Protecting Planets from Their Stars

Vidotto: Exoplans et and the st ellar wind

he continuous flow of matter that escapes out of the solar gravitational well is Tknown as the solar wind. As the material flows out of the Sun, it is accelerated to between 400 and 800 km s–1. Through this continuous flow of material, the Sun loses more than one million tonnes every second. But this is just a tiny fraction, corresponding to 2 × 10 –14 of a solar mass that is lost every year. However, the solar wind does not only involve particles, it also carries the solar magnetic field lines. The magnetized solar wind permeates the entire solar system, having an effect on any body encountered on its way. The meeting of the solar wind and a solar system planet can result

in complex interactions that depend on the Downloaded from characteristics of both the local solar wind and the planet. Factors such as whether the planet is magnetized or whether it has an atmosphere can play different roles in this interaction. Planets that are weakly magnetized or not http://astrogeo.oxfordjournals.org/ magnetized, such as Mars, can have their atmospheres exposed to the impact of the solar wind. The solar wind may then remove the atmosphere of the planet by sputtering processes. Evidence suggests that Mars once had a thicker atmosphere, but most of it may have disappeared due to interaction with the young Sun’s wind, believed to have been more intense in the distant past. In the case of a magnetized planet such as the at St Andrews University Library on September 23, 2014 Earth, the planet’s field acts as a large obstacle for the solar wind: the flow of particles pro- duced by the Sun cannot penetrate all the way to the surface of our planet, but ends up being 1: Artist’s view of the interaction between the solar wind and the Earth’s protective deflected around the Earth’s magnetic field lines magnetic field. (SOHO [ESA & NASA]) (see figure 1). Because of the speed of the flow, the impact of the solar wind in the magnetosphere of the Earth produces a bow shock that surrounds the dayside magnetosphere of our planet (the side towards the Sun). Protecting planets The formation of bow shocks is not the only signature of the interaction between the magnetized solar wind and a magnetized planet. Mediated by magnetic reconnection events, from their stars energetic electrons are released in the system. Some of these electrons spiral along planetary magnetic field lines, giving rise to cyclotron Aline Vidotto explores how planets interact with the stellar wind and emission at radio wavelengths via a process how this evidence might help us find and characterize exoplanets. called cyclotron maser instability. Because this kind of interaction is regulated by magnetic hosting star and its surrounding planetary sys- wind–obstacle interaction (Zarka 2007). This reconnections, it can only exist between mag- tem. In fact, planetary interactions could even tight linear correlation is also known as the netized bodies. be more intense, as is certainly the case for stars radiometric Bode’s law. Because the pressure of that harbour more powerful stellar winds, or the solar wind is larger at smaller distances from Exoplanetary interactions with wind that have planets orbiting at close proximity the Sun, this correlation led scientists to propose Here, three (related) processes resulting from (<0.05 au, also called close-in planets). that, if a planet were orbiting at close distance the interaction between the solar wind and a In the case of planetary radio emission, to a star identical to the Sun with the same type planet were outlined, namely: the formation although the specific details of the interactions of stellar wind, this planet could produce radio of a bow shock, atmospheric erosion (in non- that generate these emissions are yet to be elu- emissions 103–105 times more intense than Jupi- or weakly magnetized planets) and auroral cidated, it has been recognized that the amount ter (Zarka 2007, Jardine and Cameron 2008). radio emission (in magnetized planets). There of energy released at radio wavelengths by the The detection of such intense auroral radio is no reason to doubt that similar interactions giant planets in the solar system correlates signatures from exoplanets would be a direct occur between the stellar wind of an exoplanet- tightly to the energy dissipated in the solar planet-detection method, as opposed to the

A&G • February 2013 • Vol. 54 1.25 Vidotto: Exoplans et and the st ellar wind Vidotto: Exoplans et and the st ellar wind

