Документ взят из кэша поисковой машины. Адрес оригинального документа : http://xmm.vilspa.esa.es/sas/5.4.1/doc/edetect_chain.ps.gz Дата изменения: Fri Jan 10 23:37:28 2003 Дата индексирования: Tue Oct 2 11:59:14 2012 Кодировка: Поисковые слова: interacting galaxies

XMM-Newton Science Analysis System Page: 1
edetect chain
January 10, 2003
Abstract
The task edetect chain performs source detection on a user specied set of EPIC
images using the tasks eexpmap, emask, eboxdetect, esplinemap, emldetect,
and esensmap.
1 Instruments/Modes
Instrument Mode
EPIC MOS: IMAGING
EPIC PN: IMAGING
2 Use
pipeline processing yes
interactive analysis yes
3 Description
This chain script runs the SAS tasks eexpmap, emask, eboxdetect (local mode), esplinemap,
eboxdetect (map mode), emldetect, esensmap in sequence.
3.1 Overview on detection chain constituent tasks
eexpmap:
The task eexpmap uses the attitude le, the CAL vignetting information, and the exposure and bad
pixel extensions of the photon event les to calculate an exposure map corresponding to the input image.
emask:
Creates a detector mask (values 0 or 1) based on exposure or exposure gradient cuto values. The
detection tasks optionally use the mask to limit the detection to areas where the mask is 1.
xmmsas 20030110 1802-5.4.1

XMM-Newton Science Analysis System Page: 2
eboxdetect (local mode):
In local mode, (parameter usemap="no"), eboxdetect uses a 5x5 (or 3x3) pixel box and a surrounding
(+2 pixels) background area to search for signicant sources simultaneously in all input images.
esplinemap:
esplinemap uses the source list from eboxdetect (local mode) to remove sources from the input images
and creates smooth background maps by tting a 2-D spline to the images. The parameter nsplinenodes
determines the number of nodes per dimension. With nsplinenodes=16 esplinemap is able to model
the approximate shape of large extended objects, this is necessary for the reliable detection of point
sources. For the detection of extended objects (e.g. clusters of galaxies) it is recommended to use a
smoother background map (nsplinenodes=10-12).
From version 3.2 on esplinemap can read an EPIC PN event set and calculate the background cuased by
out-of-time events (visible as streaks on both sides of bright sourcces in readout directions). This works
only (and is only necessary) for PN images. This feature is switched on with the esplinemap parameter
withootset=yes, parameter ooteventset species the name of the event set. The event set should have
the same selections in TIME and FLAG as the esplinemap input image.
eboxdetect (map mode):
In map mode, (parameter usemap="yes"), eboxdetect uses a 5x5 (or 3x3) pixel box and the values from
the background map to search for signicant sources simultaneously in all input images.
emldetect:
This task uses the output list from eboxdetect and determines parameters for each input source by
means of a maximum likehood t to the input images. The following edetect chain parameters are
passed to emldetect:
likemin: Likelihood threshold (emldetect parameter mlmin).
ecf: Energy conversion factors for each input image (unit: 10 11 cts cm 2 erg 1 ).
eml list: Output source list name.
eml fitextent: Determine extent of sources?
eml determineerrors: Determine statistical errors of source parameters?
eml nmaxfit: Maximum number of sources to be t simultaneously.
eml nmulsou: Maximum number of model sources per input source position.
eml withsourcemap: Generate a source map (background image + model sources)?
eml withhotpixelfilter: Suppress detection of hot pixels?
eml usecalpsf: Use PSF model from CAL? Otherwise use analytical PSF.
Please note that the using the options eml fitextent, eml nmaxfit, and eml nmulsou can lead to
signicant increases in computation times. It is not advisable to use extent tting with a full set of 5
images, instead run extent tting on only 1 or 2 input images. Using the emldetect internal analytical
PSF (eml usecalpsf=no) saves computation time when performing extent tting.
esensmap:
For each input image esensmap calculates a sensitivity map in units of counts/sec for a likelihood
threshold given by the parameter esen mlmin.
xmmsas 20030110 1802-5.4.1

