bernhard & Frank (2007). (6) Pandey (2005) (2) also mentions that reddening outside of the eclipse lends support to the starspot hypothesis.
Uncertainties and corrections
Figure 5 represents a typical observing night's data showing the relative magnitude between the reference star, target and check stars. The curves have been offset vertically.
Most notable in Figure 5 is the step in the data, seen after the telescope underwent a pier flip.
Data collected from the mark one PIRATE facility demonstrated a systematic gradient in the East-West plane that calibration could not remove. As our target and reference star both lay in this plane the FWHM differences before and after the meridian flip were most accentuated by this effect, which manifested itself as a 'step' in the pre and post meridian flip data. (This effect is discussed at length in Holmes et al. (2011). (5) PIRATE MKII does not exhibit this gradient).
From our study of this effect we chose to apply a uniform post-flip correction to the data for each band: B (-0.055), V (-0.045), R (-0.035). The confidence of the ANOVA fit in PERANSO was used in determining these values. Figure 6 shows the uncorrected V data with pre- and post-flip data shown separately. The good correlation in the two curves allowed us to be confident in the uniform pier-flip correction.
We have chosen not to show uncertainty bars on the lightcurve plots for clarity. Typical photometric uncertainties were calculated as 0.03 magnitude. An example of such uncertainties applied to our data is provided in Figure 7, and shows the lightcurve around minimum on May 16. As can be seen, two thirds of the data points lie on, or are touching, the added polynomial trend-line indicating good data quality for this night.
Table 4 shows the standard deviation of check star no.1 from the analysis and gives an indication of the seeing on any particular observing night.
Our period is calculated as 0.4955 ± 0.0001 days, significantly different from the All Sky Automated Survey (15) period of 0.330973 days. The ASAS data is noisy and less well sampled, which may be one reason for the discrepancy.
Bernhard & Frank (2006-2010) (4,6) have studied the lightcurve of GSC 2038-0293 since 1999. They have observed that while the magnitude of the primary minimum stays roughly the same, the magnitude of the secondary minimum varies from year to year, which is believed to be caused by starspot activity. Our observations support these findings with a lightcurve that fits both in general shape and a secondary-minimum cycle in close agreement to the proposed 6 to 8 year starspot cycle. See Figure 8, where we have superimposed our data (for 2010) onto existing 1 SWASP data from years 2004, 2007 and 2008.
There are marked similarities between our data and the 2006 curve by Bernhard & Frank (4) which suggests a shorter cycle of ~6 years duration rather than the previously quoted 6-8 year cycle. (Note: The 2004 and 2007 data used only one SWASP survey camera, which greatly reduces the problems when comparing magnitudes. 2006 uses two different cameras and also some data from two further cameras. This explains the increased spread of the data for that year).
When the peak to peak amplitudes of the secondary minima are plotted over time, the pattern is more apparent. Figure 9 shows how the amplitude of the second half of the lightcurve (phase 0.25 onwards) shows two clear maxima in 1999 and 2005 and minima in 2001-2002 and 2007. 2007 was also the year of lowest starspot activity so far recorded. This suggests that there are cyclical variations, possibly in the order of 6 to 8 years. Similar cycles have been observed for other RS CVn stars (Berdyugina & Tuominem 1998). (7)
Bernhard & Frank (4,6) state that the starspot activity on the larger star is the chief cause of the variations in the depth of the secondary minima. Times of low activity correspond to the datasets where the secondary minima are all but absent. The implications of this are that the secondary eclipse is either very shallow or barely detectable, which places restrictions on the size/temperature of the secondary component.
Kozhevnikova et al. (2007), (8) in a study of three RS CVn systems state that starspotted regions were concentrated at low latitudes up to 32°, that the spots covered up to 29% of the stellar surface and all stars showed non-axisymmetric spot distributions (active longitude structures separated by approximately half the orbital period). This half-phase separation seems to be a common factor in this type of system. The active regions can migrate or change but seem to remain intact over long periods. There is also evidence from these papers of short-term variability on timescales of weeks rather than years.
Our data also show scatter around the secondary minimum lightcurve, supporting the probability of short term variability and the presence of starspots, and hence implying increased coronal activity, as do Korhonen et al. (2010) (16) who observed variable line strength in Ha spectroscopy.
