Andrew Collier Cameron, Moira Jardine, John Barnes, Sandra Jeffers, Duncan Mackay, Kenny Wood (University of St Andrews) Jean-François Donati (Obs. Midi-Pyrenees, Toulouse). Meir Semel (Obs. de Paris, Meudon) Magnetic activity on rapidly rotating stars I: Surface flux distributions Activity proxiesActivity proxies Surface coverage of active regionsSurface coverage of active regions Polar spotsPolar spots Diffusion and advection of surface magnetic fieldsDiffusion and advection of surface magnetic fields Filling factors and flux emergence rates.Filling factors and flux emergence rates.

Overview Why are rapidly rotating stars useful?Why are rapidly rotating stars useful? We can:We can: –map their surfaces! –determine the latitude distribution of active regions –estimate the flux emergence rate and spot lifetime –map the magnetic polarity distribution in the network What does this all tell us about dynamos?What does this all tell us about dynamos? –Spectral-type dependence of surface flux distribution –Spectral-type dependence of differential rotation –Cyclic behaviour: spot coverage, differential rotation –Meridional circulation?

Magnetic activity proxies Broad-band optical modulationBroad-band optical modulation –=> dark starspots Collier Cameron et al 1999 Kürster et al 1997

Magnetic activity proxies Emission cores in strong UV/optical linesEmission cores in strong UV/optical lines –=> chromospheres Linsky et al 1979Sun in Ca II nm filter

Magnetic activity proxies Emission cores in strong UV/optical linesEmission cores in strong UV/optical lines –  chromospheres –  rotation periods –  activity cycles Vaughan et al 1981

Magnetic activity proxies Emission cores in strong UV/optical linesEmission cores in strong UV/optical lines –  chromospheres –  rotation periods –  activity cycles –  differential rotation? –Secular changes in Ω  range of surface  range of surface rotation rates. rotation rates. – Period-DR relation: –BUT: No reliable latitude information information Donahue, Saar & Baliunas 1996

Magnetic activity proxies Emission cores in strong UV/optical linesEmission cores in strong UV/optical lines –  chromospheres –  rotation periods –  activity cycles –  dynamos? Noyes et al 1984

Magnetic activity proxies Soft X-ray emissionSoft X-ray emission –  magnetically confined coronal plasma XMM spectrum and light-curve of star in IC2391 (Marino et al 2003)

Magnetic activity proxies Soft X-ray emissionSoft X-ray emission –  “Saturation” … Vilhu 1984

Magnetic activity proxies Soft X-ray emissionSoft X-ray emission –  “Saturation” … –  and “super-saturation” Prosser et al. 1996, alpha Persei cluster Stauffer et al. 1997

Magnetic activity proxies Decrease in rotation with ageDecrease in rotation with age –Ultra-fast rotators found in young clusters only –Earlier spectral types spin faster after ~0.3 Gyr –=> Hot magnetically channelled winds dΩ/dt ~ -Ω 3 Barnes, S. 2001

Convection and rotation F, G, K, M spectral typesF, G, K, M spectral types –  outer convective zones Activity indicators increase with rotationActivity indicators increase with rotation –  Rotation drives activity Evidence of differential rotation: can we map it?Evidence of differential rotation: can we map it? Spindown rates depend on spectral typeSpindown rates depend on spectral type –  Convection zone depth is important Do young stars really have up to 50% starspot occupancy?Do young stars really have up to 50% starspot occupancy? For the fastest rotators L x decreases with Ω !For the fastest rotators L x decreases with Ω !

Evidence for dense spot coverage TiO bands occur in spots only.TiO bands occur in spots only. O’Neal, Neff & Saar 1996

Evidence for dense spot coverage TiO bands occur in spots only.TiO bands occur in spots only. 7055Å/8860Å band ratio gives spot temperature.7055Å/8860Å band ratio gives spot temperature. O’Neal, Neff & Saar 1998

Evidence for dense spot coverage TiO bands occur in spots only.TiO bands occur in spots only. 7055Å/8860Å band ratio gives spot temperature.7055Å/8860Å band ratio gives spot temperature. Band strength gives spot covering fraction.Band strength gives spot covering fraction. O’Neal, Neff & Saar 1998 Composite model spectrum Spot filling factor Continuum brightness ratio Normalised spot spectrum Normalised photospheric spectrum

Evidence for dense spot coverage TiO bands occur in spots only.TiO bands occur in spots only. 7055Å/8860Å band ratio gives spot temperature.7055Å/8860Å band ratio gives spot temperature. Band strength gives spot covering fraction.Band strength gives spot covering fraction. Active stars have filling factors f s ~20% to 40%Active stars have filling factors f s ~20% to 40% O’Neal, Neff & Saar 1998