(a) (b) Downloaded from

2: The lowest energy state of the coronal magnetic field oft Boo as seen at different observing epochs. Colours denote the surface radial magnetic field and the solid line represents the neutral line (when the stellar magnetic field changes polarity). widely used indirect methods of radial veloc- etary magnetic field. Therefore, planets with migration (Lovelace et al. 2008). In the solar ity measurements or transit events. Moreover, magnetic field strengths of a few G, for exam- system, this process is negligible, but could the detection of exoplanetary radio emission ple, would emit at a frequency that could not be important in particular circumstances of would also demonstrate that the planet has a be observed from the ground either due to the stars that harbour strong magnetic fields and magnetic field. Earth’s ionospheric cut-off, or because it does dense winds (Vidotto et al. 2009, 2010b) or for However, despite many attempts, exoplan- not correspond to the operating frequencies of synchronizing stellar rotation with the orbital etary radio emission has not yet been detected. available instruments. In that regard, the low- motion of planets during the pre-main-sequence One of the reasons for the lack of success is operating frequency of LOFAR (currently under phase (Lanza 2010). thought to be the beamed nature of the elec- commission), jointly with its high sensitivity at Independently of the process involved, it is tron–cyclotron maser instability. Because the this low-frequency range, makes it an instru- worth noting that, in order to study the inter- emission occurs over a small solid angle, it ment that has great potential to detect radio action of the planets with the local environment would have to be directed towards the Earth emission from exoplanets. in which they are immersed, a key step is to to be detected. Poor instrumental sensitivity Note that different properties of star–planet understand the magnetic coronae and winds of would also explain the lack of detection of radio systems can also give rise to physical inter the host stars. emission from exoplanets. Another reason for actions that are absent or negligible in the solar the failure to find exoplanets this way may be system. For instance, it has been suggested that Stellar magnetic fields because of a frequency mismatch: the emission the winds of young Sun-like stars could change Although we seem to comprehend reasonably process is thought to occur at cyclotron frequen- the orbital angular momentum of planets by well the properties of the solar wind (especially cies, which depend on the intensity of the plan- the action of dragging forces, causing planetary because we are immersed in it), it is much more

1.26 A&G • February 2013 • Vol. 54 Vidotto: Exoplans et and the st ellar wind

3: Final configuration one order of magnitude smaller (about 2 years of the steady-state as opposed to 22 years for the solar magnetic magnetic field of cycle). The polarity reversals in t Boo seem to t Boo for June 2006. occur roughly every year, switching from a Colours denote the negative poloidal field near the visible pole in stellar wind velocity June 2006 (the intensity of the surface field is on the equatorial plane (xy plane). colour-coded in figure 2a) to a positive poloidal Stellar rotation axis field in June 2007 (figure 2b), and then back is along positive z. again to a negative polarity in July 2008 (figure 2d; Catala et al. 2007, Donati et al. 2008c, Fares et al. 2009). The nature of such a short magnetic cycle in t Boo remains an open question. Surface differential rotation is thought to play an important role in the solar cycle. The fact that t Boo pre-