The lightcurve obtained by Norton et al. (2007) (1) is similar in structure to our phased lightcurve but it is clear that their secondary minimum appears to occur earlier in the cycle than in our 2010 data. This may indicate that starspots migrate, and/or vary in size and position.
The anomalies in the data referred to in the Results section are discussed here.
1 We note a general brightening over part of the lightcurves of the target over the nights from May 13 to 25 which requires further investigation. This may be due to calibration issues but the observations are not conclusive.
2 Figures 2, 3 and 4 show a marked increase in scatter for V data beginning at phase 0.5, just prior to the secondary minimum. This appears for data collected on May 10 and June 13 as a collection of data points between phases 0.5 and 0.6. An increase in the slope of the curve is also apparent in the R and B traces for observations on June 13, as sharp upturns at 0.6. Unfortunately the B and R data immediately following this phase were not obtained. Scatter around the secondary minimum caused by starspot activity is a known feature of RS CVn binaries. Our observations may indicate that, as well as varying over several years, there is shorter term variation due to starspot activity.
3 Figures 2 and 4 also show a second area of data point scatter at phase 0.8 visible in V band traces. This clump was only observed on one observing night on May 25, suggesting either a genuine transient event at the target, or local environmental problems at the observatory. We are aware that flares are common in chromospherically active RS CVn stars and a similar increase can be seen in the lightcurve of Figure 1 of Zhang (2010)12 at phase 0.35. However, their flare data profile is much more sharply defined than ours and we have concluded that this is unlikely to be evidence of a flare.
RS CVn classification
Due to the short period of ~0.4955 days deduced from our observations we can be reasonably certain that tidal forces will ensure that the star's rotation will be synchronised with the orbital period. This is confirmed by Korhonen et al. (2010), (16) who using low resolution spectroscopy deduced the rotation period to be 0.495410 days.
The start and end points of the primary eclipse are also clearly defined despite the magnitude asymmetry, indicating the components are detached.
Dragomir et al. (2007) (11) obtained evidence of CaII, H and K lines with strong core emission, which strongly supports RS CVn classification.
Figure 10 provides a qualitative interpretation of the observed lightcurve based on our understanding of known RS CVn stars. The assumption is that there is a marked difference in radii (as stated above) in that one star is a giant and the other a dwarf.
The basic configuration shown in Figure 10 qualitatively supports the changes in the flux we observed. Point D is the point of greatest luminosity with both stars showing unspotted faces in configuration 4. Point C is the point of lowest luminosity, that being the primary eclipse as shown in configuration 3.
Point B, at configuration 2, is similar to D but with a starspotted primary face which causes a drop in luminosity compared to the reference point D.
Point A should be less luminous than point B but the difference in luminosity will depend on the ratio of the stars' radii, the difference in temperature and the amount of starspotting. If the secondary is a small yellow dwarf with a temperature around 5,000K and since the primary is shown by Korhonen et al. (2010) (16) to be a late K type with a temperature of 4,750K then the radii difference could be >10:1. This gives a large stellar disc surface area difference and a secondary eclipse may be difficult to ascertain.
It is clear that further modelling, requiring more time and specialist modelling software, should be conducted to establish the actual configuration.
This model also supports the 'reddening' effect that we see in our data and in previous papers.
Further spectroscopy is required to confirm the secondary component (suggested by Korhonen et al. (2010). (16) The determination of the mass ratio and radii would help to refine the starspot models. It would also be interesting to see a 'cleaned' curve that removes the effect of the starspots and to verify any additional cyclic variations in the target.
We find GSC 02038-00293 to be a short period eclipsing RS CVn star. From our data alone we calculate the following ephemeris: JD 2455327.614 + 0.4955(1) x E
This period is in close agreement with that of Bernhard & Frank (2006) (4) and Norton et al. (2007) (1) but significantly different from that found by the All Sky Automated Survey 15 of 0.330973d.
The eclipse period is very close to the rotation period calculated by Korhonen et al. (2010) (16) of 0.495410 days suggesting a locked synchronous orbit.
Our data show a clear primary eclipse with notable asymmetry. The secondary minimum is extended and subject to short-term variability. From our observations and the literature, we believe that this is likely to be the result of starspot activity. Due to this we cannot conclusively say the secondary eclipse was detected.
The depth and short-term variability of the secondary minimum is most likely to be a result of starspot activity on one of the system components. This is in line with Bernhard & Frank (2006) (4) and also lends support to their observed 6-8 year cycle of variability in the depth of the secondary minimum. The observed reddening at the minima would also lend support to this hypothesis.
