© ESO, 2011

1. Introduction

Since the pioneering discovery by Mayor & Queloz (1995) – a little more than a decade ago – of a planet orbiting the star 51 Peg, more than 450 other exoplanets have been found at the time of writing (e.g., Schneider 2010). The discovered planets have masses ranging from 4 Earth masses to 11 Jupiter masses. They can be found at distances of several AU or close to the parent star, with orbital periods ranging from a few days to a few years. High eccentricity is a common parameter connecting these lanets (e.g., Marcy et al. 2001).

These discoveries have inspired intensive studies of the physical properties of the planets and of their parent stars, including a possible star-planet interaction. Indeed, in an analogy to binary stars, which show a higher activity level compared to single stars, at very close distances, one might also expect planets to play a role in the level of activity of their host stars. Nevertheless, observational data, still at a very informative level, demonstrate that star-planet interaction appears to be more complex than the widely accepted star-star interaction in close binary systems. For instance, Shkolnik et al. (2003, 2008) report a planet-induced chromospheric activity on two stars with planets, HD 179949 and υ   And, apparent from the night-to-night modulation of the CaII H and K chromospheric emission phased with the hot Jupiter’s orbit. In addition, Kashyap et al. (2008) claim that stars with close-in giant planets are on average more X-ray active than those with planets that are more distant, an observational result consistent with the hypothesis that giant planets in close proximity to the parent stars could influence stellar magnetic activity.

By contrast, Poppenhaeger et al. (2010) find no significant correlations between X-ray luminosity and planetary parameters, suggesting no major average activity enhancement in the corona of stars with planets. Indeed, these authors find no additional detectable effects in coronal X-ray luminosity that could be associated to coronal manifestations of star-planet interaction. According to Poppenhaeger et al. (2010), any trends in the X-ray luminosity versus planet orbital parameters seem to be dominated by selection effects. In addition, Fares et al. (2010) reports no clear evidence of star-planet magnetospheric interactions in HD 189733. In this context, let us recall that, more recently from a comparison of the distribution of the rotation of stars with and without detected planets, Alves et al. (2010) have shown that the vsini distribution for these two families of stars is drawn from the same population distribution function.

Because the observational basis of stellar activity enhancement due to star-planet interaction has clearly not yet been established, in the present study we report a statistical analysis of stellar CaII chromospheric emission flux for a sample of 74 stars with planets, as described in the following sections. For a more solid analysis of the behavior of chromospheric and coronal activity of stars with planets, it is essential to conduct a comparative analysis of stars without detected planets. This is one of the major goals of the current study. In this context, we also analyzed the CaII chromospheric emission flux of a comparison sample of 26 stars without detected planets. The paper is organized as follows. In Sect. 2 we present the characteristics of the working samples, in Sect. 3 we describe our findings with a brief discussion, and finally we outline our conclusions in Sect. 4.

2. Stellar working sample and data

Different methods, such as radial velocity measurements (Butler et al. 1996; de Medeiros et al. 2009) and photometric light curves (Konacki et al. 2003, 2004), have been used to discover new planetary systems around other stars than the Sun. Aiming to make our sample homogeneous, we only use stars that had their planets discovered with radial velocity method.

The present stellar working sample consists of selected objects from the base of extrasolar planets maintained by Jean Schneider (Schneider 2010), updated on October 6, 2009, when there were 235 planets cataloged by the method in variation in the radial velocity (RV). We checked which of these stars presented Ca II emission flux detected in the catalog of Wright et al. (2004). After this selection, we filtered the complete stellar sample to obtain a subset of main-sequence stars that are within 20 pc in the solar neighborhood. Our final working sample consists of 74 stars with planets with spectral types in the F-G-K interval, as presented in Table 1.

The chromospheric activity indicator, , for all stars was computed from the Ca II H and K line-core emission index listed by Wright et al. (2004), following the procedure of converting emission index S to the flux at the stellar surface proposed by Noyes et al. (1984). To diagnose coronal activity of the stars forming the above-mentioned stellar sample, we used the coronal activity indicator, , that was computed using the X-ray fluxes listed by Kashyap et al. (2008) and Poppenhaeger et al. (2010). Readers are referred to these works for a discussion of the observational procedures, data reduction, and error analysis.

Stellar luminosities were determined as follows. First, apparent visual magnitudes (m_v) and trigonometric parallaxes, both taken from the HIPPARCOS catalogue (ESA 1997), were combined to yield the absolute visual magnitude (M_v). Bolometric correction (BC), computed from Flower (1996) calibration, was applied to obtaining the bolometric magnitude, which was finally converted into stellar luminosity. The effective temperature was computed using Flower (1996) (B − V) versus T_eff calibration. The stellar and planetary parameters, as well as chromospheric and coronal activity indicators for the entire sample of stars with planets, are listed in Table 1.

