Spectral
observations on 2006 June 4 at NSO/SP
Welcome! Let’s work on spectrum.
For a better communication and cooperation, any suggestion are of appreciation.
Come often, to see what happen.
Updated on: 2007/02/27
Scientific
Goals:
· Study
quantitatively the Evershed flow at different heights
· Wilson
· Properties
of “moustache”
· ……
The
Sun on 2006/06/04
The
Sun on 2006/06/05, the observed AR was labeled NOAA 10892
Observations
We
observed the east-limb sunspot group (NOAA 10892) using the Horizontal
Spectrograph (HSG) at NSO/SP. A sample of the raw data is shown in the
following figure.
|
Observing Log on 2006
June 04, Sunday Seeing:
0.3 ~0.5 local with some upper level
problems Guider angle: 13.3 Slit: make heliocentric North-South along the
slit Length of slit (between hair
lines): 85” Slit width: 0.3” or 40 micron Step width: 0.2997” FOV(roughly): 85”
x 77” Time per slit position: 2.5
s Time per scan(roughly): 12.5
min (for 300 steps) Solar Field Imaged on Slit: Image size on slit jaw is
22.514mm (horizontal and vertical) 7.51” per mm 169.2” total 1” = 133.5 μm =
0.1335mm 30 micron slit = 0.225 arcsec 40 micron slit = 0.3”
|
Data Reduction
Step 1:
·
Look at CCD4
slitjaw images.
·
Make movie for each
scan. See how the slit scans the
solar surface.
·
From each movie,
compute a displacement map that will be used later to assemble the images.
·
also need the
target frames
Results from Step 1:
Data files
Information:
|
Raw data directory: |
'/rawdata/sp2006/spectra/Jun0406/ccd4/’ |
|
Total 7 scans (6 scan NOAA10892): |
|
|
060604.082548.0.ccd4.Kodak.gband.slitjaw.fits |
scan #0 (NOAA 10892), 280 steps, South 6.7, East 72.0, seeing 0.3 |
|
060604.084219.0.ccd4.Kodak.gband.slitjaw.fits |
scan #1 (NOAA 10892), 300 steps, South 6.7, East 71.8 |
|
060604.085726.0.ccd4.Kodak.gband.slitjaw.fits
|
scan #2 (NOAA 10892), 300 steps, (good scan, maybe the best) |
|
060604.091233.0.ccd4.Kodak.gband.slitjaw.fits |
scan #3 (NOAA 10892), 300 steps, |
|
060604.092239.0.ccd4.Kodak.gband.slitjaw.fits |
scan #4 (NOAA 10892), 300 steps, (good scan) |
|
060604.093245.0.ccd4.Kodak.gband.slitjaw.fits |
scan #5 (NOAA 10892), 300 steps, |
|
Calibration data (problematic) |
|
|
060604.094729.0.ccd4.Kodak.gband.slitjaw.flat.fits |
flat with hair lines in, in focus, |
|
060604.095139.0.ccd4.Kodak.gband.slitjaw.flat.fits |
same as above but out of focus with slit at home position |
|
060604.095806.0.ccd4.Kodak.gband.slitjaw.dark.fits |
dark (for all cameras, 64 exposures) |
|
060604.100458.0.ccd4.Kodak.gband.slitjaw.grid.fits |
0.5 mm grid for all cameras, 1.88 arcsec/0.5mm for all cameras. |
|
060604.100721.0.ccd4.Kodak.gband.slitjaw.grid.fits |
0.5 mm grid with secondary positions – rotated 90 degrees. |
|
060604.101012.0.ccd4.Kodak.gband.slitjaw.target.fits |
32 frames, Settings are same as above. |
|
1 scan |
|
|
060604.102232.0.ccd4.Kodak.gband.slitjaw.fits |
scan #6 (Sun center, quiet Sun), 274 steps, North 0.0, West 0.1, guider 13.2. Seeing at start is fair but become bad from the middle of the scan. See possibly saturation at suncenter (because using limb expo. times), especially for Na. The above flats taken at suncenter may also saturate, need to retake. |
|
Calibration data (taken in Sun center) |
|
|
060604.104149.0.ccd4.Kodak.gband.slitjaw.flat2.fits |
Retake flat. Same as above flat but use reduced expo. times (see following expo. times) |
|
060604.104736.0.ccd4.Kodak.gband.slitjaw.dark2.fits |
dark for the above new flat expo. times |
|
|
|
|
060604.105324.0.ccd4.Kodak.gband.slitjaw.flat.fits |
flat, fail |
|
Calibration data (taken near limb AR) |
|
|
060604.105548.0.ccd4.Kodak.gband.slitjaw.flat.fits |
flat using original expo. times but mapping just off the limb AR, no saturation. This flat will compensate on Doppler shift. |
|
060604.110254.0.ccd4.Kodak.gband.slitjaw.dark.fits |
dark for the above flat using original expo. time. |
Exposure times:
Original
at limb Reduced
at SC FOV
CCD1
(SI805-205 Halpha 6563) 80
ms 40 1024
x 1024
CCD2
(SI805-206 Fe 6302) 100 60 1024
x 1024
CCD3
(SI805-170 NaD 5889) 200 160 1024
x 1024
CCD4 (Kodak
1.6i slitjaw G-band) 20 10 1536
x 1032
A sample movie of scan #5 (click
on the image)
Step 2:
· Calibrate the spectra (dark, flat field correction;
spectrum rotation and curvature correction).
