Wednesday, 5 September 12
Introduction
‘The
Carrington
- The
experience’
Dynamic Corona
The ‘Carrington’ Experience
Wednesday, 5 September 12
THE UNIVERSITY OF WARWICK
Centre for Fusion, Space and Astrophysics
1
2
Erwin Verwichte - Sun Observation - ASSSP12
Overview of aspects of solar observation focused mainly on
EUV imaging of the solar corona.
Erwin Verwichte
Sun Observation
ASSSP12 University of Warwick 2-7 Sep 2012
104”
1000”
100”
10”
1”
0.1”
Wednesday, 5 September 12
SoHO/EIT
1000s
hour
100s
minute
OSO-7&8
1971-1978
10s
1s
second
TRACE 3EUV full-disk mosaic 1999/08/02
3
3
SoHO/EIT 30/10/2005
The Sun’s appearance changes
throughout the atmosphere
4
Erwin Verwichte - Sun Observation - ASSSP12
temporal
resolution
rs
age
m
i
e
Flar
RHESSI
2000-
SMM/HXIS
1980-1998
Yohkoh/SXT
1991-2001
CS: speed of sound, VA: Alfvén speed
CMEs
Loops
Prominences
P
P
CS
VA
λ=
λ=
CoMP
2004-
SDO/AIA
FASR
Hinode/XRT
2010
2010?
2006Hinode/EIS
NORH
2006-
Nanoflares
Solar Orbiter
1996Microflares
Skylab/S056
1973-1974
LD Flares
2006-
s
gerTRACE
a
m
li
1998ona
Cor STEREO/EUVI
The Solar atmosphere
Wednesday, 5 September 12
Radio
X-ray
EUV
1 Rsun
100 Mm
1 Mm
100 km
spatial
resolution
3
Erwin Verwichte - Sun Observation - ASSSP12
With current instruments we resolve dynamics on macroscopic scales,
where magnetohydrodynamics is applicable.
Observational Scales
5 min
TRACE 3EUV full-disk mosaic 1999/08/02
Wednesday, 5 September 12
TRACE 3EUV full-disk mosaic 1999/08/02
The Solar atmosphere
Wednesday, 5 September 12
The Solar atmosphere
3
3
4
SoHO/EIT 30/10/2005
The Sun’s appearance changes
throughout the atmosphere
4
Erwin Verwichte - Sun Observation - ASSSP12
SoHO/EIT 30/10/2005
The Sun’s appearance changes
throughout the atmosphere
Erwin Verwichte - Sun Observation - ASSSP12
TRACE 3EUV full-disk mosaic 1999/08/02
Wednesday, 5 September 12
TRACE 3EUV full-disk mosaic 1999/08/02
The Solar atmosphere
Wednesday, 5 September 12
The Solar atmosphere
3
3
4
SoHO/EIT 30/10/2005
The Sun’s appearance changes
throughout the atmosphere
4
Erwin Verwichte - Sun Observation - ASSSP12
SoHO/EIT 30/10/2005
The Sun’s appearance changes
throughout the atmosphere
Erwin Verwichte - Sun Observation - ASSSP12
TRACE 3EUV full-disk mosaic 1999/08/02
Wednesday, 5 September 12
The corona is optically thin, filled
with rarified plasma at millions
of degrees Kelvin and its
thermal part emits through lineemission from minority species
excited by collisions. This gives
rise to emission line spectra,
visible in the EUV and X-ray.
Solar spectrum
Wednesday, 5 September 12
The Solar atmosphere
3
5
4
Culhane & Acton (1974)
The largest part of solar radiation comes
from the photosphere, which can be
modeled as a black-body radiator visible
in the visible, UV and IR, with elements
providing absorption lines,
Erwin Verwichte - Sun Observation - ASSSP12
SoHO/EIT 30/10/2005
The Sun’s appearance changes
throughout the atmosphere
Erwin Verwichte - Sun Observation - ASSSP12
by the moon
ac
a
et
01
National Astronomical Observatory of Japan
Invented by Bernard Lyot in 1939
by an artificial obstruction
Downsides: at mercy of orbital mechanics of the Moon and
at mercy of the Earth’s atmospheric conditions.