sents a much higher level of surface differential Downloaded from rotation than that of the Sun may be responsible for its short observed cycle. In addition, t Boo hosts a close-in planet that, due to its proximity to the star, may have been able to synchronize, through tidal interactions, the rotation of the http://astrogeo.oxfordjournals.org/ shallow convective envelope of the host F-type difficult to probe and identify the properties of 2006a, 2008a; Morin et al. 2008, 2010) and star with the planetary orbital motion. This the winds of other stars. Even if we concentrate high-mass stars (Donati et al. 2006b). Donati presumed synchronization may have enhanced on a subsample of stars with similar masses to and Landstreet (2009) present a recent overview the shear at the tachocline, which may have our Sun, differences in coronal temperatures, of the survey. influenced the magnetic cycle of the star (Fares stellar rotation rates, magnetic field intensities, Although some objects host fields that can et al. 2009). etc, imply different stellar wind properties. resemble the large-scale solar field, there are Because the stellar winds of cool stars are Because of that, theoretical and numerical mod- also fascinating differences. For example, solar- magnetic by nature, variations of the stellar els are essential in the progress of our under- type stars that rotate about twice as fast as our magnetic field during the cycle directly influ- standing of winds of other stars. Sun show a substantial toroidal component of ence the outflowing wind. Therefore, the rapid at St Andrews University Library on September 23, 2014 One factor of particular relevance for stellar magnetic field, a component that is almost non- variation of the large-scale magnetic field of winds is the geometry of the stellar magnetic existent in the large-scale surface solar magnetic t Boo implies that the environment surround- field. The Sun can again be a useful illustration. field (Petit et al. 2008). The magnetic topology ing the close-in planet should be varying quite During periods of minimum activity, the solar of low-mass (<0.5 M ) very active stars seem to rapidly too. In order to characterize such an ⊙ field topology resembles an aligned dipole. The be dictated by interior structure changes: while environment and the related interactions with fast solar wind emerges from the polar regions partly convective stars possess a weak non- the exoplanet, Vidotto et al. 2012 performed where magnetic field lines are open (coronal axisymmetric field with a significant toroidal three-dimensional magnetohydrodynamics holes) and the slow solar wind emerges above component, fully convective ones exhibit strong simulations of the host star’s wind, taking into the low-latitude active regions (latitudes of up poloidal axisymmetric dipole-like topologies account the observed surface magnetic maps to 30–35° around the equator). In contrast, dur- (Morin et al. 2008, Donati et al. 2008a). of t Boo. ing periods of maximum activity, the topology All this recent insight into the magnetic topol- To incorporate these surface maps in the of the field becomes more complicated, affect- ogy of different stars can now be incorporated simulations, one needs a way to extrapolate the ing the solar wind: by the time the Sun is at in stellar wind models, which is a key step to surface magnetic field to the stellar corona. A maximum activity, the poles also emit the slow make them more realistic. Ultimately, it is the common method for doing this is by using the solar wind. characterization of the stellar wind that will assumption that the magnetic field is at its low- Although the richness of details of the mag- constrain the local environment surrounding est energy state (Altschuler and Newkirk 1969, netic field configuration is only known for our exoplanets and, consequently, their interactions Jardine et al. 1999). Figure 2 shows the derived closest star, modern techniques have made it with the host-star plasma. extrapolated coronal magnetic field lines for possible to reconstruct the large-scale surface each of the observed epochs for which surface magnetic fields of other stars. The Zeeman– The t Boo system maps have been observationally reconstructed. Doppler Imaging (ZDI) technique is a tomo- A particular exoplanetary system that has been The lowest-energy magnetic field configuration graphic imaging technique (e.g. Donati and placed under scrutiny is the t Boo system. The (the potential field) is used in simulations of stel- Brown 1997) that allows the reconstruction of host-star, t Boo (spectral type F7V), is a remark- lar winds as an initial condition only. In fact, the the large-scale magnetic field (intensity and ori- able object, not only because it hosts a giant interaction between the stellar wind particles entation) at the surface of the star from a series planet orbiting very close to it (0.046 au), but and the magnetic field lines removes the coronal of circular polarization spectra. This method also because it is the only star other than the field from its lowest energy state. This is illus- has now been used to investigate the magnetic Sun for which a full magnetic cycle has been trated in figure 3, which shows the final steady- topology of planet-hosting stars (Fares et al. reported in the literature. So far, two polar- state configuration of the magnetic field lines 2009, 2010, 2012), solar-type stars (Petit et al. ity reversals have been detected (Donati et al. of t Boo at the observed epoch of June 2006. 2008, 2009), young solar-type stars (Marsden 2008c, Fares et al. 2009), suggesting that the The stress in the magnetic field lines is verified et al. 2006; Donati et al. 2008b, 2010; Hus- star undergoes magnetic cycles similar to the by the presence of twisted magnetic field lines sain et al. 2007), low-mass stars (Donati et al. Sun, but with a complete period that is about around the rotation axis (z-axis). Figure 3 shows

A&G • February 2013 • Vol. 54 1.27 Vidotto: Exoplans et and the st ellar wind Vidotto: Exoplans et and the st ellar wind

