|
This cookbook illustrates how to measure photometry for point sources from MIPS
70 micron images with uniform background. We show how to
use APEX, a photometry package provided by the
Spitzer Science Center. APEX is a part of a general package called
MOPEX: MOsaicker and Point source
EXtractor. We STRONGLY recommend that users check the
intermediate outputs, either images or tables, generated by APEX. This
allows you to decide if the APEX input parameters are appropriate for
your specific data set.
It is important to note that this tutorial is specifically written for
MIPS 70 micron data, which has unresolved point sources and uniform
background. The recommendations from this cookbook are applicable to
data sets similar to those from the Spitzer COSMic Origins Survey
(SCOSMOS). The SCOSMOS MIPS 70 micron data were taken in scan map
mode with medium scan rate. The coverage per pixel is about 130-160.
This type of 70 micron data were taken over a small area (0.25sq deg)
in Spitzer Cycle 2, and were taken again over the full 2 sq. degrees
in Spitzer Cycle 3. The GO2 reduced data has been released to the
public. This will be the data we use for this cookbook.
Requirements:
You must have MOPEX installed in order to follow along with this tutorial.
MOPEX performs both image mosaicking and source photometry extraction. The source extraction part
of MOPEX is also called APEX. MOPEX/APEX can be used in either GUI mode or command line mode.
For the first time users, we recommend the GUI version, which allows easy control and visualization of each processing step.
MOPEX can be downloaded here.
Outline:
- Download the example data set
- Extract Point Source Photometry
- APEX Parameters for COSMOS Data
To follow the step-by-step instructions in this cookbook, you will need to
download the data set associated with Program 20070 (PI = D. Sanders).
This program is Spitzer COSMOS (SCOSMOS) Legacy survey approved in Cycle 2. The MIPS
70 micron data were taken in scan mode. The final mosaic image was made by D. Frayer, a member of the SCOSMOS
team, and a former member of the MIPS instrument support team. This cookbook will use
the mosaic 70 micron image from the cycle 2 data, released by the SCOSMOS team via
IPAC InfraRed Science Archive (IRSA).
The data can be downloaded by going to http://irsa.ipac.caltech.edu/data/COSMOS/images/spitzer/mips/.
From this link, you can download three images: mips_70_go2_cov_10.fits (coverage map), mips_70_go2_sci_10.fits (science image) and
mips_70_go2_unc_10.fits (uncertainty image). This data set provides the mosaic MIPS 70 micron image with fairly uniform background
and mostly unresolved point sources. It should be noted that the SCOSMOS GO2 70 micron data has two different types of
coverage --- the central small area of 0.25sq. deg is deep, with 130-160 per pixel coverage, and the full field of 2 sq. deg
is shallow, with only about 10 per pixel. Readers should use the deep portion of the data to follow the step-by-step
cookbook.
The primary purpose of this cookbook is to present guidelines on how
to extract point source photometry from a 70 micron mosaic with fairly
uniform background. To begin, we recommend that users read the extensive
documents on APEX software at
http://ssc.spitzer.caltech.edu/postbcd/mopex.html. APEX can
provide PSF-fitted photometry as well as aperture photometry (circular
apertures). Because MIPS 70 micron data has a PSF of roughly 16 arcsec
(FWHM), PSF fitting does a better job in separating blended sources,
and also can handle noise more reliably for faint sources than using
aperture photometry. For these reasons, we recommend obtaining
PSF fitted photometry if your targeted science data are point sources
with moderate to low signal-to-noise.
When you download MOPEX, the MIPS 70 micron Point source Response
Function (PRF) is a part of the package. This fits file,
mips70_prf_mosaic_4.0_4x.fits, is usually in a sub-directory called
"/cal". As its name indicates, this PRF was made for mosaicked 70
micron images. This PRF was made from a mosaic with a pixel scale of
4 arcseconds, generated from the Extragalactic First Look Survey
(xFLS). It has a resampling factor of 4. It works well for sources of
low and moderate brightnesses (<200 mJy). In addition, this PRF was
derived to be large enough to go beyond the first Airy ring and its
background was tuned to the zero level. If you use APEX for
photometry extraction, you need to be sure that your 70 micron mosaic
has the same pixel scale (4 arcsecond) as the PRF provided by APEX.
The second important issue to understand before running APEX is the
background level in your mosaic image. This point is briefly mentioned
in the previous discussion of the PRF. The principle is that whatever
PRF you use has to be consistent with the mosaic image on which source
extraction will be carried out. In addition to matching pixel scales
as mentioned above, the background in the PRF image must match that in
the mosaic image. The delivered 70 micron PRF image included in the
APEX package was made to have a background level of zero. Therefore,
users should ensure that within APEX the PRF fitting procedure is
performed on an image which has a zero-background level. To achieve
this goal, you can either directly input a background-subtracted image
into APEX; or you can input a mosaic with non-zero background, then
perform the background subtraction before measuring PRF-fitted
photometry.
Here are the two ways to make background subtracted images from which
PRF-fitted photometry can be measured:
- Manually subtract the
background before starting APEX :
One can derive the background level by finding the median of the data
distribution for source-free pixels within the mosaic. If one uses all pixels within
a region, one can first apply multiple iterations of sigma rejection of outlier pixels before
computing the median value.
