THE BASICS OF REDUCING (new)
HIRES
DATA
Quick & Dirty method:
1. Run CCDPROC with
OVERSCAN & TRIM turned on, everything else off. Run it three
times -- once on each chip hiresdata.fits[Y], where Y =1,2,3.
This is easily done in a script with the command
ccdproc ("hires0001.fits[Y]",
output="hires0001.Y.fits", ccdtype=" ", fixpix=no, overscan=yes,
trim=yes, zerocor=no, darkcor=no, flatcor=no, illumcor=no,
fringecor=no, readcor=no,
scancor=no, biassec="[687:713,*]", trimsec="[5:686,*]")
This is for 3x1 binning. For 2x1 binning, use
biassec="[1031:1070,*]", trimsec="[7:1030,*]". If you do a
CCDLIST on all the images, they should have the [OT] tag, for
overscan-corrected and trimmed.
2. I now make directories for each chip, ./chipY.
Then I make bias and flat directories within each directory:
./chipY/bias and ./chipY/flats. Each chip now looks like this:
1 pixel = 15 microns =
0.12"
Chip 1 will have the highest orders (shortest wavelengths), and Chip3
will have the lowest (longest wavelengths). If you display these
images in DS9/SAOimage, you should flip XY and rotate 90 degrees (both
in the zoom menu) to make it look like old-school HIRES data, with the
dispersion increasing along the positive horizontal axis, and the
spatial direction increasing along the positive vertical axis.
Again, this chip is binned 3x1. For a 2x1 binning, the spatial
axis would go from 0:1023, and for 1x1 binning it would be from
0:2047. NOTE: the gaps between the chips are 4 pixels, or
60 microns, or 0.48". Use the HIRES
simulator to make sure the gaps don't fall on orders you
need!
This is what our chips look like (for He 2-10, a nearby dwarf
starburst), with orders marked. These were observed with the blue
setup and a 7" x 1.148" slit.
Chip 3:
(order 60: He I 5876A emission + Na D 5890/5896A absorption)
Chip 2:
(top down emission lines: [OIII] 5007A & 4959A, HBeta 4861A, HGamma
4341A)
Chip 1:
(top down emission lines: HDelta 4102A, HEpsilon 3971A, Ca II
3934/3968A,
Hwhatevercomesafterepsilon 3888, OII 3726/3729A)
3. In ./chipY/bias, use ZEROCOMBINE
(combine = average, reject = minmax, weight = none) to combine all the
bias frames. This makes zero.chipY.fits.
4. Use this bias frame in CCDPROC
to correct all the images. Turn off everything except for
ZEROCOR. I use a script to do this as well, with the command
ccdproc
("hires0020.Y.fits", output="", ccdtype=" ", fixpix=no, overscan=no, trim=no, zerocor=yes, darkcor=no,
flatcor=no, illumcor=no, fringecor=no, readcor=no, scancor=no, readaxis="line",
zero="zero.chipY.fits")
Now all of your files should show [OTZ] in CCDLIST.
5. Now it's time to flat-field your images.
a. Use FLATCOMBINE to combine your
flats. Use the one that looks the best for each chip -- the bluer
chip(s) may need a longer exposure than the redder chip(s). (For
me, 2 sec was good for chip 3, and 3 sec was good for chips 1 and
2).
b. Trace the orders using a trace star spectrum in APTRACE. I use "o" to change
the order #s so they match up with the actual order #s, instead of the
default.
c. Use APFLATTEN to
flatten the images (note: all these tasks are in the ECHELLE package),
using the trace star as your aperture reference image, and make
apflat.Y.fits for each chip Y. I used an 8th order Legendre
polynomial and was able to easily flatten to within 5%.
d. Run CCDPROC one last
time, with flatfielding on and everything else off, and using
apflat.Y.fits as your
6. Use APSUM to extract
the Thorium-Argon spectrum in each chip: thar.Y.ec.fits. Once
again, use the trace star as an aperture reference/profile image.
Extract the spectra in echelle format.
7. Identifiy lines in the arc lamp spectra using ECIDENTIFY. We used a 1-sec
exposure of Th-Ar 2 for all chips. My solutions fit a
4th order Legendre polynomial, and were good to 0.005 Angstroms.