XMM-Newton Science Analysis System Page: 3
3.2 Use of edetect chain
The user can specify an arbitrary combination of images from dierent energy bands and dierent EPIC
instruments. All images must have identical binning and WCS keywords. Up to 3 instruments (MOS1,
MOS2, PN) with up to 5 images in dierent energy bands can be processed simultaneously (but beware
of the memory limitations of your machine).
The detection tasks eexpmap,eboxdetect , and emldetect require the energy boundaries of the input
images as input parameters pimin and pimax. edetect chain checks whether the number of input values
for pimin and pimax matches the number input images. Otherwise edetect chain will be terminated
with an error.
The most important parameters of the detection tasks are accepted as input parameters by edetect chain
and passed on to the tasks. Nevertheless edetect chain can be run with a relatively small set of
mandatory parameters, leaving the parameters of the constituent tasks at their default values.
If the exposure maps have already been created in an earlier run of edetect chain (or the PPS pro-
duced exposures maps have been renamed to the edetect chain standard le names), the calculation of
exposure maps can be omitted by setting witheexpmap="no".
If the event lists corresponding to the images are not available, a dummy EPIC le (e.g. the image itself)
can be given as parameter eventsets. This will lead to a less accurate calculation of the exposure maps.
3.3 Examples
1. Simultaneous detection run on the 5 standard energy bands of the MOS1 detector:
edetect_chain imagesets=''M1IMAGE_1000.FIT M1IMAGE_2000.FIT M1IMAGE_3000.FIT
M1IMAGE_4000.FIT M1IMAGE_5000.FIT''
eventsets=M1EVLI.FIT
attitudeset=atthk.dat
pimin=''200 500 2000 4500 7500''
pimax=''500 2000 4500 7500 12000''
ecf=''1.8035 1.9872 0.75557 0.26502 0.028794''
Note: Here we use energy conversion factors for the thin lter (see
http://xmmssc-www.le.ac.uk/pubdocs/SSC-LUX-TM-0059.ps.gz
for other ecf values). The default value of the ecf parameter is 1.0 (in units of 10 11 counts cm 2 erg 1 ),
using the default value will lead to incorrect values in the FLUX columns of the eboxdetect and emldetect
output source lists.
2. Simultaneous detection run on energy bands 2 and 3 and all EPIC detectors, use event set PNEVLI.FIT
to calculate background due to out-of-time events:
edetect_chain imagesets=''M1IMAGE_2000.FIT M1IMAGE_3000.FIT
M2IMAGE_2000.FIT M2IMAGE_3000.FIT
PNIMAGE_2000.FIT PNIMAGE_3000.FIT''
eventsets=''M1EVLI.FIT M2EVLI.FIT PNEVLI.FIT''
attitudeset=atthk.dat
pimin=''500 2000 500 2000 500 2000''
pimax=''2000 4500 2000 4500 2000 4500''
xmmsas\_20030110\_1802-5.4.1

XMM-Newton Science Analysis System Page: 4
esp_withootset=yes
esp_ooteventset=PNEVLI.FIT
Note: The calculation of the background due to out-of-time events is only used for EPIC PN data, the
events set given in esp ooteventset has to be an EPIC PN event list and should have the same TIME
and FLAG selections as the EPIC PN input images.
3. Simultaneous detection run on PN images 2 and 3, using emldetect to determine the extent of the
sources:
edetect_chain imagesets=''PNIMAGE_2000.FIT PNIMAGE_3000.FIT''
eventsets=PNEVLI.FIT
attitudeset=atthk.dat
eml_fitextent=yes
pimin=''500 2000''
pimax=''2000 4500''
esp_nsplinenodes=12
eml_fitextent=yes
eml_ecut=0.93
usecalpsf=no
Note: The parameter eml ecut determines the cutout radius of the sub-image used for the ML t.
eml ecut is given as encircled energy fraction (EEF) of the EPIC PSF. For an on-axis source eml ecut=0.93
corresponds to a cutout radius of  45 arcsec. The use of the analytical PSF (usecalpsf="no" will save
computing time when performing extent tting.
4 Parameters
This section documents the parameters recognized by this task (if any).
Parameter Mand Type Default Constraints
attitudeset yes lename attitude.ts
Name of attitude le
eventsets yes lename
list
events.ts
Names of event list les
imagesets yes lenamelist image.ts
Names of input images
pimin yes integer
list
500 0,20000
Lower energy boundary in PI channels
pimax yes integer
list
4500 0,20000
Upper energy boundary in PI channels
likemin no oat 10. 2.,1000.
xmmsas 20030110 1802-5.4.1

