This is an electronic newsletter edited by the IUE Data
Analysis Center (IUEDAC)
intended to inform the numerous remote and local users of the IUEDAC
software of recent software updates and project changes.
IUEDAC Software and Hardware Updates
Recent Software Changes (since 13-Dec-1995)
A routine called MAN is now available for graphics terminal
users for getting help on intrinsic IDL commands. For
example, typing MAN,'SMOOTH' will display information on the
IDL SMOOTH command.
The IUE merged log has been updated. All Goddard entries
should now be included as well as NEWSIPS status entries through
12/5/95. The new version also includes galactic coordinates which
can be accessed via the "expert mode" database
A new version of MGEX was implemented which corrects problems
when individual spectra are extracted from multiple exposure
NEWSIPS SILO files.
The IUEDAC staff continues to prepare the DAC analysis software
for the arrival of the NEWSIPS high dispersion data. Efforts
thus far have concentrated on the basic extraction and display
utilities. Are there particular capabilities and/or software that
you would like to see become part of the DAC software package?
This is your chance to send us your input. Send correspondence to
Recently we discovered a small error in the processing history portion
of the FITS headers for some images. The error affects only LWP
low-dispersion data taken after Nov. 12, 1992 (i.e. data that may be
affected by solar scattered light). There is a message in the processing
history under RAW_SCREEN that states that, except for trailed and
multiple exposure images, the spectral extraction is set to point
source, in order to minimize any possible scattered light contamination.
This is correct. In some cases, the RAW_SCREEN processing history also
says that the extraction was set to "extended source"; this is
incorrect. The affected images are being reprocessed to remove this
possible source of confusion.
The IPC has now processed all of the available NASA SWP and LWP low
dispersion data. Data which are still proprietary have not yet been
processed through NEWSIPS, but this will be done as data become publicly
accessible. Some images have not been processed because of missing
database information or other problems. We are currently working on
getting those images taken care of and processed. Note also that we are
not processing VILSPA data; VILSPA is doing so.
If you need images that seem unavailable, please contact Cathy Imhoff
. She can check on them and in many cases fix the problem so that the
images can be processed.
NEWSIPS Data Availability
Recently NSSDC has placed a large volume of NEWSIPS data into the NDADS
system (see below). If you have been trying to get some images which are
listed as processed but not yet available from NDADS, you may wish to try
IUE and Education
Are you interested in using the IUE satellite or its data to support
science education? We are too! We have recently formed a small
informal working group to exchange ideas and experiences. If you are
interested in joining, please contact Cathy Imhoff.
Much of the SWP and LWP low dispersion Final Archive data are now
available to the community. The reaction to the NEWSIPS low dispersion
processing techniques has been very positive. The high dispersion
Final Archive processing system is in final testing for the SWP images.
The processing system should be implemented in a few weeks. The SWP
ripple correction and absolute flux calibration are being derived by
our collaborators at VILSPA. Processing of SWP high dispersion images
will begin when these calibrations are received and operational testing
of the software is complete. There will be several new data products
for the NEWSIPS high dispersion processing system. The SIHI image
is a rotated and geometrically rectified image. Individual orders
can be custom-extracted easily from this image, which includes
wavelength calibration for the horizontal echelle orders. Another
output product of the high dispersion NEWSIPS system is the COSMIC RAY
image, which is created by masking out the spectral orders and
flagging the deviant background pixels, allowing visual identification
of grazing incidence cosmic ray events.
Re-Extracting Certain NEWSIPS Data
The NEWSIPS extraction technique for low dispersion was developed to
handle all of the images in the archive in a fully automated way.
However, some spectra are not properly extracted with the automated
extraction and should be re-extracted using an interactive
version of the optimal extraction method or an alternate extraction method.
Below are 2 examples of situations where a re-extraction of the
NEWSIPS data is required. Investigators are advised that the SILO image
should be examined for every image used in their analysis to fully
understand the data. Only examining the MXLO files can be misleading.
