Processing AIA cutout data
This document goes through the steps I follow for processing
AIA cutout data. In general I convert an AIA image into an IDL
map for my work.
My own routines that I mention below are available at http://files.pyoung.org/idl/sdo/.
Step 1 - request and download level-1 cutout data from the
JSOC
Go to the page Procedure
for obtaining SDO cutouts from the JSOC service at the EIS Wiki for
information about this. Note that I do not use the 'register' option
when requesting cutouts so that I'm able to re-spike the cutouts if
necessary.
Step 2 - re-spike the cutouts (if necessary)Small-scale
dynamic features can be removed from AIA data by the de-spiking
routine, so it's often necessary to "re-spike" the data. Go to the page
Re-spiking
AIA images at the EIS Wiki for information about this. I save the
re-spiked images as FITS files that can be read with read_sdo.
Step 3 - convert to map
I have a routine called sdo2map.pro that reads an SDO FITS
file and converts it to a map. The
routine does not call aia_prep. It
does do the following things, though:
- it divides the data by the exposure time.
- removes badly saturated frames.
- rotates the data to correct for the image rotation angle
The call is:
IDL> map=sdo2map(filelist)
Step 4 - remove "sawtooth" effect
If you make a movie of a set of cutout frames, you'll see a
jerkiness
that's due to the pixelization of the frames. This is removed by doing
an interpolation of the frames using the IDL bilinear.pro routine. I
encode this in my routine sdo_align_map.pro:
IDL> newmap=sdo_align_map(map)
I also use this routine for pulling out a sub-field from the maps:
IDL> submap=sdo_align_map(map,xrange=xrange,yrange=yrange)
Why not use aia_prep?
The following reasons:
- the routine is slow for long cutout sequences (it's better to write out the prepped images as FITS files)
- it doesn't correct the images for the rotation angle
- when using /use_ref it gives each frame the same xcen,ycen
value so you lose the correct pointing information
An important feature of aia_prep, though, is that it checks the latest
pointing information for your file and updates it (if necessary).
However, it seems to be rare that the pointing information is incorrect.
Another feature is that aia_prep scales different filters to
the same plate scale. This is important if you plan to make multi-color
images/movies with one filter on top of another. I prefer to use
contours to do this sort of thing, and the map software handles this
without having to scale the images to the same plate scale.
Page maintained by Dr Peter R Young.
|