Here are the solutions -- this will only work for our values of ECHANGL
= -0.00004228 and XDANGL = 1.24399996, but if should give you a good
idea of the dispersion in these orders.
thar.3.ec.fits:
ap = 59, w1 = 5990.187, w2 = 6096.485, dw = 0.025958, nw = 4096
thar.3.ec.fits: ap = 60, w1 = 5890.289, w2 = 5995.029, dw = 0.025578,
nw = 4096
thar.3.ec.fits: ap = 61, w1 = 5793.675, w2 = 5896.879, dw = 0.025202,
nw = 4096
thar.3.ec.fits: ap = 62, w1 = 5700.185, w2 = 5801.877, dw = 0.024833,
nw = 4096
thar.3.ec.fits: ap = 63, w1 = 5609.672, w2 = 5709.878, dw =
0.02447, nw = 4096
thar.3.ec.fits: ap = 64, w1 = 5521.994, w2 = 5620.741, dw = 0.024114,
nw = 4096
thar.3.ec.fits: ap = 65, w1 = 5437.021, w2 = 5534.339, dw = 0.023765,
nw = 4096
thar.3.ec.fits: ap = 66, w1 = 5354.629, w2 = 5450.549, dw = 0.023424,
nw = 4096
thar.3.ec.fits: ap = 67, w1 = 5274.7, w2 = 5369.255, dw
= 0.02309, nw = 4096
thar.3.ec.fits: ap = 68, w1 = 5197.124, w2 = 5290.349, dw = 0.022765,
nw = 4096
thar.2.ec.fits: ap = 70, w1 = 5048.787, w2 = 5139.431, dw = 0.022135,
nw = 4096
thar.2.ec.fits: ap = 71, w1 = 4977.678, w2 = 5067.077, dw = 0.021831,
nw = 4096
thar.2.ec.fits: ap = 72, w1 = 4908.548, w2 = 4996.73, dw =
0.021534, nw = 4096
thar.2.ec.fits: ap = 73, w1 = 4841.314, w2 = 4928.307, dw = 0.021244,
nw = 4096
thar.2.ec.fits: ap = 74, w1 = 4775.899, w2 = 4861.729, dw =
0.02096, nw = 4096
thar.2.ec.fits: ap = 75, w1 = 4712.231, w2 = 4796.925, dw = 0.020682,
nw = 4096
thar.2.ec.fits: ap = 76, w1 = 4650.24, w2 = 4733.822, dw =
0.020411, nw = 4096
thar.2.ec.fits: ap = 77, w1 = 4589.861, w2 = 4672.356, dw = 0.020145,
nw = 4096
thar.2.ec.fits: ap = 78, w1 = 4531.032, w2 = 4612.463, dw = 0.019885,
nw = 4096
thar.2.ec.fits: ap = 79, w1 = 4473.694, w2 = 4554.083, dw = 0.019631,
nw = 4096
thar.2.ec.fits: ap = 80, w1 = 4417.791, w2 = 4497.161, dw = 0.019382,
nw = 4096
thar.2.ec.fits: ap = 81, w1 = 4363.27, w2 = 4441.642, dw =
0.019138, nw = 4096
thar.2.ec.fits: ap = 82, w1 = 4310.081, w2 = 4387.475, dw =
0.0189, nw = 4096
thar.2.ec.fits: ap = 83, w1 = 4258.176, w2 = 4334.613, dw = 0.018666,
nw = 4096
thar.2.ec.fits: ap = 84, w1 = 4207.508, w2 = 4283.007, dw = 0.018437,
nw = 4096
thar.2.ec.fits: ap = 85, w1 = 4158.035, w2 = 4232.615, dw = 0.018213,
nw = 4096
thar.1.ec.fits: ap = 86, w1 = 4109.608, w2 = 4183.329, dw = 0.018003,
nw = 4096
thar.1.ec.fits: ap = 87, w1 = 4062.393, w2 = 4135.239, dw = 0.017789,
nw = 4096
thar.1.ec.fits: ap = 88, w1 = 4016.253, w2 = 4088.241, dw = 0.017579,
nw = 4096
thar.1.ec.fits: ap = 89, w1 = 3971.15, w2 = 4042.298, dw =
0.017374, nw = 4096
thar.1.ec.fits: ap = 90, w1 = 3927.05, w2 = 3997.375, dw =
0.017173, nw = 4096
thar.1.ec.fits: ap = 91, w1 = 3883.92, w2 = 3953.439, dw =
0.016977, nw = 4096
thar.1.ec.fits: ap = 92, w1 = 3841.727, w2 = 3910.456, dw = 0.016784,