1) Multiple spectra
This situation occurs when two or more spectra are recorded through the
large aperture, or when the aperture is filled with an extended source
with a continuum point source also in the aperture (e.g., PN with
central star). The IUESIPS extraction technique was a simple boxcar
extraction. If multiple spectra were present in the large aperture,
the flux from all spectra was summed for the IUESIPS extraction.
The extraction technique used in NEWSIPS is based on the optimal
extraction technique (Kinney et al. 1991). The NEWSIPS code assumes
the cross-dispersion profile varies slowly with wavelength and that
only one spectrum is present in the large aperture. In the case of
multiple spectra, the NEWSIPS extraction will center on the strongest
spectrum. If the additional spectrum or full aperture illumination from
an extended source includes strong emission lines, the NEWSIPS extraction
will possibly flag the emission lines as deviant pixels based on the
statistics of the noise model and the deterimined profile. The flux
from the pixels in the emission lines flagged as cosmic ray hits
(deviant pixels) is not included in the extraction of the spectral flux.
Therefore, the total flux in the emission lines can be less with NEWSIPS
than with IUESIPS if a strong continuum source is also present
in the aperture.
For this situation, the investigator should use an alternate extraction
technique, depending on the goals of the science. A boxcar extraction
would yield the total flux in the aperture in the case of an extended
source and a point source. If multiple point source spectra are present,
they can be extracted individually with an interactive version of
optimal extraction or a gaussian extraction.
The fact that the NEWSIPS extraction requires that the cross-dispersion
profile vary slowly with wavelength and that it cannot model
more than one spectrum in the large aperture at a time in its automated
mode, as well as the algorithm used for flagging deviant pixels for
exclusion from the flux calculation, are presented in the NEWSIPS Image
Processing Information Manual, Newsletter #53.
The NEWSIPS extraction method uses splines to fit the spectrum in
successive lines parallel to the dispersion. The fits are used to
create the array of cross-dispersion profiles used to weight the
extracted data. The number of nodes used in the spline fits is
determined from the signal-to-noise ratio, with 15 being the maximum
number of nodes for high signal-to-noise data, and 2 being the
minimum for low signal-to-noise data. We have found that when
2 nodes have been used in the spline fit, the spectrum may
not be well modeled by the fit and the resulting flux can be
less accurate than flux determined using an assumed profile, such
as a gaussian profile, the default optimal profile, or a boxcar
profile. Investigators are advised to use caution when using
NEWSIPS data that has been extracted with a 2-nose spline fit.
This information is recorded in the HISTORY portion of the FITS
label. It would be a good idea to re-extract such spectra with
an alternate extraction technique to verify the NEWSIPS results.
Questions concerning the NEWSIPS image processing techniques,
the NEWSIPS output products, and IUE calibrations can be directed
to Joy Nichols (firstname.lastname@example.org).
Availability of NEWSIPS Data Via NDADS
Most low dispersion SWP and LWP data processed by GSFC with NEWSIPS
through early January, 1996, are now available from NDADS. The status
flags accessed by the project status report marker are being updated
and should be up-to-date early in February. Some VILSPA SWP low
dispersion data are now available through NDADS and more will become
available over the next few weeks.
Some users attempting to retrieve NEWSIPS data from NDADS may have
been told that the data was proprietary. Due to a programming error,
some data was incorrectly flagged as proprietary. All of the erroneous
flags have now been reset.
The archive has received 19th episode data from VILSPA, but none of it
is yet available through NDADS. This data should become available
sometime in February. High dispersion data have been processed with
IUESIPS and the low dispersion data have been processed with NEWSIPS.
Please send questions concerning retrieval or availability of IUE data
from NDADS to Karen Levay.
Dr. Michael Carini has left the IUEDAC staff to assume an academic
position. Ms. Lyla Taylor has left the DAC to pursue a graduate degree.
The IUE project wishes both of them well and both will be missed.
Mr. James Caplinger joins the DAC staff. Many IUE users probably
know Jim as one of IUE's excellent group of Telescope Operators.
Dr. Richard Arquilla is the current IUEDAC Task Leader.