For comparative purposes, we took a sample of stars not known to have any planetary-mass companions. This new sample was taken from Schmitt et al. (1997), as done by Poppenhaeger et al. (2010), and is composed of 26 F-G-K type dwarf stars with all stars having Ca II emission flux measured by Wright et al. (2004) and X-ray fluxes detected by Schmitt (1997). The chromospheric and coronal activity indicators were calculated as before for the sample of stars with planets. However, we should be cautious about this sample, which derives from a list of stars that are surveyed for planets, but for which none have yet been found. This certainly does not mean that such stars have no planetary companions whatsoever. In fact, they might host planets with very low mass and/or a long orbital period that are more difficult to detect with radial velocity surveys. Stellar properties and chromospheric and coronal activity indicators for stars without detected planets are given in Table 2.

Fig. 1 Chromospheric activity indicator  as a function of the planetary semi-major axis log (1 / apl) for our sample of stars with planets.

3. Results and discussion

The aim of this work is to point out additional observational constraints that support or reject major effects of star-planet interaction in stellar activity, based on CaII chromospheric and X-ray emission fluxes. To this end, we dedicated most of our efforts to identifying qualitative trends between CaII and X-ray fluxes and planetary parameters. We chose and as indicators of chromospheric and coronal activity, respectively, because they are independent of stellar radius-induced effects. Indeed, it is expected that any planet-induced activity changes should therefore be more evident in (e.g.: Poppenhaeger et al. 2010). In the context of any planet-induced chromospheric activity, one should also expect more evident changes in  for such an aspect.

Following the same strategy as used by Poppenhaeger et al. (2010) in their analysis of the coronal X-ray emission behavior in stars with planets, we show the distribution of the chromospheric activity indicator  of stars with planets as a function of planetary distance log (1 / a_pl) (Fig. 1). A close comparison of the distribution of  versus log (1 / a_pl) with the distribution of the coronal activity indicator  versus log (1 / a_pl), illustrated in Fig. 5 of Poppenhaeger et al. (2010), shows that, at least qualitatively, the behavior of the chromospheric activity indicator appears to be comparable to that of the coronal activity indicator. The same aspect can be seen in Fig. 2 where we show the distribution of the chromospheric activity indicator  of stars with planets as a function of the product of the planetary mass with the reciprocal distance log (1 / a_pl).

Still following the strategy proposed by Poppenhaeger et al. (2010), we also carried out a statistical analysis of both  and  activity indicators of stars with close-in planets, namely planets with a_pl  ≤  0.20 AU, searching for correlations with planetary mass and semi-major axis. Table 3 shows the result of such an analysis, with the Spearman’s ρ rank correlation for various combinations of stellar quantities and planetary parameters, for two samples. The first is composed of 19 stars with X-ray lumisonity from Kashyap et al. (2008), and the second is composed of 13 stars with X-ray luminosities from Poppenhaeger et al. (2010). The first interesting aspect of this analysis is a possible anticorrelation between the chromospheric activity indicator and the planetary semi-major axis in stars hosting planets with a_pl  ≤  0.20 AU for both samples. When we consider the entire sample of stars, we observe an absence of any significant correlation between these two parameters, without any distinction of a_pl values.

Fig. 2 Chromospheric activity indicator  as a function of the product of the planetary mass with the reciprocal semi-major axis log (1 / apl  ×  Mpl) for our sample of stars with planets.

Fig. 3 The cumulative distributions of  for stars with apl  ≤  0.2 AU (solid curve) and apl  ≥  0.5 AU (dashed curve).

Two possible correlations are found in the present analysis: one of planetary mass log (M_pl) with  and the other of products of the planetary mass with the reciprocal distance log (1 / a_pl  ×  M_pl) with . Stars with giant and close-in planets exhibit higher  values than those with small outermost planets, a result corroborating the one obtained by Poppenhaeger et al. (2010) for the relationship between X-ray luminosities and planetary parameters. In this context, the Spearman coefficients from the correlation study of  with , also displayed in Table 3, show a significant correlation between chromospheric and coronal activity indicators for stars with close-in planets. This correlation indicates that the relations found by Poppenhaeger et al. (2010) between the coronal activity indicator and the planetary parameters must be the same as in our analysis of stellar chromospheric activity and the presence of planets around a star.