0)
Wednesday, 5 September 12
7
Erwin Verwichte - Sun Observation - ASSSP12
The corona can be seen in the visible light because its electrons
(K-corona) and dust (F-corona) scatter light from the photosphere.
Intensity is proportional to density (info on structuring), scattering
has no temperature info.
Natural and unnatural coronagraphs
Wednesday, 5 September 12
molecular
There are two windows to observe space from the ground: visible and radio
atomic
6
Erwin Verwichte - Sun Observation - ASSSP12
Earth atmosphere is a barrier to observe the corona?
s
Pa
ff
ho
2
l. (
9
Erwin Verwichte - Sun Observation - ASSSP12
Polarization
Wednesday, 5 September 12
P / sin2 (✓B ) I
V / B cos ✓B
✓
dI
d!
◆
•Line-of-sight magnetic field from V
(Zeeman, tricky in corona as V/I < 0.001)
•Plane-of-sky magnetic field direction from
Q and U (resonance scattering)
The Stokes parameters contain
information about the magnetic field:
also
P = [Q2 + U2]1/2
I = |Ex|2 + |Ey|2 Total intensity (coherent + incoherent)
Q = |Ex|2 - |Ey|2 Horizontal-vertical linear polarization balance
U = |Ea|2 - |Eb|2 45 degrees linear polarization balance
V = |EL|2 - |ER|2 Left-right circular polarization balance
b
L
y
R
a
x
Stokes parameters describe the coherence and polarization of light.
We measure the light as a vector in the plane of the sky using polarizing filters.
Stokes parameters
Ground-measurements are
flexible and can be sensitive
and high time-cadence, but
again you are at the mercy
of the Earth’s atmosphere!
Example: Polarimeter which measures the polarization of the light with a
coronagraph at several wavelengths around the emission line.
Wednesday, 5 September 12
8
Erwin Verwichte - Sun Observation - ASSSP12
There are ‘forbidden’ emission lines in the visible spectrum that are sensitive
to coronal temperatures that offer more, e.g. Fe XIII at 1074.7 nm!
Natural and unnatural coronagraphs
CoMP (Tomczyk et al. 2008)
Wednesday, 5 September 12
A first instrument that uses normal incidence (reflection of a mirror) is the
Extreme-ultraviolet Imaging Telescope (EIT) on board SoHO.
Extreme-ultraviolet Imaging Telescope (EIT)
Wednesday, 5 September 12
11
Erwin Verwichte - Sun Observation - ASSSP12
One traditional method is using diffraction grating (e.g. Skylab, MOSES rocket).
You cannot use refraction (lenses) in the EUV domain because matter
absorbs this light. EUV telescopes thus need to use reflection (mirrors).
Skylab overlappograph
10
Erwin Verwichte - Sun Observation - ASSSP12
An EUV imager takes images of the Sun through a broadband wavelength
filter in which one (or several) dominant EUV emission line(s) are present.
Imaging in the EUV
(x,y)
Wednesday, 5 September 12
(λ): CCD quantum efficiency [1]
D(λ,t): Degradation over time [1]
A: open aperture area [cm2] = 11.97
TEF(λ): Entrance Filter transmissivity [1]
RPM(λ): Primary Mirror reflectivity [1]
RSM(λ): Secondary Mirror reflectivity [1]
TFW(λ): Filter Wheel transmissivity [1]
TSL(λ): Stray-Light Filter transmissivity [1]
Ae↵ ( ) = A TEF ( ) RPM ( ) RSM ( ) TFW ( ) TSL ( ) ✏( ) D( , t)
To compare with other instruments we use the Effective Area:
0
Z1
Z
I(x, y) = Ae↵,i ( ) G( ) F (x, y) P ( , ✓, ne , T ) ⇤ PSF(✓)d✓ d
The throughput (transfer function) of the telescope:
Extreme-ultraviolet Imaging Telescope (EIT)
Wednesday, 5 September 12
CCD is 1024x1024 back-illuminated.
