next previous
Up: Photometric variability of

3. Photometric variability

The largest photometric campaign of Southern TTS is presented by Covino et al. (1992), after monitoring a sample of 30 stars in tex2html_wrap_inline1468. They confirm the previously reported period of CV Cha, Sz 6 and Sz 68 and establish new periods for four other stars (AS 205, Wa Oph/2, WA Oph/3 and Wa CrA/2). They are unable to confirm previously reported periods for T Cha and S CrA or to find any periodical behavior for SZ Cha and GQ Lup.

We analyze the light curves of the 26 T Tauri stars listed in Table 1 where the first 13 have only recently been identified as PMS stars.

Table 2 shows the periods determined in this work and the previously published ones. We label each star in Table 2 as weak or classical following the earlier criteria based solely on the strength of the Htex2html_wrap_inline1522 emission: wTTS if EWtex2html_wrap_inline1524 is less than 10 Å  and cTTS if larger. However, the most complete set of criteria to distinguish cTTS from wTTS is discussed by Edwards et al. (1993). These are the near infrared color excess, the optical continuum veiling and the forbidden emission line of [OI] at tex2html_wrap_inline1526 6300. Such data are presently lacking for the majority of our targets.

   

PDS/HBC w/cHtex2html_wrap_inline1522 (Å) Lit. tex2html_wrap_inline1532 tex2html_wrap_inline1532
PDS 1 w0.21--
45w2.01--
50w12.5/8.01--
59w5.01-11.73
66c 471-5.75
70w2.01-5.10
77w2.01- 2.50:
81w0.71--
82c/w29.0/9.01--
83w10.01-5.52
89w2.21--
99w5.01--
101c 521--
HBC 565 c 492 7.63, 6.24, 6.15,65.97
567 c262 8.63, 8.04, NP6NP
570 c 562-9.86:
578 c 71274, 7.23 2.56
583 c 522-5.34
590 c 322-13.00:
591 c2.013.244,10, NP7 -
603 c 172- NP
605 w 7.02NP7NP
620 w0.52 -4.05:
656 c5483.12,NP
657 c 2884.05
663/664c1025.2011, 4.8395.15
Table 2: Published Data for the Program Stars and Computed Periods. In the first column, the PDS or HBC numbers are listed. These pre main sequence stars are classified in the second column as wTTS (w) or cTTS (c) if indications of an accretion disk are present or not. See text for details. Htex2html_wrap_inline1522 line strengths are indicated in the third column. Periods that are found in the literature and determined in this work are respectively listed next

1) GHETAL92, 2) Herbig & Bell 1988, 3) Kappelman & Mauder 1981, 4) Mauder & Sosna 1975, 5) Schaeffer 1983, 6) Bouvier et al. 1986, 7) Covino et al. 1992, 8) Torres CAO, private communication, 9) Torres et al. 1983 (see text), 10) Hoffmeister (1965), 11) Chugainov (1974).

The EWtex2html_wrap_inline1524 are taken from the HBC or PDS. For a low mass PMS star, Htex2html_wrap_inline1522 line strength indicates (but not conclusively) the presence of a circumstellar disk since important components of this line are formed either in a wind or in an accretion column - both scenarios requiring disks (Johns & Basri 1995). Thus, the Htex2html_wrap_inline1522 of Hen 892 and BZ Sgr suggest the presence of a disk. All the remaining PDS stars have small but overt EWtex2html_wrap_inline1524. Further observations of these wTTS must be carried out, especially in the near infrared, in order to constrain their circumstellar environments.

3.1. Stars with no firm detection of periodicity

3.1.1. PDS 1 (Hen 1), PDS 89 and PDS 99

These objects were not observed with enough regularity to enable time series analysis. There are only 4 observations per star.

Hen 1 is not an IRAS source and does not indicate near infrared excess of luminosity (GHETAL92). It is situated at high galactic latitude tex2html_wrap_inline1622 and is not associated with any known molecular cloud. The level of variability (as deduced solely from our 4 observations) is one of the lowest in our sample - at a level which is unusual for pre-Main Sequence stars.

PDS 89 is a serendipitous discovery of PDS since it lies outside of the error box of an IRAS source. The source is actually associated with PDS 149 in the small cloud L152 where HBC 651 is located (Torres et al. 1995). PDS 99 belongs to the CrA association and was previously identified as an Htex2html_wrap_inline1522 emission line star by Marraco & Rydgren (1981).