As an additional statistical test to verify that the present data sets, segregated by planetary distances, are signiÞcantly different from one another, we applied the Kolmogorov-Smirnov test (Press et al. 1992), which calculates the probability that two distributions are derived from the same parent distribution. According to the K-S test, the p value indicates the probability that both distributions come from the same origin. We conducted a K-S analysis, considering the  of stars with planets within a_pl  ≤  0.20 AU (20 stars) and stars with planets beyond a_pl  ≥  0.50 AU (43 stars), once again following the strategy adopted by Poppenhaeger et al. (2010). Figure 3 shows the cumulative functions for both  distributions. The errors were calculated considering the errors in  by applying random fluctuations with Gaussian distribution. The final values of D are the average of the standard deviation after the application of the fluctuations. The probability value of about 53% obtained on the K-S test is consistent with the two distributions drawn from the same population. In agreement with Poppenhaeger et al. (2010) in their analysis of coronal X-ray emission, the present result indicates that both distributions of chromospheric activity indicator  of stars with close-in and far-out planets are very similar.

Fig. 4 Chromospheric activity indicator  as a function of color index (B − V). Open circles denote stars with planets and solid circles stars without planets.

Fig. 5 The cumulative distributions of  for stars with planets (solid curve) and stars without planets (dashed curve). For the sample of stars hosting planets, we are only considering stars with planets with apl  ≤  0.20 AU.

3.1. Comparison of chromospheric activity indicator in stars with and without planets

In a search for systematic differences, we compared the behavior of chromospheric activity indicator  in stars with and without detected planets. In Fig. 4 we show  versus color index (B − V) for both samples of stars, with no discernible differences. Both samples tend to follow the well known behavior of , characterized by a decrease in chromospheric activity with increasing color index. In addition, in order to check for systematic differences between stars with and without detected planets, we compared the distributions of  for the two aforementioned samples, considering only stars hosting close-in planets, in other words, stars with planets with a_pl ≤ 0.20 AU. Figure 5 shows the cumulative functions for both distributions. A K-S test reveals that the probability of both samples being drawn from the same parent distribution is 27%, reinforcing the previous scenario, which shows no clear evidence of enhanced chromospheric activity associated to the presence of planetary companions.

4. Conclusions

We analyzed a sample of 74 stars with planets in the solar neighborhood and present chromospheric activity indicator , searching for possible effects of star-planet interaction on the stellar chromosphere. From these analyses, we found no signiÞcant correlations between the chromospheric activity indicator  and planetary parameters: semi-major axis and product of the planetary mass with the reciprocal semi-major axis. However, we found a possible correlation between the  and mass and between the product of planetary mass and reciprocal semi-major axis, indicating that massive close-in planets are often found around stars with an enhanced chromospheric activity indicator. Such a result supports the one obtained by Poppenhaeger et al. (2010) in their analysis of X-ray coronal luminosity in stars with planets. According to these authors, this dependence can be ascribed to selection effects, since the Doppler method for planet detection favors small and far-out planets around stars with low activity. Indeed, the present analysis shows a strong correlation between chromospheric and coronal activity indicators of stars with planets. Additionally, a statistical comparison between the  indicator in stars with and without detected planets shows no clear evidence of enhanced chromospheric activity associated with planetary companions. In summary, our analysis reveals no clear evidence of enhanced planet-induced activity in the chromosphere of the stars. In agreement with the conclusions drawn by Poppenhaeger et al. (2010) in their analysis of the X-ray luminosity behavior of stars with planets, any trends observed in the present study seem to be mostly the result of selection effects.

Acknowledgments

Research activities at the Stellar Board of Universidade Federal do Rio Grande do Norte are supported by continuous grants from the Brazilian agencies CNPq and FAPERN (J. R. De Medeiros and B. L. Canto Martins). M. L. Chagas and S. Alves acknowledge graduate fellowships from the CAPES Brazilian agency. L. P. de Souza Neto and I. C. Leão

acknowledge fellowships of the CNPq Brazilian agency. We warmly thank the anonymous referee for a careful reading and for suggestions that largely improved this paper.

References

All Tables

All Figures

	Fig. 1 Chromospheric activity indicator  as a function of the planetary semi-major axis log (1 / apl) for our sample of stars with planets.
In the text

	Fig. 2 Chromospheric activity indicator  as a function of the product of the planetary mass with the reciprocal semi-major axis log (1 / apl  ×  Mpl) for our sample of stars with planets.
In the text

	Fig. 3 The cumulative distributions of  for stars with apl  ≤  0.2 AU (solid curve) and apl  ≥  0.5 AU (dashed curve).
In the text

	Fig. 4 Chromospheric activity indicator  as a function of color index (B − V). Open circles denote stars with planets and solid circles stars without planets.
In the text

	Fig. 5 The cumulative distributions of  for stars with planets (solid curve) and stars without planets (dashed curve). For the sample of stars hosting planets, we are only considering stars with planets with apl  ≤  0.20 AU.
In the text