(a) optical transit

5: Comparison between the near-UV and the optical transits of WASP-12b. Filled circles represent the spectroscopic narrow-band near-UV transit of WASP-12b observed using HST/COS by Fossati (b) near-UV et al. (2010), obtained at five consecutive orbits of the satellite. The solid line shows the broadband transit optical transit. Note that the near-UV transit starts before the optical ephemeris, but finishes at Downloaded from about the same phase, illustrating that there is an asymmetric distribution of material around the planet. Monte Carlo radiation transfer simulations (dashed line, Llama et al. 2011, see also figure 6) of the near-UV transit of WASP-12b supports the hypothesis that a bow shock surrounding the planet can explain both the early ingress in the near-UV data, as well as the excess absorption compared to the optical transit. (Figure adapted from Fossati et al. 2010b and Llama et al. 2011) http://astrogeo.oxfordjournals.org/

of t Boo, showing that this star probably has a radio detection with ground-based observations denser wind than that of the Sun, with mass-loss from planets with magnetic field intensities≲ 4 G 4: Sketches of the light curves obtained rates that are two orders of magnitude larger should not be possible (Vidotto et al. 2012). · · · through observations in the (top) optical than the solar value M (M 135 M ). and (bottom) near-UV, where the bow shock ⊙ ≈ ⊙ Finding magnetized planets surrounding the planet’s magnetosphere is Planetary radio emission in t Boo b also able to absorb stellar radiation. (From A magnetic field has not yet been observed on Vidotto et al. 2011b) Such a denser wind implies that the energy dis- extrasolar planets. Detection of radio emission sipated in the stellar wind–planet interaction would not only constrain local characteristics at St Andrews University Library on September 23, 2014 can be significantly higher than the values of the stellar wind, but would also demonstrate derived in the solar system. Since, among the that exoplanets are magnetized. Fortunately, the wind velocity in the equatorial plane of the solar system planets, the wind-dissipated energy there may be other ways to probe exoplanetary star (xy-plane). Vidotto et al. (2012) showed follows linearly the planetary radio emission, it magnetic fields, in particular for transiting sys- that both the coronal magnetic field lines and is expected that the planet t Boo b should have tems, through signatures of bow shocks during the stellar wind velocity profile vary through high radio emission, provided that the planet is transit observations. the stellar cycle. magnetized and the proper viewing conditions The formation of bow shocks is one signature One important unknown parameter that is are achieved. Combined with the close prox- of the interaction of the host-star corona/wind needed in all models of stellar winds is the den- imity of the system (~16 pc), the t Boo system with an orbiting planet. What determines the sity at the base of the corona. Unfortunately, is one of the strongest candidates for verifying orientation of the shock is the net velocity of the without a good estimate of this value, the stellar planetary radio emission theories. particles meeting the planet’s magnetosphere. In mass-loss rates, one of the fundamental prop- Using the detailed stellar wind model devel- the case of the Earth, the solar wind has essen- erties of the stellar wind, cannot be inferred. oped for its host-star, Vidotto et al. (2012) esti- tially only a radial component, which is much One way to constrain coronal base densities, mated radio emission from t Boo b, exploring larger than the orbital velocity of the Earth. and therefore mass-loss rates, is to perform a different values for the assumed planetary mag- Because of that, the shock forms facing the Sun direct comparison between derived mass-loss netic field. They showed that, for a planet with (a dayside shock). However, for close-in exo- rates from the simulations and those deter- a magnetic field similar to Jupiter’s ≃( 14 G), the planets that possess high orbital velocities and mined observationally. However, mass-loss radio flux is estimated to be≃ 0.5–1 mJy, occur- are frequently in regions where the host star’s rate measurements exist only for the Sun, so ring at an emission frequency of ≃34 MHz. wind velocity is comparatively much smaller, a indirect approaches are often employed (see, for Although small, this emission frequency lies shock may develop ahead of the planet (called example, Wood 2004). in the observable range of current instruments, the ahead shock). In general, we expect that Vidotto et al. 2012 employed an indirect such as LOFAR. To observe such a small flux, shocks are formed at intermediate angles. methodology to constrain the mass-loss rate an instrument with a sensitivity lying on a mJy Due to their high orbital velocities, close-in of t Boo using emission measure (EM) values level is required. planets offer the best conditions for transit derived from X-ray spectra. The EM probes The same estimate was made considering that observations of bow shocks. If the compressed both the electron and ion densities inside the planet has a magnetic field similar to the shocked material is able to absorb stellar radia- regions of closed magnetic field lines in the stel- Earth (≃1 G). Although the radio flux is not tion, then the signature of bow shocks may be lar corona. Therefore, by tuning the coronal significantly different to the previous case, the observed through both a deeper transit and an densities in the simulation, it is possible to find emission frequency (≃2 MHz) falls at a range early ingress of some spectral lines with respect a best match of the predicted EM values from below the ionospheric cut-off, preventing any to the broadband optical ingress (Vidotto et al. the simulations to the observed ones for t Boo. possible detection from the ground. In fact, 2012). The sketches shown in figure 4 illustrate · This approach constrained the mass-loss rate M because of the ionospheric cutoff at ~10 MHz, this idea.