3.1.2. PDS 45 (CoD-29 8887) and PDS 50 (Hen 600)

These two PDS stars together with PDS 54, PDS 55 and the cTTS TW Hya lie within a circle of about 7tex2html_wrap_inline1626 at high galactic latitude tex2html_wrap_inline1628. de la Reza et al. (1989) attempt to physically connect these objects as members of a common parental cloud which has already dissipated. Jensen et al. (1996) did not detect PDS 45 nor Hen 600 at 800 tex2html_wrap_inline1630m but did detect PDS 55, which curiously is not an IRAS source. We have less than 4 entries for PDS 54 and PDS 55, and they are not included in this work.

Our photometry in the V-band of PDS 45 and 50 shows the low range of variability typically ascribed to solar-type active regions on the stellar surface (see Rodonò 1986). The amplitude of the variations reaches merely a few times the intrinsic errors in our measurements thereby preventing us from establishing a reliable rotational period.

3.1.3. PDS 77 = CoD-41 10484

PDS 77 has the spectroscopic characteristics of a wTTS and is a member of the Lupus association. If the strength of EWtex2html_wrap_inline1524 alone is the indicator of a circumstellar disk then PDS 77 is a wTTS and probably devoid of an active disk. However, Herbst et al. (1994) argue that wTTS generally have light fluctuations less than 0.8 mag in V. Only accretion disks are capable of powering light variations that are larger than this value in the V-band.

We have gathered a sparse and small number of data points (7). Nevertheless, we observed a significant range of variability in all colors (0.91 at V). Several possibilities are indicated by the DCDFT code and they all cluster about 2.50 days with mild acceptance. We suggest this as a preliminary result and we stress that additional observations are needed to confirm this.

3.1.4. PDS 81 and PDS 82 = VV Sco

These two stars are found in the same region of the tex2html_wrap_inline1642 Oph complex. PDS 81 presents a range of variability of 0.15 mag in all colors. It is a wTTS with Htex2html_wrap_inline1522 sometimes filled in. It is associated with the cloud B40 for which HBC lists 5 TTS, and the PDS finds another 3. A total of 11 exposures in tex2html_wrap_inline1468 were gathered in 1988/1989 (4/7) and the BSP algorithm does not indicate any reliable period. The DCDFT method indicates a period of 11.11 days but with a modest degree of confidence (62tex2html_wrap_inline1648 in V) which leads us to not include it in the pool of TTS with determined rotational periods.

VV Sco is a binary system with a separation of 1.5''. It lies at the border of the complex and cannot be resolved photometrically. We search for periodical patterns in the light curve with no conclusive results.

3.1.5. PDS 101 = BZ Sgr

The PMS nature of BZ Sgr was established quite recently (GHETAL92) but its Htex2html_wrap_inline1522 profile in emission had been known since earlier prism surveys (Stephenson & Sanduleak 1977). GHETAL92 associated this object with the molecular cloud No. 159 (Magnani et al. 1985). We discover a very faint companion less than 5'' apart. This possible companion has a firm emission in Htex2html_wrap_inline1522 but the noise prevents any conclusion as to the existence of the Li I line.

This star presents remarkable variability and striking color-dependent amplitudes that are typical of cTTS. The strength of its Htex2html_wrap_inline1522 also leads us to believe that this object is a classical TTS. Jensen et al. (1996) detect the star at 800 tex2html_wrap_inline1630m which is additional evidence for the presence of a circumstellar disk.

Neither algorithm was able to find any periodic pattern for BZ Sgr using our data set.

3.1.6. HBC 567 = TW CHA

Previous periods of 8.6 and 8.0 days have been reported for TW Cha as well as a null result (see Table 2). TW Cha is a cTTS with a reportedly large degree of variability (Bouvier et al. 1986). This is confirmed here. In spite of the large amplitude, no convincing period-folded light curve presented itself. The reported periods of 8.0 and 8.6 days do not fold our data into a smooth light curve.

3.1.7. HBC 570 = CT CHA

We gather 10 entries in 1995 distributed over two months for CT Cha. It presents, in general, mild variability. The BSP algorithm and the DCDFT code indicate two periods with observations well distributed in phase and have large confidence levels: 9.86d and 13.68d. Several other possible periods with acceptance larger than 90tex2html_wrap_inline1648 cluster at about the former value. We tentatively suggest 9.86d as the computed period for the season.