Measuring spot coverage with HST Eclipsing binary SV CamEclipsing binary SV Cam G0V + K5VG0V + K5V Edge-on orbitEdge-on orbit K5V transits primaryK5V transits primary Light-curve analysis  radiiLight-curve analysis  radii Measure missing-flux spectrum at mid eclipseMeasure missing-flux spectrum at mid eclipse Use HIPPARCOS parallax to get solid angle  surface brightnessUse HIPPARCOS parallax to get solid angle  surface brightness Jeffers et al. 2004

Eclipsed-flux deficiency in SV Cam Eclipsed flux is ~30% less than best-fit T eff indicates.Eclipsed flux is ~30% less than best-fit T eff indicates. f S ~40%f S ~40% Jeffers et al. 2004

Evolutionary effects of flux blocking Star expands slightlyStar expands slightly Photospheric T eff increasesPhotospheric T eff increases –  significant effects on HR diagrams of young open clusters, e.g. Pleiades Stauffer et al 2003Spruit & Weiss 1986

Imaging of stellar surfaces Direct imaging?Direct imaging? Stellar Imager mission concept:Stellar Imager mission concept: –Goal is 50,000 km resolution on a Sunlike star 4 pc away –Requires angular resolution µas –0.5-km space-based UV-optical interferometer array ?

Rotational broadening of photospheric lines Rotational Doppler shift dominates broadening of stellar photospheric lines in rapid rotators.Rotational Doppler shift dominates broadening of stellar photospheric lines in rapid rotators. Rotation profile contains information about surface features (Goncharsky et al 1977, Vogt & Penrod 1983)Rotation profile contains information about surface features (Goncharsky et al 1977, Vogt & Penrod 1983) Stauffer et al 1997

Starspot “bumps” in spectral lines Intensity AA v sin i-v sin iv(spot) Velocity v sin i-v sin iv(spot) Velocity

Imaging of stellar surfaces on a budget Combine profiles of all recorded photospheric lines to boost S:N.Combine profiles of all recorded photospheric lines to boost S:N. Compute synthetic line profiles from trial image.Compute synthetic line profiles from trial image. Iterate to target  2 at maximum entropy.Iterate to target  2 at maximum entropy. Get simplest image that fits data.Get simplest image that fits data. Nearly always get a dark polar cap.Nearly always get a dark polar cap. Starspot signatures in photospheric lines -v sin i +v sin i

Example: Speedy Mic (K3V) Spots present at all latitudes including polar regions.Spots present at all latitudes including polar regions. Barnes et al 2004

Example: HDE Strassmeier et al 1998: WTTS, v sin i = 78 km s –1Strassmeier et al 1998: WTTS, v sin i = 78 km s –1

Polar fields Schrijver & Title (2002) modelled flux emergence on stars of different rotation rates.Schrijver & Title (2002) modelled flux emergence on stars of different rotation rates. Rapid rotators develop rings of opposite polarity at poles.Rapid rotators develop rings of opposite polarity at poles. Note reversal of polar fields over cycle.Note reversal of polar fields over cycle. Also Schüssler (1997) modelled buoyant flux tube emergence. Flux tubes deflected to high latitudes on rapid rotators.Also Schüssler (1997) modelled buoyant flux tube emergence. Flux tubes deflected to high latitudes on rapid rotators.

Andrew Collier Cameron, Moira Jardine, Duncan Mackay, Kenny Wood, John Barnes, Sandra Jeffers (University of St Andrews) Jean-François Donati (Obs. Midi-Pyrenees, Toulouse). Meir Semel (Obs. de Paris, Meudon) Magnetic activity on rapidly rotating stars II:Temporal evolution Tracking starspotsTracking starspots Time-varying differential rotationTime-varying differential rotation Differential rotation along the main sequenceDifferential rotation along the main sequence Stellar magnetogramsStellar magnetograms 3D coronal structure3D coronal structure

What else can we learn from stellar surface maps? Snapshots:Snapshots: –Unpolarized: Latitude distributions of spots –Locations of slingshot prominence complexes –Circularly polarized: Magnetic topology of corona Days-weeks timescale:Days-weeks timescale: –starspots trace surface differential rotation and meridional flows Weeks-months:Weeks-months: –Lifetimes of individual spots and magnetic regions Years:Years: –Stellar butterfly diagram: Dynamo cycles –Polarity reversals?

