Large amplitude transverse oscillationsin a multi-stranded EUV prominence

centre for fusion, space and astrophysics

J. M. HarrisC. Foullon, V. M. Nakariakov, E. Verwichte

Prominence Oscillations

☼ Solar prominences float in the corona, held in place by the magnetic field

☼ They are cooler and denser than the coronal plasma, therefore appearing:

• bright against a dark background in 304 Å (70,000 K),

• but dark against the bright corona in 195 Å (1.5 million K).

☼ Oscillations traditionally observed in Hα lines as ‘winking filaments’

• e.g. Dyson et al (1930), Brusek (1951), Ramsey & Smith (1966).

• The line of sight velocity can be measured using the Doppler shift.

195 Å

EUV images of a prominence on the NE limb (SOHO/EIT)

304 Å

Prominence Oscillations

☼ Now also observed by EUV imagers

• e.g. Foullon et al (2004, 2010), Isobe & Tripathi (2006), Pinter et al (2008).

• This enables long periods and the field of view velocity to be measured.

☼ Small amplitude oscillations: velocity ≈ 2-3 km/s.

☼ Large amplitude oscillations: velocity > 20km/s.

• can be triggered by nearby flares or EIT waves,

• observational analyses are scarce, see Tripathi et al (2009) for a review.

☼ Analysis of these oscillations enables us to:

• measure plasma parameters (e.g. magnetic field) via coronal seismology,

• understand more about solar prominences (e.g. link to prominence eruptions),

• verify oscillation and damping theories.

Prominence Oscillation on 30th July 2005

☼ Observed using SOHO/EIT:

• 195 Å, 12 min cadence,

• 304 Å, 6 hour cadence.

☼ Two successive trains of transverse oscillations,

☼ triggered by EIT waves from two flares in the same remote active region:

• X1.3 class flare at 06:17

• C8.9 class flare at 16:39

EIT Waves

16:45 - 16:35 16:55 - 16:45 17:07 - 16:55 17:17 - 17:07

EIT

Wav

e 2

☼ 1st EIT wave seen over only 1 frame using running difference images (Type II radio burst, 1801 km/s)

☼ 2nd EIT wave seen over 4 fast frames (no Type II reported)

☼ Intensity depletion is larger following the 1st EIT wave

16:35 - 16:35

06:44 - 06:02 17:07 - 16:35

EIT

Wav

e 1

EIT

Wav

e 2

Evolution of the Apparent Height

1cos1 00

LtL+

R

h=

R

Lh

oo

Foullon & Verwichte (2006)

Image using ratio of 304/195 Åwith region of interest indicated

L0

= Carrington longitude

= Carrington longitude when over the limb

tL= apparent height

= actual height

= solar radius

Lhh 0R o

rotation

h 0

Lh

0L=tLposition over the limb when

positionat tL

Evolution of Apparent Height

008.070.160 =R

h

o

1.027.50 =L

5 daysflare 1flare 2 Region of interest moves with

the rotation of the prominence

Time - Distance Plots: slit 1

Time - Distance Plots: slit 4

Analysis of Time Series: slit 1

P = 122 ± 23 minΤ = 131 ± 94 minv = 12 ± 5 km/s

P = 101 ± 1 minΤ = 218 ± 47 minv = 32 ± 10 km/s

Analysis of Time Series: slit 4

P = 97 ± 21 min Τ = 274 ± 497 minv = 5 ± 4 km/s

P = 94 ± 2 min Τ = 119 ± 19 minv = 30 ± 4 km/s

P = 108 ± 2 min Τ = 277 ± 80 minv = 17 ± 3 km/s

P = 111 ± 4 min Τ = 362 ± 212 minv = 11 ± 3 km/s

Results: Amplitude & Period

Results: Damping Times

0.100.91±Pτ consistent with damping via resonant absorption

Coronal loop oscillation data:Nakariakov et al (1999) Aschwanden et al (2002)Wang & Solanki (2004) Verwichte et al. (2004)Van Doorsselaere et al. (2007) Hori et al. (2007)Van Doorsselaere et al (2009) Verwichte et al. (2009)Verwichte et al. (2010)

Prominence oscillation data:Harris et al (2010)

Conclusions

☼ Large amplitude transverse (horizontal) prominence oscillation.

☼ Velocity amplitudes:

• Generally increasing with height, up to 32km/s following the X1.3 class flare and up to 12km/s after the C8.9 class flare, as expected due to the difference in flare energy.

☼ Periods of around 100 minutes (±10 minutes):

• Generally increasing with height and varying for different strands, indicating that the prominence doesn’t oscillate as a solid body but according to its filamentary structure.

• Around 10% shorter during the 1st oscillatory train than the 2nd (c.f. 15x flare energy, 5x amplitude), suggesting the period is largely dependent on the properties of the prominence rather than the triggering mechanism, as expected for an MHD mode.

☼ Damping times:

• 2 to 3 periods for most strands. when combined with data from loop oscillations, this is consistent with damping via resonant absorption.

• Other strands exhibit much longer decay times, but the errors in these cases are very large.

0.100.91±Pτ

Harris et al. 2010, in preparation