3.1.8. HBC 591 = T CHA

The pre Main Sequence nature of T Cha is established by Alcalá et al. (1993). In spite of an EWtex2html_wrap_inline1524 typical of weak TTS, these authors report evidence of disk accretion - as inferred from the inverse P Cygni profile in the Htex2html_wrap_inline1522 - and robust spectral variability. We also have several medium resolution data targeting the Htex2html_wrap_inline1522 line that show strong variability (the PDS archives). The line goes from absorption to emission on a time scale of days, as is also reported in Alcalá et al. (1993).

Hoffmeister (1965) made extensive visual observations and proposed a period of 3.2436 days. A periodicity of 3.2d is also found by Mauder & Sosna (1975), although not confirmed by the more recent observations of Covino et al. (1992). This star consistently undergoes large photometric and spectroscopic variability (Covino et al. 1992; Alcalá et al. 1993).

Neither the BSP nor the DCDFT method reveal any conclusive period for our modest sample. Due in part to the variability revealed in the medium resolution data, we re-analyzed the data sets of Hoffmeister (1965) and Mauder & Sosna (1975) using the DCDFT method. We averaged their data into a maximum of 3 data points per night. Hoffmeister's observations give a period of 3.221 days with a false alarm probability of 10-8. No period was found for the Mauder & Sosna data set unless the observations taken after JD = 244141399 are omitted, in which case we are able to reproduce the period of 3.22 days. Our 7-exposure light curve does not fold smoothly into a 3.22 day period.

We hypothesize that T Cha is another case of a TTS that undergoes phases of periodical behavior (see for example Vrba et al. 1989) and phases of irregular variations.

3.1.9. HBC 590 = SZ 45 and HBC 620 = SZ 108

These two stars are cTTS that show mild to low variability. Because of weather conditions we gathered a maximum of 10 points distributed over two consecutive years.

For SZ 45, the BSP periodogram method indicates two periods at 14.12 days and 14.75 days with comparably low tex2html_wrap_inline1496's. The period-folded light curves, however, are not well distributed in phase. The DCDFT method computes several possibilities, all clustering about a period of 13.00 days. We suggest the latter as a tentative value.

For SZ 108, both methods indicate similar solutions at about 4.05d with a high level of confidence. However, the data are not well distributed in phase and due to the modest number of entries we regard the result as tentative.

3.1.10. HBC 603 = SZ 77

SZ 77 shows a moderate range of variability in the course of our campaign (Table 1). Analysis of the near infrared excess indicates active disk accretion in this object (Batalha & Basri 1993). Neither methods provide definitive period determinations.

3.1.11. HBC 605 = SZ 82

SZ 82 was monitored by Covino et al. (1992) and no periodicity was found. It has signatures of disk accretion, with a mild near infrared excess indicative of reprocessed stellar light by overlying circumstellar dust. The ratio between the stellar and systemic luminosities tends to reinforce the presence of an active circumstellar disk (Batalha & Basri 1993).

Two possible periods are indicated by the BSP: 1.16d and 1.61d. The DCDFT method does not indicate a conclusive solution for the V-band, nevertheless two maxima in the power spectra are achieved at 7.42d and 1.55 d, both aliases. The confidence of both periods are 78tex2html_wrap_inline1648 which leads us to conclude that no periodicity was established for SZ 82.

3.2. Stars with detected periodicity

3.2.1. PDS 59

PDS 59 is situated close to T Cha and may have a faint companion at 12 arcsec. The EWtex2html_wrap_inline1680 (500 mÅ) and the EWtex2html_wrap_inline1524 (less than 10 Å) are typical of wTTS (GHETAL92).

We have a total of 10 data entries distributed over three consecutive years (1989-1991) which is enough to cast suspicion on any result. Nevertheless, we proceed with the analysis having in mind that wTTS's have, in general, stable spot distributions (Herbst et al. 1994). The BSP method indicates 14.07d and 11.73d as possible periods, with the former folded light curve indicating a double wave shape. The DCDFT method reveals a peak in the power spectrum at 11.73d in all colors, except U, with a modest confidence level of 30tex2html_wrap_inline1648, 98tex2html_wrap_inline1648, 96tex2html_wrap_inline1648 and 83tex2html_wrap_inline1648 in the colors B, V, R and I respectively. The peak and formal acceptance are significantly raised if all colors are included simultaneously. We propose a tentative period of 11.73d and the folded light curve is presented in Fig. 1 (click here).

  figure334
Figure 1: Period-folded light curve for PDS 59

3.2.2. PDS 66 = Hen 892

Hen 892 is a cTTS on the basis of its strong Htex2html_wrap_inline1522, Li (370 mÅ) and IRAS colors. The large EWtex2html_wrap_inline1524 indicates accretion which implies the presence of veiling and/or color excess, especially in the blue where these effects are most strongly felt. Nevertheless, the amplitude of variability of Hen 892 is merely 0.1 mag in all colors. This variability is less than that found among active field stars (Rodonò 1986) and is atypical of cTTS. This object is isolated from any known star forming region.