Polar spots and convective-zone depth LQ Lup (G2)LQ Lup (G2) Donati et al (2000)Donati et al (2000)

Polar spots and convective-zone depth HE 699 (G2-3V; alpha Per G dwarf)HE 699 (G2-3V; alpha Per G dwarf) Jeffers et al (2002)Jeffers et al (2002)

Polar spots and convective-zone depth HK Aqr (M1)HK Aqr (M1) Barnes et al (2004)Barnes et al (2004)

Polar spots and convective-zone depth RE J (M1)RE J (M1) Barnes et al (2001)Barnes et al (2001)

Polar spots and convective-zone depth G3V G6V G8V K0V K3V M1V

Surface brightness: 1996 Dec Equator rotatesEquator rotates faster than pole faster than pole –solar-like shear –Prot ~ 0.5 d –Equator laps pole by 1 cycle every ~ 120d 1 cycle every ~ 120d

Surface shear: 1996 December CCF for surface- brightness imagesCCF for surface- brightness images CCF for magnetic images:CCF for magnetic images:

Starspots as flow tracers Individual spot trails have their own recurrence periods.Individual spot trails have their own recurrence periods. Velocity amplitude of sinusoid:Velocity amplitude of sinusoid: Rotation rate at latitude  Stellar radius Axial inclination

Matched-filter analysis Travelling gaussian:Travelling gaussian: Spot velocity amplitude: Foreshortening and limb darkening Inclination Latitude Spot phase angle relative to observer’s meridian Intrinsic line width Spot rotation rate:

Optimal scaling g ij : equatorial spot at phase 0.5 Badness of fit: Scale factor: x ij (phase binned on trial period) 22

Optimal scaling g ij : equatorial spot at phase 0.5 Badness of fit: Scale factor: x ij (phase binned on trial period) 22

Differential rotation: 1988 Dec Model fit:Model fit:

Differential rotation: 1992 Jan Model fit:Model fit:

Differential rotation: 1993 Nov Model fit:Model fit:

Differential rotation: 1995 Dec Model fit:Model fit:

Differential rotation: 1996 Dec Model fit:Model fit:

Differential rotation: 1998 Dec Model fit:Model fit:

Differential rotation: 2000 Dec Model fit:Model fit:

Differential rotation: 2001 Dec Model fit:Model fit:

Differential rotation Differential rotation rate doubled in 3 years from 1988 Dec to 1992 Jan.Differential rotation rate doubled in 3 years from 1988 Dec to 1992 Jan. As equator speeds up, polar regions slow down.As equator speeds up, polar regions slow down. Rotation rate at  ~ 40 o remains ~ constant.Rotation rate at  ~ 40 o remains ~ constant.

Plot estimates in Ω eq -dΩ plane Interpret differences as: distinct anchoring depths of tracers within CZ temporal changes in angular rotation profile within CZ Impact on convective zones Angular rotation in convective zoneAngular rotation in convective zone

plot estimates in Ω eq -dΩ plane changes in differential rotation:  powered with a few % of L *  correspond to a field of ≈10 kG in the whole CZ  sufficient to generate orbital period fluctuations of binary stars (Applegate 1992) Impact on convective zones Angular rotation in convective zoneAngular rotation in convective zone

Differential rotation along the main sequence Barnes et al. 2004

Comparison with other techniques Barnes et al. 2004

Zeeman Doppler Imaging Zeeman effect:Zeeman effect: –  component: Linear polarization, no shift –  components: Elliptical polarization,  ~ ± gB Field orientation and line-of-sight:Field orientation and line-of-sight: –Circular polarization (  cpt) strongest when B // line of sight. How stellar rotation helps:How stellar rotation helps: –Rotational Doppler effect separates magnetic regions in velocity space. –Field orientation relative to line of sight changes as magnetic region crosses disc. Landé g-factor for line Local magnetic field strength

The Semel Polarimeter Aberration-free linear polarizing beamsplitter /4 plate Optical axis switched at ±45 o relative to beamsplitter axes AAT f/8 Cass focus Focal reducer Dual beams analyzed for opposite circular polarization states Bowen- Walraven image slicer at UCLES slit position UCLESUCLES Semel et al 1993: A&A 278, 231 Dual fibre feed to UCLES slit area

Stokes V: weak-field approximation I ( ) V ( ) LeftRight Difference High g Low g

Multi-line imaging Essential for ZDIEssential for ZDI –Stokes V signature is typically < 10 –4 times continuum. –Typical S:N is 300 in continuum. –Weighted least-squares deconvolution procedure recovers profile information from up to 4600 images of 2700 lines. Nice for Stokes I tooNice for Stokes I too –Huge sensitivity gain – turns the AAT into a 160-m telescope! –Allows use of full time and wavelength resolution. –Unprecedented amounts of surface detail recoverable. * = Weight =g-factor * depth