GSC 02038-00293 is a rare, active, very-short period eclipsing RS CVn star which warrants further investigation, specifically the following:
--Detailed spectroscopy at other epochs to determine the presence of the binary component;
--Detailed lightcurve modelling allowing subtraction of binary curve effects to allow proper modelling of the spot sizes and locations;
--Continuing B, V & R photometric observations to follow secondary changes over the 6-8 year cycle and to determine the nature of the B-V & V-R anti-correlation.
[Members of the 2010 Open University Module S382 Undergraduate Programme]
Special thanks to Andrew Norton, Ulrich Kolb, Stefan Holmes and James Smith of the Open University's PIRATE observatory for their guidance and feedback. Also to Jay Jina, Colm Kilmurry, Brad Rose and Cyrille Tijsseling for their assistance with observing and analysis.
The authors wish to thank the referees whose comments and general guidance greatly improved the paper.
Address: (TR) 1, Rivermede, Ponteland, Newcastle-upon-Tyne NE20 9XA. [firstname.lastname@example.org]
(1) Norton A. J. et al., 'New periodic variable stars coincident with ROSAT sources discovered using SuperWASP', A&A, 467, 785-905 (2007)
(2) Pandey J. C. et al., 'Unravelling the nature of HD 81032--a new RS CVn binary', J. Astroph. Astron. 26(4), 359-376 (2005)
(3) Hall D. S., 'The RS CVn binaries and binaries with similar properties', in Multiple Periodic Variable Stars, (Ed.) Fitch W. S., Proceedings of IAU Colloquium 29, Budapest, Hungary, 1975 Sept 1-15, 1975; Astrophys & Space Science Lib. 60, p. 287, Reidel, Dordrecht & Boston, 1976
(4) Bernhard K. & Frank P., 'GSC 2038.0293 is a new short-period eclipsing RS CVn variable', IAU Inf. Bull. Var. Stars, 5719, 1-4 (2006)
(5) Holmes S. et al., 'PIRATE: A remotely-operable telescope facility for research and education', PASP, 123, 1177-1187 (2011)
(6) Frank P. & Bernhard K., 'CCD photometry of the short-period eclipsing RS CVn variable GSC 2038.0293', Open Eur. J. on Var. Stars, 71, 1 (2007)
(7) Berdyugina S. V., 'Starspots: A key to the stellar dynamo', Living Reviews in Solar Physics, 2, 8 (2005)
(8) Kozhevnikova A. V. et al., 'Long-term starspot activity of three short- period RS CVn stars: BH Vir, WY Cnc and CG Cyg' A&A trans., 26(1-3), 111-112 (2007)
(9) Bernhard K. & Frank P., 'Wie lange dauert ein Sternfleckenzyklus auf GSC 2038.0293?', BAV Rundbrief, 57, 163 (2008)
(10) Sarma C. V. S. et al., 'A study of the distortion wave in the RS CVn eclipsing binary SV Camelopardalis', J. Astroph. Astron. (ISSN 02506335), 12, 49-67 (1991)
(11) Dragomir D. et al., 'Spectral classification of optical counterparts to ROSAT all-sky survey X-ray sources', Astron. J., 133, 2495-2501 (2007)
(12) Zhang L. et al., 'A CCD photometric study of the newly identified RS CVn star DV Piscium,' New Astron., 15(4), 362-366 (2010)
(16) Korhonen H. et al., 'Photometric and spectroscopic observations of three rapidly rotating late-type stars: EY Dra, V374 Peg and GSC 0203800293', Astron.Nachr., 331(8), 772 (2010)
(18) Lucas R. J. & Kolb U., 'Software architecture for an unattended remotely controlled telescope', J. Brit. Astron. Assoc., 121(5), 265-269 (2011)
Received 2011 July 11; accepted 2012 May 30