We gathered a total of 9 entries in 1989/1990 (1/8). The BSP method indicates two acceptable solutions in the colors V, R and I: 4.67d and 5.71d. The DCDFT method indicates 5.75 tex2html_wrap_inline1714 0.03 days with a confidence of 97.2tex2html_wrap_inline1648 and 4.70 tex2html_wrap_inline1714 0.02 with a confidence of 93.5tex2html_wrap_inline1648. The criteria to adopt 5.75 days as the period for the season is based on the shape and the phase coverage of the folded light curve (Fig. 2 (click here)).

  figure344
Figure 2: The same as Fig. 1 for Hen 892

3.2.3. PDS 70 = CoD-40 8434

The Htex2html_wrap_inline1522 line of PDS 70 varies in between 2-4 Å. It shows a low range of variability (  0.1) which is independent of color.

We gathered 9 entries tex2html_wrap_inline1468 during 1989/1990 (1/8). The BSP method fits the data well at 5.1d or 5.6d, with comparable and low tex2html_wrap_inline1496 and good phase coverage. The power spectrum from the DCDFT method peaks at 5.1d and 4.26d with a large degree of confidence in the V-band. The confidence of the latter value is not consistently reproduced in the other colors. Therefore, we suggest a period of 5.1d during these two years (Fig. 3 (click here)).

  figure353
Figure 3: The same as Fig. 1 for PDS 70

3.2.4. PDS 83 = V896 Sco

A steady increase in the amplitude towards the blue can be seen in the light curves of V896 Sco. Our data set is very modest with one entry taken in 1989 and six others in the following year. The periodogram analysis done with the BSP method indicates two possible periods at 5.54 days and 7.93 days yielding good phase coverage and low tex2html_wrap_inline1496 values. The DCDFT method indicates 5.52 days with a large confidence level after adding the magnitudes of all colors. Thus, inspection of both period-folded light curves leads us to suggest 5.52 days as the rotational period of V896 Sco (Fig. 4 (click here)).

  figure360
Figure 4: The same as Fig. 1 for V896 Sco

3.2.5. HBC 565 = SY CHA

SY Cha is an active T Tauri showing, at times, complex light curve behavior which suggests the presence of more than one dominant spot (Bouvier & Bertout 1989) and changes in the spot distribution. Periodicities have been reported for this object in Bouvier & Bertout (1989); Schaeffer (1983); Kappelman & Mauder (1981) and Mauder & Sosna (1975). These authors report 6.0, 6.129, 7.6 and 6.2d, respectively.

We gathered a total of 22 observations during 1994 and 1995 (6/16) and we confirm the periodic behavior of SY Cha. The BSP algorithm indicates two periods at 5.97d and 6.07d. The former gives more complete phase coverage. The DCDFT method suggests 5.97 days as the most acceptable period for the season with confidence levels of 99.99tex2html_wrap_inline1648. We conclude that during our campaign SY Cha had a period of 5.97d and the final folded light curve is presented in Fig. 5 (click here).

  figure372
Figure 5: The same as Fig. 1 for SY Cha

3.2.6. HBC 578 = VZ CHA

VZ Cha is a typical representative of cTTS, with very large EWtex2html_wrap_inline1524 and IRAS fluxes. Studies based on spectroscopic observations indicate a period of about 7 days (Mauder & Schulz 1978). Another period of 7.2 days - determined with sparser time sampling - is also pointed out by Kappelman & Mauder (1981).

The BSP method does not indicate any significant minimum. The DCDFT algorithm presents peaks in the power spectrum at 2.577 days (99.8tex2html_wrap_inline1648 in I and 98.0tex2html_wrap_inline1648 in V), and, with lower significance, 12.00 days. We suggest 2.56 days as the final period, after inspecting the period-folded light curve (Fig. 6 (click here)).

  figure382
Figure 6: The same as Fig. 1 for VZ Cha

3.2.7. HBC 583 = WY CHA

We have photometry of WY Cha taken during two seasons. The BSP method indicates two acceptable periods of 5.43d and 2.10d, with the former showing better phase coverage. The DCDFT method indicates the period of 5.34 days with a very high level of acceptance. The former period gives the minimum tex2html_wrap_inline1496 for the adjusted curve and is chosen as the period for the season (Fig. 7 (click here)).