A stray-light filter eliminates any
reflections off instrument parts.
The filter wheel allows to add some
additional Al-filtering.
13
Erwin Verwichte - Sun Observation - ASSSP12
The optics are a Ritchey-Chretien design (a bit like a Cassegrain). They
are multi-layer coated with Mo and Si. Each quadrant is different, with a
different EUV wavelength window.
A sector wheel allows to select one of the four quadrants of the optics.
The shutter lets the CCD see or not
see the EUV light.
12
Erwin Verwichte - Sun Observation - ASSSP12
Al-filters block out the visible part of the solar spectrum (just as eclipse
glasses would). They are mounted on two filter support grids.
Extreme-ultraviolet Imaging Telescope (EIT)
(Dere et al. 2000).
12398
g
3.65
Wednesday, 5 September 12
AIA/SDO full PSF (Grigis et al 2012)
The point-spread function PFS is the shape of a point source at the CCD
after it passed through all the optics. This includes entrance and other
filter diffraction (remember grid!) and the two mirrors
core PSF convolved with a
pixel for EIT 304Å bandpass
15
Erwin Verwichte - Sun Observation - ASSSP12
EIT CCD flat-field
Extreme-ultraviolet Imaging Telescope (EIT)
Wednesday, 5 September 12
where apix is the CCD pixel area and f
the telescope focus, which is a function
of wavelength
Also, integration over the space-angle
θ for the pixel can be written as
✓
◆2
Z
apix
d✓ =
f
(x,y)
F(x,y) is the CCD flat-field. Not all pixels are of identical quality.
with g how many electrons makes one DN.
G( ) =
14
Erwin Verwichte - Sun Observation - ASSSP12
G is the gain of the CCD-camera in DN (Digital Number) per photon. In
other words, how much image value is a photon worth?
Extreme-ultraviolet Imaging Telescope (EIT)
LOS
ne (T, z) nH (T, z) dz
0
Wednesday, 5 September 12
Z1
EM = DEM(T ) dT
and
dEM
dz
DEM =
= ne nH
or DEM =
dT
dT
T
✓
dz
dT
◆
dT
O’Dwyer et al. (2010)
ne nH
TZ
+dT
The Differential Emission Measure (DEM) is defined as the distribution of
emissivity (ne nH) in the line-of-sight as a function of temperature(linear version).
EM =
Z
17
Erwin Verwichte - Sun Observation - ASSSP12
First, we define the emission EM to be the line-of-sight integration emissivity
ne nH (we assume resonance emission line excited by collisions):
Extreme-ultraviolet Imaging Telescope (EIT)
Wednesday, 5 September 12
Created using CHIANTI
where G contains the atomic information of all emission lines (and free). This
function is usually much stronger temperature than density dependent.
0
Z1
P ( , ne , T ) = G( , ne , T ) DEM(T ) dT
16
Erwin Verwichte - Sun Observation - ASSSP12
Also, the incident solar flux [photons cm-2 s-1 sr-1 Å-1 ] may be split into the
atomic emission line details and the emissivity of the plasma:
Extreme-ultraviolet Imaging Telescope (EIT)
0
Wednesday, 5 September 12
Erwin Verwichte - Sun Observation - ASSSP12
19
Because multiple emission lines
contribute to H(T), its shape
can be multi-peaked.
✓
◆2
Z1
apix
H(T ) = Ae↵,i ( ) G( )
G( , ne (T ), T ) d
f
Extreme-ultraviolet Imaging Telescope (EIT)
Wednesday, 5 September 12
0
Z1
I = H(T ) DEM(T ) dT
18
Erwin Verwichte - Sun Observation - ASSSP12
Putting all together, we can describe the instrument’s response in terms
of temperature bandpasses H(T):
Extreme-ultraviolet Imaging Telescope (EIT)
Wednesday, 5 September 12
)
284/195
195/171
195/171 ratio
I1
H1 (T0 )
=
= F12 (T0 ) ) T0 = F121 (I1 /I2 )
I2
H2 (T0 )
Assume the coronal plasma is isothermal, i.e. DEM(T) = EM0 δ(T-T0). Then,
I1 = H1 (T0 ) EM0
I2 = H2 (T0 ) EM0
21
Erwin Verwichte - Sun Observation - ASSSP12
We can take the ratio of two bandpasses and in a certain interval of
temperature, where the ratio is monotonic and near the peak responses of
the two bandpasses, we can estimate the temperature crudely.