Detecting magnetic fields Zeeman effect in spectral linesZeeman effect in spectral lines circular and linearcircular and linear polarisation in line profiles polarisation in line profiles amplitude usually < 0.1% amplitude usually < 0.1%

Detecting magnetic fields Zeeman effect in spectral linesZeeman effect in spectral lines circular and linearcircular and linear polarisation in line profiles polarisation in line profiles amplitude usually < 0.1% amplitude usually < 0.1% Stack line profiles withStack line profiles with Least-squares Least-squares Deconvolution Deconvolution

Stokes V time-series spectra Stokes I & V dynamic spectra of AB Dor Demonstrates rotationalDemonstrates rotational modulation of Zeeman modulation of Zeeman signature signature Yields location ofYields location of magnetic regions & magnetic regions & orientation of field lines orientation of field lines

Radial field: 1996 Dec

The shape of a stellar corona Altshuler & Newkirk (1969):Altshuler & Newkirk (1969): –fitted potential field models to solar surface magnetograms. –Mimic transition from closed corona to solar wind by imposing a “source surface” at several solar radii. Field beyond source surface is radial. Can we do this for other stars?Can we do this for other stars?

Coronal topology and X-ray emission Jardine, Wood, Cameron, Donati & Mackay(2002) MNRAS, 336:Jardine, Wood, Cameron, Donati & Mackay(2002) MNRAS, 336: Potential field.Potential field. –1995 Dec Dec magnetograms AB Dor rotation rate.AB Dor rotation rate. Isothermal coronaIsothermal corona –T=10 7 K Base pressure  B 2.Base pressure  B 2. –But p=0 on open lines Soft X-ray emissivity  n e 2. Soft X-ray emissivity  n e 2. Monte Carlo RT code.Monte Carlo RT code. Centrifugal compression/stripping:Centrifugal compression/stripping: – p=0 on field lines where p > B 2 / 2µ (cf. Mestel & Spruit 1987)

Centrifugal stripping and supersaturation P = 0.51 d P=0.17 d Jardine 2004

Centrifugal stripping and supersaturation Co-rotation radius shrinks as Ω increasesCo-rotation radius shrinks as Ω increases Loops near co-rotation burst openLoops near co-rotation burst open Coronal X-ray emission measure decreasesCoronal X-ray emission measure decreases Jardine 2004

Slingshot prominences: signatures AB Dor, AAT/UCLES, 1996 Dec 29AB Dor, AAT/UCLES, 1996 Dec 29 Donati et al 1999Donati et al 1999 Starspot signatures in photospheric lines -v sin i +v sin i Absorption transients in H alpha -v sin i +v sin i

Coronal condensations: single stars Detected in 90% of young (pre-) main sequence stars with P rot <1 dayDetected in 90% of young (pre-) main sequence stars with P rot <1 day –AB Dor (K0V): Collier Cameron &Robinson 1989 –HD =“Speedy Mic” (K0V): Jeffries 1993 –4 G dwarfs in  Per cluster: Collier Cameron & Woods 1992 –HK Aqr = Gl 890 (M1V): Byrne, Eibe & Rolleston 1996 –RE J : Eibe 1998 –PZ Tel: Barnes et al 2000 (right) P rot = 1 day (slowest yet) –Pre-main sequence G star RX J (Donati et al 2000) --prominences in emission!

Emission signatures Seen only in the most rapidly-rotating, early G dwarfs, e.g. RX J (Donati et al 2000):Seen only in the most rapidly-rotating, early G dwarfs, e.g. RX J (Donati et al 2000): Star is viewed at low inclination; uneclipsed H  - emitting clouds trace out sinusoids

Tomographic back-projection Clouds congregate near co-rotation radius (dotted).Clouds congregate near co-rotation radius (dotted). Little evidence of material inside co-rotation radius.Little evidence of material inside co-rotation radius. Substantial evolution of gas distribution over 4 nights.Substantial evolution of gas distribution over 4 nights.