1.28 A&G • February 2013 • Vol. 54 Vidotto: Exoplans et and the st ellar wind

6: Sequence of images from Monte Carlo radiation transfer simulations of the near-UV transit of WASP-12b. (Adapted from Llama et al. 2011)

This suggestion proposed by Vidotto (2010a) al. (2011). Figure 6 shows a sequence of images We assume the stand-off distance to trace the was motivated by transit observations of the depicting the transit in the near-UV, where we extent of the planetary magnetosphere. At the close-in giant planet WASP-12b. Based on note that the shocked material can be very tenu- magnetopause, pressure balance between the Hubble Space Telescope/Cosmic Origins Spec- ous, but as long as there is enough material inte- coronal total pressure and the planet total pres- trograph (HST/COS) observations using nar- grated along the line of sight, it can cause the sure requires that

row-band near-UV spectroscopy, Fossati et al. absorption level observed in the near-UV data. Downloaded from [B (a)]2 [B (r )]2 2 ______c ______p M (2010b) showed that the transit lightcurve of ρc Δ u + + pc = + pp (4) Planetary magnetic fields: a new 8π 8π WASP-12b (filled circles in figure 5) presents detection method? both an early ingress when compared to its where rc, pc and Bc(a) are the local coronal optical transit (solid line in figure 5), as well as An interesting outcome of the observations of mass density, thermal pressure, and magnetic excess absorption during the transit, indicat- bow shocks around exoplanets is that it per- field intensity, and pp and Bp(rM) are the planet http://astrogeo.oxfordjournals.org/ ing the presence of an asymmetric distribution mits one to infer the magnetic field intensity of thermal pressure and magnetic field intensity of material surrounding the planet. This result the transiting planet. By measuring the phases at rM. In the case of a magnetized planet with was reiterated by Haswell et al. (2012) in a more at which the near-UV and the optical transits a magnetosphere of a few planetary radii, the recent HST/COS set of observations. begin, one can derive the stand-off distance planet total pressure is usually dominated by WASP-12b orbits a late-F main-sequence from the shock to the centre of the planet. In the contribution from the planetary magnetic star which has mass M = 1.35M and radius the geometrical consideration below, we assume pressure (i.e. p ~ 0). ⁎ ⊙ p R = 1.57 R , at an extremely small orbital that the planet is fully superimposed on the disc Vidotto et al. (2010a) showed that, because ⁎ ⊙ radius of a = 0.023 au, which corresponds to of the central star, which is a good approxima- WASP-12b is close to the star, the kinetic term a distance of only 3.15 R (Hebb et al. 2009). tion for the cases of, for example, small plan- of the coronal plasma may be neglected in equa- ⁎ Due to its close proximity to the star, the flux ets and transits with small impact parameters tion 4. They also neglected the coronal thermal at St Andrews University Library on September 23, 2014 of coronal particles impacting on the planet b. Consider the sketches presented in figure 4, pressure (justified by the low coronal densities), should come mainly from the azimuthal direc- where dop and duv are, respectively, the sky- so that equation 4 reduces to Bc(a) ≃ Bp(rM). Fur- tion, as the planet moves at a Keplerian orbital projected distances that the planet (optical) and ther assuming that stellar and planetary mag- 1/2 –1 velocity of uK = (GM ⁄a) ~ 230 km s around the system planet + magnetosphere (near-UV) netic fields are dipolar, we have ⁎ the star. Therefore, stellar coronal material travel from the beginning of the transit until the R /a 3 ______∗ Bp = B (5) is compressed ahead of the planetary orbital middle of the optical transit ∗ ( R /r ) p M motion, possibly forming a bow shock ahead d = (R2 – b2)1/2 + R (1) op ⁎ p of the planet. Vidotto et al. (2010a) suggest that and where B and B are the magnetic field intensi- ∗ p this shocked material is able to absorb enough d = (R2 – b2)1/2 + r (2) ties at the stellar and planetary surfaces, respec- uv ⁎ M stellar radiation, causing the asymmetric light- where b is the impact parameter derived from tively. Here as well, the planetary magnetic field curve observed in the near-UV (see figures 4 and transit observations, Rp is the planetary radius, can be related to observed quantities. 5), where the presence of compressed material and rM is the distance from the shock nose to Therefore, by determining the phase at which ahead of the planetary orbit causes the early the centre of the planet. The start of the optical the near-UV transit begins, one can derive the ingress, while the lack of compressed material transit occurs at phase f1 ≡ f op (point 1 in figure stand-off distance (equation 3) and then esti- behind the planetary orbit causes simultaneous 4a), while the near-UV transit starts at an ear- mate the intensity of the magnetic field of the egresses both in the near-UV transit and the lier phase f1ʹ ≡ f uv (point 1ʹ in figure 4b). Taking planet (equation 5), provided that the stellar optical one. the optical mid-transit phase at f = fm ≡ 1, we magnetic field is known. For WASP-12, we use To verify this idea, Llama et al. (2011) per- note that d is proportional to (1 – f ), while the upper limit of B < 10 G (Fossati et al. 2010a) op op ∗ formed Monte Carlo radiation transfer simu- duv is proportional to (1 – fuv). Using equations and the stand-off distance obtained from the lations of the near-UV transit of WASP-12b. 1 and 2 we find (Vidotto ______et al. 2010a, 2011d) near-UV transit observation rM 2= 4. Rp (Lai et They confirmed that the presence of a bow al. 2010) and we were able to predict an upper ___rM ______(1 – fuv) ___ R 2 ____bR 2 shock indeed breaks the symmetry of the transit = ⁎ – ⁎ + 1 limit for WASP-12b’s planetary magnetic field Rp (1 – fop) [  ( R p ) ( R p ) ] lightcurve, supporting the hypothesis proposed ______(3) of Bp < 24 G. by Vidotto et al. (2010a). In their simulations, R 2 bR 2 – ___ ⁎ – ____⁎ Searching for magnetic fields in the geometry of the shock was varied in order  ( R p ) ( R p ) to fit the near-UV data. They found that no Note that, by measuring the phase at which other exoplanets fine-tuning is required. There are several shock the transit starts in the near-UV (fuv), one can In theory, the suggestion that through tran- geometries that could still provide a good fit relate the normalized stand-off distance (rM/Rp) sit observations one can probe the planetary to the HST/COS data; the current data is not to observed quantities such as the planet–star magnetic field is quite straightforward – all it yet adequate to fully constrain the bow shock radius ratio (R /R ), the impact parameter (b, requires is a measurement of the transit ingress p ⁎ geometry. The dashed line in figure 5 shows one in units of R ), and the phase of optical first phase in the near-UV. In practice, however, ⁎ of the several possible fits found by Llama et contact (fop). acquisition of near-UV transit data requires the