Temperature bandpass ratio
Wednesday, 5 September 12
20
Erwin Verwichte - Sun Observation - ASSSP12
Example of degradation of the CCD. Because the entrance filter grid
throws a shadow onto the CCD, those pixels in the shadow have seen
less total light and have less degraded. Hence, their gain is higher. In
images we see the grid as a bright feature!
Extreme-ultraviolet Imaging Telescope (EIT)
Wednesday, 5 September 12
22
23
Erwin Verwichte - Sun Observation - ASSSP12
Lemen et al. (2012)
An improved time resolution but not dramatically different from TRACE
Same spatial resolution as TRACE
What advantages does the SDO-era bring?
Solar Dynamics Observatory
Wednesday, 5 September 12
22
Erwin Verwichte - Sun Observation - ASSSP12
The Atmospheric Imaging Assembly (AIA) on board SDO is the current
generation of EUV imager but works on similar principles as EIT but at higher
spatial and temporal resolution and with more bandpasses.
Solar Dynamics Observatory
Wednesday, 5 September 12
Solar Dynamics Observatory
Wednesday, 5 September 12
Combination with STEREO is useful for geometry estimates
Multiple simultaneous bandpasses: explore phenomena
beyond bulk to hot and cool coronal plasma.
No SAA occurring just when it is interesting
Statistical studies are now only limited by man-power!
24
23
Erwin Verwichte - Sun Observation - ASSSP12
Full-sun FOV and synoptic program means no event will be
missed.
SDO/AIA can be relied upon for spectroscopic studies
22
Erwin Verwichte - Sun Observation - ASSSP12
An improved time resolution but not dramatically different from TRACE
Same spatial resolution as TRACE
What advantages does the SDO-era bring?
Solar Dynamics Observatory

1
2
✓
T
T0
T
◆
2
T)
Wednesday, 5 September 12
2
=
n
1
2
i
X
Ii
noise
EM0 hi (T0 , T )
4) We find the best values of T0 and ΔT, and hence EM0, by minimizing the
difference with the observations
✓
◆
i
3) For every value of T0 and ΔT, we calculate EM0:
P
Ii
i
EM0 = P
hi (T0 , T )
0

✓
◆
Z1
Hi (T )
1 T T0
p
Ii = EM0
exp
dT = EM0 hi (T0 ,
2
T
2⇡ T
2) The intensity (DN/s) of each bandpass is then of the form:
26
Erwin Verwichte - Sun Observation - ASSSP12
which in the limit of ΔT ⟶ 0, becomes DEM(T) = EM0 δ(T-T0).
1
DEM(T ) = EM0 p
exp
2⇡ T
1) We assume that DEM(T) is of the form:
25
Erwin Verwichte - Sun Observation - ASSSP12
Again, we may estimate single temperature and density in the line-of-sight.
DEM method
Wednesday, 5 September 12
Combining bandpasses (171,193 and 211) in threecolour channels gives an idea of temperature.
Solar Dynamics Observatory
j
j
j
i
X
hij1 Ii
Wednesday, 5 September 12
Hannah & Kontar (2012)
DEMj =
from Ii:
However, this inversion is often ill-posed as
the information from a limited number of
bandpasses is not enough. Careful
regularization techniques then need to be
employed and image noise taken into
account.