Latitude dependence AB Dor prominences need to be anchored at high latitude to cross stellar disk, since i = 60 degrees.AB Dor prominences need to be anchored at high latitude to cross stellar disk, since i = 60 degrees. What about other stars with different inclinations?What about other stars with different inclinations? –BD : Low inclination, no transients found: Jeffries et al 1994 

Where do bipoles emerge on young stars? Solar-type starSolar-type star Bipoles emerge at solar-like latitudes as cycle progresses.Bipoles emerge at solar-like latitudes as cycle progresses. Solar transport coeffs.Solar transport coeffs. Flux emergence rate 30 times solar.Flux emergence rate 30 times solar. Solar meridional flow rate (11 m/sec)Solar meridional flow rate (11 m/sec) 2 cycles per movie loop.2 cycles per movie loop. Van Ballegooijen code, modified by Duncan Mackay.Van Ballegooijen code, modified by Duncan Mackay. Confirms earlier work by Schrijver & Title for similar flux emergence rate.Confirms earlier work by Schrijver & Title for similar flux emergence rate.

Where do bipoles emerge on young stars? Solar-type starSolar-type star Bipoles emerge at latitudes 70 deg to 10 deg as cycle progresses.Bipoles emerge at latitudes 70 deg to 10 deg as cycle progresses. Solar transport coeffs.Solar transport coeffs. Flux emergence rate 30 times solar.Flux emergence rate 30 times solar. Solar meridional flow rate (11 m/sec)Solar meridional flow rate (11 m/sec) 2 cycles per movie loop.2 cycles per movie loop. Flux vanishes at all latitudes around activity minimum.Flux vanishes at all latitudes around activity minimum. Mostly unipolar caps.Mostly unipolar caps.

Where do bipoles emerge on young stars? Solar-type starSolar-type star Bipoles emerge at latitudes 70 deg to 10 deg as cycle progresses.Bipoles emerge at latitudes 70 deg to 10 deg as cycle progresses. Solar transport coeffs.Solar transport coeffs. Flux emergence rate 30 times solar.Flux emergence rate 30 times solar. 9 x Solar meridional flow rate (100 m/sec)9 x Solar meridional flow rate (100 m/sec) 2 cycles per movie loop.2 cycles per movie loop. Flux vanishes at all latitudes around activity minimum.Flux vanishes at all latitudes around activity minimum. Multipolar caps.Multipolar caps.

Where do bipoles emerge on young stars? Solar-type starSolar-type star Bipoles emerge at range of latitudes around 35 degrees (no butterfly diagram).Bipoles emerge at range of latitudes around 35 degrees (no butterfly diagram). Solar transport coeffs.Solar transport coeffs. Flux emergence rate 30 times solar.Flux emergence rate 30 times solar. 9 x Solar meridional flow rate (100 m/sec)9 x Solar meridional flow rate (100 m/sec) 2 cycles per movie loop.2 cycles per movie loop. Flux vanishes at all latitudes around activity minimum.Flux vanishes at all latitudes around activity minimum. Multipolar caps.Multipolar caps.

Where do bipoles emerge on young stars? Solar-type starSolar-type star Bipoles emerge at range of latitudes around 35 degrees (no butterfly diagram).Bipoles emerge at range of latitudes around 35 degrees (no butterfly diagram). Solar transport coeffs.Solar transport coeffs. Flux emergence rate 30 times solar.Flux emergence rate 30 times solar. 9 x Solar meridional flow rate (100 m/sec)9 x Solar meridional flow rate (100 m/sec) Slow-motion action replay.Slow-motion action replay. Flux vanishes at all latitudes around activity minimum.Flux vanishes at all latitudes around activity minimum. Multipolar caps.Multipolar caps.

Slingshot prominences and polar fields McIvor et al 2003McIvor et al 2003 Possible polar field configurations:Possible polar field configurations:

Slingshot prominences and polar fields McIvor et al 2003McIvor et al 2003 Corresponding coronal field configurations:Corresponding coronal field configurations: Unipolar cap supports fewer high-latitude prominences.Unipolar cap supports fewer high-latitude prominences.

Summary and prospects Rotational shear ∆ΩRotational shear ∆Ω –Decreases strongly with increasing convective-zone depth –Increases weakly with increasing stellar rotation rate. Differential rotation rate shows year-to-year variabilityDifferential rotation rate shows year-to-year variability –Consistent with Applegate (1992) mechanism for binaries. Polar spot activity appears stronger in shallow convective zones.Polar spot activity appears stronger in shallow convective zones. Advection and diffusion of emergent bipoles can give rise to flux pile-up at polar caps.Advection and diffusion of emergent bipoles can give rise to flux pile-up at polar caps. Prominence distribution suggests mixed-polarity caps.Prominence distribution suggests mixed-polarity caps. Enhanced poleward meridional flows may be needed.Enhanced poleward meridional flows may be needed. –Should be detectable if present.