A&G • February 2013 • Vol. 54 1.29 Vidotto: Exoplans et and the st ellar wind use of space-borne facilities, making follow-up the planet’s magnetic field through bow shock frequency of flares and CMEs may be much and new target detections rather difficult. observations), which would otherwise remain higher than for the present-day Sun. In order to optimize target selection, Vidotto unknown. We live in exciting times for both theoretical et al. (2011a) presented a classification of the Stellar wind–planet interactions are two-way and observational investigations of the inter known transiting systems according to their roads. On one side, the stellar wind plays a deci- action between host stars and their exoplan- potential for producing shocks that could cause sive role in the characterization of the magnetic ets. Several aspects remain unsolved, such as observable light curve asymmetries. The main environment around the planet. On the other, how different stellar coronal environments assumption considered was that, once the con- planetary observations of wind-related pro- surrounding a planet can affect habitability ditions for shock formation are met, planetary cesses (transit asymmetries, planetary radio (Vidotto et al. 2013 submitted). The findings shocks absorb in certain near-UV lines, in a emission) can help constrain the local proper- reviewed here highlight the importance of similar way to WASP-12b. In addition, for it to ties of the stellar wind. understanding and characterizing the magnetic be detected, the shock must compress the local Recent observations of transiting systems environment of exoplanets, and provide guid- plasma to a density sufficiently high to cause an reveal that transit asymmetries are common ance for future work. ● observable increase in optical depth. and occur at various wavelengths. This last hypothesis requires Transit asymmetries are believed Aline Vidotto is an RAS Research Fellow, SUPA, How