The problem then becomes a matter of inverting the matrix hij to find
0
where DEMj(T) are known profiles. Then the intensity (DN/s) has the form:
1
XZ
X
X
Ii =
Hi (T ) DEMj (Tj ) dT =
Hi (Tj ) DEMj (Tj ) =
hij DEMj
j
28
Erwin Verwichte - Sun Observation - ASSSP12
More formally, you can write DEM(T) as made from many temperature bins:
X
DEM(T ) =
DEMj (Tj )
DEM method
Wednesday, 5 September 12
27
Erwin Verwichte - Sun Observation - ASSSP12
Note that this is an approximate method as the DEM(T) is highly idealised.
Example from Aschwanden et al. (2011)
DEM method
Wednesday, 5 September 12
After much rain, the spread
is practically uniform.
30
In the beginning, each tile
will have much variation of
small number of rain spots.
Erwin Verwichte - Sun Observation - ASSSP12
Higher signal-to-noise (crisper)
Example of constant, uniform rain on the pavement:
Photon noise
Wednesday, 5 September 12
Lower signal-to-noise (fuzzier)
AIA193
The more photons collected the better resolved (crisper) the image
becomes.
AIA94
29
Erwin Verwichte - Sun Observation - ASSSP12
The measured intensity is subject to errors. It is important to know about to
avoid mistakes. An inherent image noise is photon noise (or shot noise).
Photon noise
I t
G( ) d
= R
I t
G( ) d
sZ
I t
=
G( )d
p
I
t
Poisson distributions as a function
of number of occurrences k
Wednesday, 5 September 12
The larger the exposure time, the less effect of photon noise.
SNR =
Hence the signal-to-noise ratio SNR becomes
2
where G(λ) is the gain [DN/photon]. The photon
noise then becomes
N = R
The standard deviation of N photons is σ = √N .
Relating back to the formula for image intensity,
N relates to I for exposure time ∆t as
31
Erwin Verwichte - Sun Observation - ASSSP12
Photons hitting a CCD pixel are independent random events (similar to rain
drops hitting the pavement). The distribution of the photons follows Poisson
statistics (for large intensity it becomes Gaussian).
Photon noise
Wednesday, 5 September 12
After much rain, the spread
is practically uniform.
30
Erwin Verwichte - Sun Observation - ASSSP12
In the beginning, each tile
will have much variation of
small number of rain spots.
Example of constant, uniform rain on the pavement:
Photon noise
2
photon (I
t) +
Wednesday, 5 September 12
Example of cosmics
Wednesday, 5 September 12
⇡
p
+
2
read
+
2
dig
+
2.3 + 0.06 I t DN
2
dark (TCCD )
Example: typical AIA 171Å image:
=
+
2
proc
+ ...
Erwin Verwichte - Sun Observation - ASSSP12
33
Aschwanden et al. (2000),
Ding & Nakariakov (2012)
2
comp
5) Processing noise: e.g. de-spiking to remove cosmic-ray hits (sometimes
better not to do).
4) Compression noise: to save bandwidth in storing and sending images to
Earth, JPEG compression is used.
3) Digitization noise: the signal is converted to an integer (range of power of
two), this introduced an uncertainty of σ = 0.5 DN.
2) Read-out noise: in a CCD the signal per pixel (photon-induced electrons)
are collected in rows towards one direction.
1) Dark current noise: even if the CCD is un-exposed, the finite temperature
induces spurious signals. That is why CCD are cooled. σ = σ(TCCD).
2
32
Erwin Verwichte - Sun Observation - ASSSP12
The processing of the measured light by the electronics is a source of noise.
Instrumental noise
Wednesday, 5 September 12
Observation
Erwin Verwichte - Sun Observation - ASSSP12
35
Extreme UltraViolet Imager (EUVI)
Selection of the best data-set for the problem.
What temporal, spatial, spectral resolution is there?
Typical data-analysis chain
Wednesday, 5 September 12
34
Erwin Verwichte - Sun Observation - ASSSP12
measurement of rotation of the Sun using coronal bright points seen by EUVI
on board STEREO.
We look at an example study of using EIV imaging data:
Example of an observational study
Enhancement of the feature or event to detect,
e.g. background subtraction, running-difference,
gradient filters.
Filtering
Filtering
Wednesday, 5 September 12
35
Detection of the feature or event. This can be done
in an automated way (objective and scaleable) or
by hand (mouse-clicking and tennis arms).
