UV ENVIRONMENT OF EXOPLANETS

Jeffrey Linsky, Kevin France, and Rachel Bushinsky

University of ColoradoBoulder Colorado

ExoClimes 2012Aspen Colorado15-20 Jan 2012

Simulated spectrum of GJ 436 (M2.5 V with a 0.073MJ exoplanet). Note emission lines and

photo-absorption cross-sections of H2O and CH4

COS spectrum of 2M1207 a young M8 brown dwarf with a CS disk (H2) and accretion (C IV) but no Si III-IV or Mg II

emission lines (France et al. ApJ 715, 596 (2010))

HST/STIS spectrum of AU Mic (M1 V) (Pagano et al. ApJ 532, 497 (2000). What line emits most of the UV flux?

UV spectra of solar-mass stars (spectral types G0-G2 dwarfs)

“Sun in time Project”. Fluxes at 1 AU from solar type stars as a function of age (Ribas et al. ApJ 622, 680 (2005)). Flux is very sensitive to age

(activity) at shorter wavelengths.

Spectral flux (for solar type stars) decreases with age most rapidly at the shortest wavelengths. Is

the same true for M dwarfs?

Reconstructing Lyman-α Line Profiles Using Information on the Local ISM (Wood et al.

ApJS 159, 118 (2005).

Left: ξ Boo A (G8 V). Right: AD Leo (M3.5 V) and AU Mic (M0 V).

Observed and reconstructed Lyman-α line profiles (Wood et al. ApJS 159, 118 (2005)

Limitations on the Lyman-α line reconstruction technique

• Must be able to resolve the D I Lyman-α line to reconstruct the H I line.

• Effective horizon at log N(HI)=18.7 (15-100 pc).

• Should also have high resolution spectra of metal lines (e.g., MgII or FeII) to determine whether many velocity components in the line of sight.

Curves for increasing HI column density are for log N(HI) = 17.5, 17.8, 18.1, 18.4, 18.7. Solid and dashed lines are for T=10,000 and 5,000 K. (Wood et al. 2005)

Observed and reconstructed Lyman-α flux for GJ667C (M1.5V)

2 exoplanets at 0.05 and 0.12 AU.

Solar Spectral Irradiance at Earth (moderate activity)

Band f(erg/cm2/s) Range FactorX-ray (ROSAT) 0.41 1010-30Å 0.013 530-100Å 0.42 3100-300Å 1.2 3300-911Å 1.4 2911-1027Å 0.23 1.5Lyman-α 5.1 1.51200-1600Å minus Ly-α

2.44 1.4

UV and X-ray flux emitted by AU Mic (M1 V)

Spectral Region Obs. flux (10-12

erg/cm2/s)Reference

Lyman-α (1216Å)

10.3 (Reconstructed)

Wood et al. (2005)

1175-1700Å except Lyman-α

1.6 (HST/STIS)

Pagano et al. (2000)

933-1176Å 0.75 (FUSE)

Redfield et al. (2002)

X-ray 36.1 (ROSAT)

Wood et al. (2005)

HST observations of M dwarfs with exoplanets (programs 12034, 12035, and 12464)

Star + exoplanets d(pc) COS/FUV STIS/Lyα+

GJ436 M3.5V+1 10.14 2012 plan 2012 plan

GJ581 M5.0V+4 6.21 07/2011 ? 07/2011 ?

GJ667C M1.5V+2 6.95 09/2011

GJ832 M1.5V+1 4.95 2012 plan 07/2011

GJ876 M5.0V+4 4.69 01/2012 11/2011

GJ1214 M4.5V+1 12.95 2012 plan 2012 plan

Maps of n(HI) computed in hydrodynamic models of stellar atmospheres (Wood et al 2005). Scales

are in AU. Solid line indicates LOS to the observer.

61 Vir has 3 exoplanets located inside of its hydrogen wall.

Interstellar, astrospheric, and heliospheric Lyman-α absorption

• Interstellar gas flow from above.

• Line of sight passes through astrosphere, ISM, and then heliosphere

• Astrospheric absorption is blue compared to ISM because H slows down in the astropause as seen by the star.

• Heliospheric absorption is red compared to ISM because H slows down in the heliopause.

• Astrospheric absorption proportional to stellar mass loss rate.

Measuring stellar mass loss rates from the astrospheric absorption feature on the blue

side of interstellar Lyman-α absorption

Mass loss rate vs activity (X-ray surface flux) and stellar age

• Solid dots indicate dwarf stars (Prox Cen and EV Lac are only M dwarfs).

• Near Fx=106 there is a break suggesting a change in magnetic field structure.

• More measurements are needed (HST GO-19 proposal).

• Ṁ ~ FX1.34±0.18 ~ t2.33±0.55

Conclusions• Stellar UV+FUV+EUV flux incident on an exoplanet’s

atmosphere photodissociates important molecules like H2O and CH4 and photoionizes important atoms like H and O.

• Reconstructed Lyman-α dominates the UV flux often by a very large factor.

• Lyman-α and UV stellar fluxes for G stars follow a regular dependence on B-V and rotation period (a proxy for stellar activity).

• Lyman-α and UV fluxes of M stars cannot be easily predicted and must be observed. We have HST observing programs for M stars with exoplanets.

• We have found a scaling law for stellar mass loss rates which important for computing charge exchange rates. We have an HST/STIS observing program to extend this to M dwarfs.

We are now writing HST proposals for new observations of

M dwarfs with exoplanets

Those interested in providing data for new targets or in theoretical analysis

(e.g., atmospheric chemistry calculations) are welcome to join the

proposing team. Please see me.