knowledge of the local ambient to be caused by an asymmetric School of Physics and Astronomy, University of St Downloaded from medium that surrounds the do‘‘ different distribution of material sur- Andrews, UK. planet. stellar coronal rounding the planet. Rappa- By adopting simplified environments around port (2012) reported a transit References hypotheses, namely that up asymmetry in KIC 12557548, Altschuler M D and Newkirk G 1969 Solar Phys. 9 131. planets affect Brogi M et al. 2012 Astron. & Astrophys. 545 L5. to the planetary orbit the stel- which they have interpreted as habitability? Catala C et al. 2007 Mon. Not. Roy. Ast. Soc. 374 L42. http://astrogeo.oxfordjournals.org/ lar corona can be treated as being caused by a trailing dust Donati J-F and Brown S F 1997 Astron. & Astrophys. in hydrostatic equilibrium and ’’ cloud. Further data on the sys- 326 1135. isothermal, Vidotto et al. (2011a) tem found variations in the transit Donati J-F and Landstreet J D 2009 Ann. Rev. Astron. and Astrophys. 47 333. predicted the characteristics of the ambi- asymmetry, leading Brogi et al. (2012) Donati J-F et al. 2006a Science 311 633. ent medium that surrounds the planet for a sam- to conclude that the dust cloud might have Donati J-F et al. 2006b Mon. Not. Roy. Ast. Soc. 370 ple of 125 transiting systems, and discussed disappeared. Temporal variations of transit 629. whether such characteristics present favourable asymmetries were also observed in the brighter Donati J-F et al. 2008a Mon. Not. Roy. Ast. Soc. 390 conditions for the presence and detection of a system HD 189733 (Lecavelier des Etangs et 545. Donati J-F et al. 2008b Mon. Not. Roy. Ast. Soc. 386 bow shock. Excluding systems that are quite far al. 2012). The observed variation was attrib- 1234. at St Andrews University Library on September 23, 2014 ( ≳400 pc), the planets that were top ranked are: uted to modifications of the properties of the Donati J-F et al. 2008c Mon. Not. Roy. Ast. Soc. 385 WASP-19b, WASP-4b, WASP-18b, CoRoT-7b, stellar wind impacting on the planet, probably 1179. HAT-P-7b, CoRoT-1b, TrES-3 and WASP-5b. caused by a stellar flare detected ~8 h prior to Donati J-F et al. 2010 Mon. Not. Roy. Ast. Soc. 409 1347. Fares R et al. 2009 Mon. Not. Roy. Ast. Soc. 398 1383. the transit. Concluding remarks Fares R et al. 2010 Mon. Not. Roy. Ast. Soc. 406 409. It is interesting to note that, in the frame- Fares R et al. 2012 Mon. Not. Roy. Ast. Soc. 423 1006. In this article, I discussed how the stellar work of near-UV transit, Vidotto et al. (2011b) Fossati L et al. 2010a Ap. J. 720 872. corona/wind can affect surrounding planets. investigated time-dependent effects on the Fossati L et al. 2010b Ap. J. Lett. 714 L222. Because the winds of cool stars are incredibly asymmetry of planetary transit light curves. Haswell C A et al. 2012 Ap. J. 760 79. Hebb L et al. 2009 Ap. J. 693 1920. tenuous, up to now there have not been any For example, stellar coronae are not axisym- Hussain G A J et al. 2007 Mon. Not. Roy. Ast. Soc. 377 direct measurements of their properties (except metric. Thus, along its orbit, the planet interacts 1488. for the Sun itself). To determine the fundamen- with stellar material of different characteris- Jardine M and Cameron A C 2008 Astron. & Astrophys. tal properties of such winds (such as mass-loss tics. In this case, differences in the surround- 490 843. Jardine M et al. 1999 Mon. Not. Roy. Ast. Soc. 305 L35. rates, terminal velocities), one has to rely on the ing material will cause variation in the size of Lai D et al. 2010 Ap. J. 721 923. very few indirect methods that probe cool stellar the planet’s magnetosphere, and therefore on Lanza L 2010 Astron. & Astrophys. 