Enhancement of the feature or event to detect,
e.g. background subtraction, running-difference,
gradient filters.
Extraction
(Segmentation)
Selection of the best data-set for the problem.
What temporal, spatial, spectral resolution is there?
Observation
Typical data-analysis chain
Erwin Verwichte - Sun Observation - ASSSP12
Selection of the best data-set for the problem.
What temporal, spatial, spectral resolution is there?
Wednesday, 5 September 12
35
Erwin Verwichte - Sun Observation - ASSSP12
Observation
Typical data-analysis chain
Enhancement of the feature or event to detect,
e.g. background subtraction, running-difference,
gradient filters.
Filtering
35
Wednesday, 5 September 12
Characterisation of the results: connect to physical
quantities, statistical trends and theoretical models.
Characterisation
(Data-mining)
Filtering
Detection of the feature or event. This can be done
in an automated way (objective and scaleable) or
by hand (mouse-clicking and tennis arms).
Enhancement of the feature or event to detect,
e.g. background subtraction, running-difference,
gradient filters.
Extraction
(Segmentation)
Selection of the best data-set for the problem.
What temporal, spatial, spectral resolution is there?
Erwin Verwichte - Sun Observation - ASSSP12
Observation
Typical data-analysis chain
Wednesday, 5 September 12
Detection of the feature or event. This can be done
in an automated way (objective and scaleable) or
by hand (mouse-clicking and tennis arms).
Selection of the best data-set for the problem.
What temporal, spatial, spectral resolution is there?
Extraction
(Segmentation)
35
Erwin Verwichte - Sun Observation - ASSSP12
Observation
Typical data-analysis chain
35
Erwin Verwichte - Sun Observation - ASSSP12
Wednesday, 5 September 12
Bright points are nice local point-like features (mini active regions) that
live long enough and can be used as tracers of the rotation of the Sun.
We select a month worth of data from STEREO-A and STEREO-B during solar
minimum when there are not many active regions and bright points are clear.
Solar rotation using bright points
36
Characterisation of the results: connect to physical
quantities, statistical trends and theoretical models.
Characterisation
(Data-mining)
Wednesday, 5 September 12
Detection of the feature or event. This can be done
in an automated way (objective and scaleable) or
by hand (mouse-clicking and tennis arms).
Enhancement of the feature or event to detect,
e.g. background subtraction, running-difference,
gradient filters.
Filtering
Extraction
(Segmentation)
Selection of the best data-set for the problem.
What temporal, spatial, spectral resolution is there?
Erwin Verwichte - Sun Observation - ASSSP12
Observation
Typical data-analysis chain
1
+1
Z
I(~r 0 )
✓
~r 0
a
~r
◆
d2~r 0
Wednesday, 5 September 12
An original image
r2
2
◆
38
Erwin Verwichte - Sun Observation - ASSSP12
r2 exp
Wavelet transform at four scales
Wavelet filtering to enhance bright points
Wednesday, 5 September 12
(~r) =
✓
The transform enhances anything that looks like the wavelet at a certain
scale a. Here we take the Mexican Hat wavelet, which is of the form:
which is basically a 2d convolution of the image with Ѱ(x,y), which is the
mother wavelet with the special characteristic that its average is zero.
1
Wa (I)(~r) = 2
a
37
Erwin Verwichte - Sun Observation - ASSSP12
We filter the EUVI images using a two-dimensional wavelet transform:
Wavelet filtering to enhance bright points
~r ~r0
p
a2 +
2
◆
Wednesday, 5 September 12
40
Erwin Verwichte - Sun Observation - ASSSP12
Bright points average position is extracted by tresholding the wavelet
transform at the optimum scale a. But not all detection are correct
(remove cosmics)!
Detection of bright-points
Wednesday, 5 September 12
I(~r) = I0 + ax + by
Also, the wavelet transform removes any background trend (as perceived
at scale a) since the transform of the following is zero!