· “flatspectrum_na.pro” do this job (modified a little bit based on “flatspectrum_practise.pro”).
Results from Step 2:
“spectrum_halfsize.sav” This is a sample of calibrated Hα spectra (scan
#2, the best scan).
The following image is constructed by the spectrum
data cube.
Click on the image to see how the Sun changes face at
different wavelengths.
The following two images show the comparison of
constructed solar images by raw and calibrated spectrum data cube.
Raw data Calibrated
data
Problem:
There are horizontal lines across the constructed
images. They become more distinct after calibrationL.
These horizontal lines might be due to dust particles
on the slit. Try a data set that is closest in time to the flat fields (scan #5).
In this case, if there are some dusts, they shouldn't move too much.
The following image groups show the result of calibration experiment.
Calibration Experiment:
1. scan #2 calibrated by flat2
2. scan #2 calibrated by flat in focus (problematic flat, this
flat may be saturated)
3. scan #2 calibrated by flat out of focus (problematic
flat, this flat may be saturated)
4. scan #5 calibrated by flat out of focus (problematic
flat, this flat may be saturated)
5. scan #5 calibrated by flat in focus (problematic flat, this
flat may be saturated)
6. scan #5 calibrated by flat2
Scan
#2 raw data at 100th column
Calibrated
data by flat2
Calibrated data by problematic flat in focus Calibrated data by
problematic flat
out of focus
Scan
#2 raw data at 500th column
Calibrated
data by flat2
Calibrated data by problematic flat in focus Calibrated data by
problematic flat
out of focus
Scan
#5 raw data at 100th column
Calibrated
data by flat2
Calibrated data by problematic flat in focus Calibrated data by
problematic flat
out of focus
Scan
#5 raw data at 500th column
Calibrated
data by flat2
Calibrated data by problematic flat in focus Calibrated data by
problematic flat
out of focus
Conclusion:
· Change from scan #2 to #5 (the nearest scan in time to
the flat field) does not show apparent improvement.
· Calibrated by “Flat in focus” seems diminish
the horizontal lines most for all the cases.
· Scan #5 looks better than scan #2.
Hairline tilt for all images and how it changes with time:
(Hairline tile, i.e., “angle1” in the
calibration program, should not change, therefore we use the same angle1
obtained from the first image to correct for all the 300 images. While the
following figure shows that “angle1” calculated from each image are
not the same.)
Step 3:
· Need to get the filter curves for the interference
filters first. Otherwise we don't have a flat continuum part for the spectra.
·
Remove the
prefilter curve so that the continuum is flat.
·
We will still
have to consider later on the different noise levels in the analysis of the
spectra. In the line wings/continuum, we received less light than we would have
with a wider prefilter.
Results from Step 3:
The following are .sav files that contain calibrated
spectrum data cube.
Flat field and Scan #6 can be used to obtain the prefilter
curves.
|
|
Scan #5 AR |
Scan #6 QS |
Flat field |
|
CCD1 Halpha |
|||
|
CCD2 Fe |
|||
|
CCD3 Na |
Step 4:
· Send Carsten a sample of calibrated spectrum.
· Carsten test the routines to compute bi-sectors and
other stuff from the data.
|
Observing Log on 2006
June 11 (for
Angelo) Weather: Good, no clouds,
crystal clear sky. Filter: Halpha
Long scan: change
from original 120-840 (a little off) to 140-860 (looks good), units 4 Take
flat field Short scan: 360-680, units 4, expo. time 40 ms Take
target: 10 air force target “calib/target1.fts”
10 with grid target “calib/target2.fts” Take
dark: 100 VIM frames “calib/dark.fts” Take
data (move FOV to a QS region near suncenter) VIM scan 60 speckle
60 (Wrong
setting for VIM, frame per z step should be 5, while we set it to 1,
therefore VIM goes much quicker than speckle. Speckle was at 40% when VIM was
done. Restart VIM with 35 scans (set frame per z step to 5) Take
another data set VIM scan 60 speckle
60 Filter: Change Halpha filter to NaD (same QS region) Long scan:
340-1020 Short scan: change from yesterday’s 500-840 to 520-860 Began
to take data: VIM: unit 4, frames per z step 5, 60 scans Speckle:
60 simultaneously. Take
another data set (change from 60 to 30 sequences): When VIM done, speckle (forgot to change, still 60)
was at 65%. We retake VIM (set to 20 scans) Take
target again at this position: Air
force target first VIM:
same setting, target 10 speckle:
target 3 Grid
target VIM:
same setting. target 10 speckle:
target 4 Take
dark frame: VIM:
dark 100 speckle: Take
flat (move telescope to suncenter QS without spot): VIM:
520-860, unit 4, frame per z step 1. Filter: change filter to Fe 6302 (try to calibrate liquid crystal and
take magnetogram) Long scan: 630-1300, unit 10, frame per z step 1, 1 scan Take
scan test |