512 A77. winds. On the theoretical front, magnetohydro- the stand-off distance observed during transit Lecavelier des Etangs A et al. 2012 Astron. & Astro- dynamics numerical simulations can definitely observations. Because the stellar rotation period phys. 543 L4. help ascertain the aspects of winds. Aiming at in general differs from the orbital period, a Llama J et al. 2011 Mon. Not. Roy. Ast. Soc. 416 L41. Lovelace R V E et al. 2008 Mon. Not. Roy. Ast. Soc. 389 characterizing such winds more realistically, series of transit observations can probe different 1233. we have implemented (for example, Vidotto et stellar material. This is the case, for instance, if Marsden S C et al. 2006 Mon. Not. Roy. Ast. Soc. 370 468. al. 2011c, 2012) in our numerical simulations the star has an oblique magnetosphere and the Morin J et al. 2008 Mon. Not. Roy. Ast. Soc. 390 567. observationally derived surface magnetic field planetary orbit takes the planet through regions Morin J et al. 2010 Mon. Not. Roy. Ast. Soc. 407 2269. Petit P et al. 2008 Mon. Not. Roy. Ast. Soc. 388 80. maps, which show the diverse magnetic field of confined and expanding stellar material. Petit P et al. 2009 Astron. & Astrophys 508 L9. strengths and topologies that these stars host. Furthermore, time-dependent intrinsic varia- Rappaport S et al. 2012 Ap. J. 752 1. Ultimately, quantifying and understanding tions of the stellar magnetism, such as those due Vidotto A A et al. 2009 Ap. J. 703 1734. winds of cool stars leads to the characteriza- to coronal mass ejections (CMEs), flares and Vidotto A A et al. 2010a Ap. J. 722 L168. tion of the environment surrounding exo- stellar magnetic cycles, can also impress variable Vidotto A A et al. 2010b Ap. J. 720 1262. Vidotto A A et al. 2011a Mon. Not. Roy. Ast. Soc. 411 L46. planets. Although these environments may be observable signatures in different transits of the Vidotto A A et al. 2011b Mon. Not. Roy. Ast. Soc. 414 1573. potentially dangerous for a planet’s atmosphere same system. Although the impact of a CME on Vidotto A A et al. 2011c Mon. Not. Roy. Ast. Soc. 412 351. (especially for close-in planets), the interaction a planet increases the local density surrounding Vidotto A A et al. 2011d Ast. Nachrichten 332 1055. between planets and the host star’s coronal the planet, its effect is transient and may not be Vidotto A A et al. 2012 Mon. Not. Roy. Ast. Soc 423 3285. Vidotto A A et al. 2013 Mon. Not. Roy. Ast. Soc submit- winds can provide other avenues for planet captured in transit observations, except maybe ted. detection (such as radio emission) and maybe for the case of planets orbiting young, magneti- Wood B E 2004 Living Reviews in Solar Physics 1 2. even assessment of planetary properties (such as cally active stars for which the magnitude and Zarka P 2007 Planet. and Space Sci. 55 598.

1.30 A&G • February 2013 • Vol. 54