For a = σ we find the maximum wavelet amplitude. So we can find the scale
of the bright point easily by scanning in a. Also, the transform is still centered
on r0, so we have no shifts (Gaussian blurring can induce such).
2
✓
we find the wavelet back again but with an amplitude that depends on a
a2
Wa (I)(~r) = 2⇡A 2
a +
39
Erwin Verwichte - Sun Observation - ASSSP12
If you take the wavelet transform of a bright-point, modeled as
a 2d Gaussian:
✓
◆
(~r ~r0 )2
I(~r) = A exp
2 2
Wavelet filtering to enhance bright points
⌦ = A + B sin2 + C sin4
Wednesday, 5 September 12
⌦ = A + B sin2 + C sin4
Bright point rotation (not height corrected)
42
Erwin Verwichte - Sun Observation - ASSSP12
And using both STEREO’s we can find the height of the bright-points in the
solar atmosphere!
Wednesday, 5 September 12
41
Erwin Verwichte - Sun Observation - ASSSP12
Looking at extracted bright-points from different times, their movement on
the disk, and hence their siderical rotation rate can be determined!
Detection of bright-points
Wednesday, 5 September 12
EUVI properties
Full disk images (304Å, 171Å, 195Å, 284Å)
1.6 arcsec CCD pixel size
Time cadence of 2.5 minutes
A 10 min oscillation is visible in
the southern loop arcade of an
active region on the East limb.
•
•
•
42
43
43
Erwin Verwichte - Sun Observation - ASSSP12
We consider an observation by STEREO/EUVI of a
transverse loop oscillation on 26/06/2007 from two
vantage points [Verwichte et al. 2009].
Another example of observational study
Wednesday, 5 September 12
EUVI properties
Full disk images (304Å, 171Å, 195Å, 284Å)
1.6 arcsec CCD pixel size
Time cadence of 2.5 minutes
A 10 min oscillation is visible in
the southern loop arcade of an
active region on the East limb.
•
•
•
42
Erwin Verwichte - Sun Observation - ASSSP12
We consider an observation by STEREO/EUVI of a
transverse loop oscillation on 26/06/2007 from two
vantage points [Verwichte et al. 2009].
Another example of observational study
Wednesday, 5 September 12
Suggests observed mode is horizontally polarised fundamental kink mode.
With the geometry known we compare simulated oscillations for various
polarisations and harmonics with the data.
Another example of observational study
42
44
44
Erwin Verwichte - Sun Observation - ASSSP12
Suggests observed mode is horizontally polarised fundamental kink mode.
Wednesday, 5 September 12
42
Erwin Verwichte - Sun Observation - ASSSP12
With the geometry known we compare simulated oscillations for various
polarisations and harmonics with the data.
Another example of observational study
Wednesday, 5 September 12
This can be done between two instantaneous views (e.g. STEREO-AIA) or
between two times (dynamic stereoscopy).
Verwichte et al. 2009, 2010
42
45
46
Erwin Verwichte - Sun Observation - ASSSP12
One may add a model to aid 3d reconstruction. For instance one may
assume that a loop is planar (all points in the same plane). Then one adjusts
the inclination of the loop plane to find the best match between the two
views. This works well for large separation angles between views, but is
always approximate.
Stereoscopy
Wednesday, 5 September 12
Aschwanden, 2009
This does not give a unique solution as we need 3 non co-planar views to
have all information. So with STEREO it works well except when structures
(such as loop) are oriented in the epipolar plane of STEREO-A, B and Sun.
Inhester 2006
42
Erwin Verwichte - Sun Observation - ASSSP12
Stereoscopy means using two view points to reconstruct a 3d picture of a
structure if the angle between the views is not too large.
Stereoscopy
Wednesday, 5 September 12
EUVI-A 171
AIA 171
~90◦ view separation
with STEREO is useful!
42
Example of a vertical TLO
47
Erwin Verwichte - Sun Observation - ASSSP12
The full sun, high-cadence synoptic, multi-wavelength view from AIA/SDO
provides rich pickings in oscillation events (>200 events!).
The combination with STEREO allows for a good estimate of 3d structure.
Another example of observational study