  figure389
Figure 7: The same as Fig. 1 for WY Cha

3.2.8. HBC 656 = AS 216

Our photometric data of AS 216 - a PMS star not associated with any known molecular cloud (Quast et al. 1987) - have been gathered since 1985. The EWtex2html_wrap_inline1524 computed at medium resolution (GHETAL92) spans values between 40 and 70 Å  which is indicative of disk-like activity. The photometric variability of AS 216 shows no significant change in the first three years. It reaches an amplitude of 0.15 mag in the V-band during the interval tex2html_wrap_inline1764, rises up to 0.25 mag during tex2html_wrap_inline1766, and then decreases again to 0.15 mag tex2html_wrap_inline1768. This behavior is consistent for all the colors. In the last two years (1989/1990) AS 216 went through an overall increase in variability, which is best observed in the U band.

Both algorithms indicate the periods 3.12d and 4.24d during the first three runs, independent of whether the data sets are used together or separately in the periodogram analysis. A slightly larger confidence is given to the 3.12d period leading us to adopt it as the photometric period of 1985/1986/1987. However, neither of these options reaches the minimum level of acceptance in the following two years (1989/1990). In fact, the BSP method gives a period of 6.69d which is not confirmed by the DCDFT method, in part because the period-folded light curve shape departs slightly from a sinusoidal one. We conclude that AS 216 is another case of a cTTS that undergoes periodical/aperiodical phases.

The folded light curve of the 3.12 day period (1985/1986/1987) is shown in Fig. 8 (click here). We note the occurrence of a maximum in brightness at phase 0.85 that is detected in the three colors. It is not an artifact of data reduction nor weather conditions. In fact, it indicates a significant increase of continuum emission which is not compatible with classical flare activity but with the optical veiling detected in PMS stars (Basri & Batalha 1990). We show in Fig. 9 (click here) the 1989/1990 photometry, folded into 3.12 days.

  figure406
Figure 8: UBV period-folded light curve for AS 216 during 1985, 1986 and 1987 observing season

  figure411
Figure 9: The 3.12 days period-folded light curve of AS 216 during the 1989/1990 season. Inspection in the Fig. 8 (click here) show that AS 216 is a typical case of cTTS that goes through phases of periodical/aperiodical behavior.

3.2.9. HBC 657 = AS 218

AS 218 has shown evidence of periodic or quasi-periodic modulation in several colors with possibly no connection to cold spots (Quast et al. 1987). It is also shown that this object is isolated from any molecular cloud (Quast et al. 1987).

The BSP method indicates periods of less than 2.0 days which are not included in our final analysis. The folded light curve with a period of 4.05 days obtained by the DCDFT method is indicated in Fig. 10 (click here).

  figure420
Figure 10: The same as Fig. 1 for AS 218

3.2.10. HBC 663/664 = FK Ser

FK Ser is classified as a post T Tauri Star by Herbig (1973) and a visual binary system with separation of 1.33'' (HBC 663/664). It presents 800 tex2html_wrap_inline1630m emission similar to that of PDS 55, and Jensen et al. (1996) conclude that FK Ser has disk properties similar to that of a typical TTS.

Quast et al. (1987) suggest that this system is isolated from any known molecular cloud, and they note that the periodic modulation is incompatible with those of cool spots. Chugainov (1974) finds a sinusoidal light variation with a period of 5.20 days and an amplitude of 0.1 magnitudes. Observations of this star in B and V done in 1974 at CTIO result in a period of 4.83 days, erroneously printed as 4.53 days in Torres et al. (1983). The origin of the apparent discrepancy in the resulting periods may reside in the fact that FK Ser is a binary system. Each component has a similar spectral classification. Therefore, the continuum modulation of each component, if present, will necessarily add to the final light curve explaining the difference in periods.

We have 30 entries taken during 10 nights in the 1986 run UBV and 17 taken during the 1988-1990 one tex2html_wrap_inline1468. In 1986 FK Ser shows large amplitude variations - a maximum in U. The period of 4.89d presents itself as the best solution of the BSP algorithm for both runs. The DCDFT method finds a period of 5.13d in the first run if the photometry of all colors are added simultaneously. The variability is conspicuous in the following run, but the indicated period of 5.15d has a lower confidence level than for the previous run. We suggest a photometric period of 5.15d for FK Ser (Fig. 11 (click here)).

  figure433
Figure 11: The same as Fig. 1 for FK Ser


next previous
Up: Photometric variability of

Copyright by the European Southern Observatory (ESO)