XMM-Newton Science Analysis System Page: 5
emldetect: detection likelihood threshold
witheexpmap no boolean true
create new exposure maps?
ecf no oat 1. 0.,1000.
energy conversion factors
eex attrebin no oat 4.0 0.0,60.0
eexpmap: Positional accuracy of attitude rebinning [arcsec]
eex usedss no boolean true
eexpmap: use DSS info from input image
eex withpatterns no boolean true
eexpmap: User selected event patterns?
eex pattern no string \ \
eexpmap: Event Pattern types
emask threshold1 no oat 0.2 0.0,1.0
emask: threshold1 (fraction of maximum)
emask threshold2 no oat 0.5 0.0,10.0
emask: threshold2 (gradient)
eboxl list no lename eboxlist l.ts
Name of eboxdetect local mode source list
eboxm list no lename eboxlist m.ts
Name of eboxdetect map mode source list
eboxl likemin no oat 6. 1.0,50.0
eboxdetect: local detect minimum detection likelihood
eboxm likemin no oat 6. 1.0,50.0
eboxdetect: map detect minimum detection likelihood
ebox withdetmask no boolean true
eboxdetect: detection mask ag
ebox withexpimage no boolean true
eboxdetect: exposure image ag
ebox boxsize no integer 5 3,5
eboxdetect: detection box size: permitted values: 3 or 5
esp withcheese no boolean false
esplinemap: ag for output of cheesed image
esp withdetmask no boolean false
esplinemap: ag to use detection mask
esp nsplinenodes no integer 16 5,20
esplinemap: number of spline nodes
xmmsas 20030110 1802-5.4.1

XMM-Newton Science Analysis System Page: 6
esp ntrun no integer 3 1,5
esplinemap: number of iterations
esp scut no oat 0.005 0.0,10.0
esplinemap:source cut-out ux level [counts/arcsec 2 ]
esp excesssigma no oat 4.0 1.0,6.0
esplinemap: threshold for excess with respect to spline
esp withexpimage no boolean false
esplinemap: ag to use exposure map
esp withootset no boolean false
esplinemap: ag to use out-of-time event set
esp ooteventset no lename events.ts
EPN out-of-time event set
eml list no lename emllist.ts
emldetect: source list
eml textent no boolean false
emldetect:t source extent
eml ecut no oat 0.68
emldetect: events cut-out radius
eml scut no oat 0.9
emldetect: source cut-out radius for multiple source ts
eml determineerrors no boolean false
emldetect: determine stat errors
eml nmaxt no integer 1 1,6
emldetect: max number of sources for multi-source ts
eml nmulsou no integer 1 1,3
emldetect: max number of sources per input source position
eml withsourcemap no boolean false
emldetect: creation of output source map
eml withxidband no boolean false
emldetect: add xid band output
eml withhotpixellter no boolean false
emldetect: use hot pixel lter
eml withdetmask no boolean false
emldetect: use detector mask(s)
esen mlmin no oat 10. 1,20
esensmap: upper limit likelihood
xmmsas 20030110 1802-5.4.1

XMM-Newton Science Analysis System Page: 7
5 Errors
This section documents warnings and errors generated by this task (if any). Note that warnings and
errors can also be generated in the SAS infrastructure libraries, in which case they would not be docu-
mented here. Refer to the index of all errors and warnings available in the HTML version of the SAS
documentation.
Parameter error (fatal)
imagesets,pimin,pimax must have same number of values
Parameter error (fatal)
maximum number of eventsets (3) exceeded
missingAttribute (fatal)
Attribute INSTRUME missing in le
missingAttribute (fatal)
Attribute INSTRUME missing in le
NoEvlist (warning)
The task eexpmap needs exposure and bad pixel information from the event list, otherwise
the exposure map will be inaccurate
corrective action: No event list for specied, no EXPOSURE information
6 Input Files
1. EPIC FITS images
2. EPIC event list les
3. attitude le
7 Output Files
1. exposure maps (from task eexpmap)
2. detector mask images (from task emask)
3. background maps (from task esplinemap)
4. eboxdetect source list (local mode)
5. eboxdetect source list (map mode)
6. emldetect source list
7. source maps, if eml withsourcemap="yes" (from task emldetect)
8. sensitivity images (from task esensmap)
xmmsas 20030110 1802-5.4.1

XMM-Newton Science Analysis System Page: 8
The task edetect chain uses the following hardcoded extensions to the input image names for the output
products:
Exposure maps: exp.
Mask images: mask.
Background images: bkg.
Source maps: smap.
Sensitivity images: sen.
The names of the output source lists from eboxdetect and emldetect can be specied via the parameters
eboxl list, eboxm list, and eml list.
8 Algorithm
see descriptions of individual tasks
9 Comments
10 Future developments
References
xmmsas 20030110 1802-5.4.1