Table 1. Journal of observations of GSC 02038-00293 Date Start Run Filter Exposure No. of (2010) time (UT) ID length (secs.) exposures May 10 20:24 A V 30 30 20:56 B V 30 60 21:52 C V 30 3 high humidity forced early dome closure May 13 20:13 A V 60 60 21:36 B V 60 60 23:05 C V 60 9 23:22 D V 60 19 23:53 E V 60 27 increasing haze/cloud ended observations early May 16 20:25 A V 30 60 21:15 B V 30 60 22:08 C V 30 60 23:01 D V 30 60 23:53 E V 30 28 00:38 F V 30 100 01:58 G V 30 100 slight haze, good conditions May 22 21:06 A V 30 60 21:57 B V 30 120 23:42 C V 30 60 good conditions May 25 20:22 A V 60 6 20:33 B V 30 30 20:58 C V 30 30 21:36 E V, B, R 30 50 each filter 23:45 F V, B, R 60 20 each filter 01:10 G V, B, R 30 20 each filter 02:00 D V, B, R 30 30 each filter clouds at start, improving towards end, bright Moon May 28 20:36 A-D V, B, R 30,60,30 30 each filter 00:05 E V, B, R 30,60,30 30 each filter 01:38 F V, B, R 30,60,30 30 each filter weather prevented early observations, near full Moon May 31 20:32 A V, B, R 30,60,30 24 each filter 21:42 B V, B, R 30,60,30 3x (3,2,3) 23:03 C V, B, R 30,60,30 10x (3,2,3) 00:25 D V, B, R 30,60,30 10x (3,2,3) 01:46 E V, B, R 30,60,30 10x (3,2,3) 02:56 F V, B, R 30,60,30 5x (3,2,3) later cloud, dome failure during eclipse, bright Moon June 3 20:23 A V, B, R 30,60,30 10x (3,2,3) 21:45 B V, B, R 30,60,30 10x (3,2,3) 23:05 C V, B, R 30,60,30 10x (3,2,3) 23:43 D V, B, R 30,60,20 10x (3,2,3) 00:50 E V, B, R 30,60,20 10x (3,2,3) 02:04 F V, B, R 30,60,20 10x (3,2,3) good conditions, R exposure lowered due to saturation June 13 20:34 A V, B, R 30,60,30 10x (3,3,3) 22:08 B V, B, R 30,60,20 1x (3,3,3) 22:30 C V, B, R 30,60,20 10x (3,3,3) 23:58 D V, B, R 30,60,20 10x (3,3,3) 01:24 E V, B, R 30,60,20 7x (3,3,3) good conditions, cloud at end of night Table 2. Reference and check stars for GSC 02038-00293 Name RA, Dec (J2000) Ref star1 GSC 0238-0652 16 03 38.05, +25 21 10.35 Check star 1 GSC02038-00987 16 03 26.80, +25 13 54.3 Check star 2 CMC14 J160300. 16 03 00.66, +25 14 06.4 6+251406 Check star 3 GSC02038-00648 16 02 46.26, +25 22 55.9 V mag B mag Ref star1 9.98 ± 0.3 * 10.694 ± 0.034 Check star 1 11.51 ± 0.37 -- (‡) Check star 2 13.363 * -- Check star 3 13.34 ± 0.4 -- (‡) * Data from VizieR using the Carlsberg Meridian Catalogue 14 (CMC14) I/304. (‡) Data from GSC v1.1, photographic magnitudes. Table 3. Approximate magnitudes for different parts of the cycle of GSC 02038-00293 (The numbers in brackets are corrected against our reference star, assuming the values B-band= 10.694 and V-band= 9.98 obtained from CMC14.) V B B-V Primary eclipse 0.70 (10.68) 0.89 (11.58) 0.19 (0.90) Primary maximum 0.46 (10.45) 0.62 (11.31) 0.16 (0.86) Table 4. Magnitude errors (The standard deviation seen in the magnitude of check star 1 for each observing session as an indication of photometric uncertainty in the data.) Date (2010) Pre-flip Post-flip May 10 0.009 Not applicable May 13 0.006 Not applicable May 16 0.007 Not applicable May 22 0.005 0.006 May 25 0.017 0.007 May 28 Not applicable 0.007 May 31 0.021 0.020 June 3 0.006 0.007 June 13 0.008 0.008
|Printer friendly Cite/link Email Feedback|
|Author:||Bruce, Alastair; Cruickshank, Stewart; Rodda, Tony; Salisbury, Mark|
|Publication:||Journal of the British Astronomical Association|
|Date:||Aug 1, 2013|
|Previous Article:||Saturn during the 2007/2008 apparition.|
|Next Article:||Ordinary Meeting, 2013 January 30: held at the Royal Astronomical Society, Burlington House, Piccadilly, London WI.|