nw = 4096
thar.1.ec.fits: ap = 93, w1 = 3800.443, w2 = 3868.398, dw = 0.016595,
nw = 4096
thar.1.ec.fits: ap = 94, w1 = 3760.037, w2 = 3827.234, dw = 0.016409,
nw = 4096
thar.1.ec.fits: ap = 95, w1 = 3720.482, w2 = 3786.935, dw = 0.016228,
nw = 4096
thar.1.ec.fits: ap = 96, w1 = 3681.752, w2 = 3747.476, dw =
0.01605, nw = 4096
thar.1.ec.fits: ap = 97, w1 = 3643.82, w2 = 3708.83, dw =
0.015875, nw = 4096
thar.1.ec.fits: ap = 98, w1 = 3606.663, w2 = 3670.972, dw = 0.015704,
nw = 4096
thar.1.ec.fits: ap = 99, w1 = 3570.257, w2 = 3633.878, dw = 0.015536,
nw = 4096
thar.1.ec.fits: ap = 100, w1 = 3534.579, w2 = 3597.526, dw = 0.015372,
nw = 4096
thar.1.ec.fits: ap = 101, w1 = 3499.607, w2 = 3561.894, dw =
0.01521, nw = 4096
thar.1.ec.fits: ap = 102, w1 = 3465.322, w2 = 3526.96, dw =
0.015052, nw = 4096
thar.1.ec.fits: ap = 103, w1 = 3431.702, w2 = 3492.704, dw = 0.014897,
nw = 4096
thar.1.ec.fits: ap = 104, w1 = 3398.729, w2 = 3459.107, dw = 0.014744,
nw = 4096
thar.1.ec.fits: ap = 105, w1 = 3366.383, w2 = 3426.15, dw =
0.014595, nw = 4096
thar.1.ec.fits: ap = 106, w1 = 3334.648, w2 = 3393.814, dw = 0.014448,
nw = 4096
thar.1.ec.fits: ap = 107, w1 = 3303.507, w2 = 3362.083, dw = 0.014304,
nw = 4096
thar.1.ec.fits: ap = 108, w1 = 3272.941, w2 = 3330.94, dw =
0.014163, nw = 4096
thar.1.ec.fits: ap = 109, w1 = 3242.937, w2 = 3300.368, dw = 0.014025,
nw = 4096
thar.1.ec.fits: ap = 110, w1 = 3213.478, w2 = 3270.352, dw = 0.013889,
nw = 4096
7. Combine object frames using IMCOMBINE,
with combine = average and reject = crreject (to get rid of cosmic
rays).
8. Use APEDIT to define
the apertures you want to extract from the object images. The
trace star can be used as an aperture reference image. Use 'b' in
the editing window to fix/redefine the background to be subtracted.
9. Use REFSPEC to define
the Th-Ar spectrum as the dispersion solution reference.
10. Use DOECSLIT to
extract the spectra, subtract the background, and apply the dispersion
solution all at once. You can also do this step-by-step using the
tasks in the ECHELLE package. Hint: in DOECSLIT you can get to
the extraction parameters by going down to sparams and typing ':e'.
11. Use SPLOT or SPECFIT to analyze your wicked cool
spectra!
KNOWN ISSUES:
i. The cross-disperser angle may
drift over the night. Especially after our run in January, when
there were known issues with the instrument, our cross-disperser angle
changed in a very detectable way (~6 pixels along the dispersion axis)
over the night. Make sure you take
arc lamp spectra before/after all important data frame sets!
Also, it's a good idea to check your Echelle grating &
cross-disperser angles throughout the night so you can fix them if
needed.
Created by Colleen Schwartz
[colleen-at-physics(dot)ucsb(dot)edu], 3.17.2005.