One can also fit a Gaussian to the central core of the data distribution after
clipping off the wings.
- Use the APEX module Extract
MedFilter:
It is possible to use this module to produce a reasonable background
subtracted image for PRF fitted photometry. The trick is to make sure
that the box size is big enough so that you are not including positive
signals from sources, thus over-subtracting the background.
With the SCOSMOS 70 micron data, we found that we can tune the APEX
median filter parameters to produce acceptable background subtracted images.
The specific values for the parameters in this module are discussed below.
For SCOSMOS data, we have tested both of these methods, and found very small differences.
It is important to note that we run APEX Med Filter several times to identify
the optimal parameters by comparing the results from both methods. For any other data,
we strongly recommend users to examine the output from MedFilter , and make comparisons
of results from two methods. It is important to be sure that the image used for the
PRF-fitted photometry has zero background.
In general, you will encounter three noise maps after you have
used MOPEX to put together all individual exposures into a
single mosaic.
The noise map with an extension of *_unc.fits
is usually based on the SSC pipeline error propagation using
individual bunc.fits images from the BCD products. The
pipeline uncertainties were tuned to slightly
under-estimate the true noise. This was tuned to be low
to ensure that on-line pipeline processing can produce
reasonable outlier rejection. We do not recommend to use this
type of noise map for source detection and computing SNR.
The second noise image is measured by the standard deviation
of the pixel stack at each sky position. This noise image is
one of the outputs from the image mosaic tool MOPEX, and has
an extension *std.fits. When your data set has good coverage,
this type of noise map is usually much closer to the true sky
noise level.
The third noise estimate is from APEX module
GAUSSNOISE. This method involves measuring the pixel
value distribution within a sliding box. In this method, the
user can clip out a certain number of pixels whose data values
deviate significantly from the median value. We will
demonstrate the specific parameter settings in Section
II. This noise image represents spatial pixel-to-pixel
noise.
APEX uses noise images for two purposes: one is to determine
the detection threshhold, another is to compute the signal-to-noise
ratio for the PRF-fitted photometry. By default, APEX uses the noise
estimated from the
GAUSSNOISE module to determine sources above the
specified detection threshold as well as for computing the SNR
of the resulting photometry. The latest version of APEX also
offers another option which allows users to choose a different
noise image for SNR computation. This parameter is in the APEX
single frame setting --- a switch allows one to use SIGMA_FILE
to estimate SNR. This SIGMA_FILE can be defined by users to
be any preferred noise map.
For the COSMOS data, we found that the noise image made by
GAUSSNOISE is a little bit higher than
mips_70_go2_unc_10.fits. In principle, if module
GAUSSNOISE is properly tuned, its result should be
close to the noise image measured from pixel stack (method 2
above). We recommend that users make comparisons among these
noise images and make the appropriate selection for their own
data. For the specific example here, we used the noise map
estimated from GAUSSNOISE for detection and computing
the SNR during the source fitting.
In this section, we describe how to run APEX step-by-step, using a
subset of SCOSMOS GO2 70 micron data over the central area of 0.25 sq
degrees instead of the full 2 sq degrees. This subset of data has
deep coverage (130-160 per pixel). With the data set covering the
full COSMOS area downloaded from IRSA, you can cut out the central
region with deep data, and put a prefix 'sub' to all the file names. We
run APEX with the following basic steps:
1. Start a new APEX
single-frame pipeline. After starting up MOPEX, go to the main
menu and select File-->New Apex Single-Frame Pipeline. In the window
that pops up, select "APEX 1frame, MIPS 70 microns". The
previously-empty MOPEX window will now be filled with a module flow
consisting of "Initial Setup" and "APEX single frame". It will look
like this:
2. Set parameters.
Only the parameters that require a change from the template default
are listed below.
Initial Settings:
| Input File Name |
mips70/sub_mips_70_go2_sci_10_0bck.fits
The background in this image has been manually subtracted and set to zero. See below
for a detailed discussion. |
| Sigma File Name |
mips70/sub_mips_70_go2_unc_10.fits |
| Coverage Map |
mips70/sub_mips_70_go2_cov_10.fits |
| Output Directory |
mips70/apex_output |
Although we specify sub_mips_70_go2_unc_10.fits as the input noise map, during the extraction and SNR calculation, we recomputed
noise within APEX using GAUSSNOISE module. For details, see below.
APEX single frame settings:
| PRF_File_Name |
/Applications/mopex/cal/mips70_prf_mosaic_4.0_4x.fits
You will have to modify your path to reflect your mopex installation directory |
| Use PSP to detect |
input image w/o background subtraction
Here you need to choose the type of image on which the source
detection is performed. In the APEX single frame setting, there are 4
options:
(1) input image w/o background subtraction, which means
source detection can directly run on it without any additional
background subtraction.
(2) Filtered image, which is essentially
smoothed by a gaussian similar to the MIPS 70 micron PSF.
(3) PSP
image, which is Point Source Probability image, generated by Point
Source probability module.
(4) background subtracted input image,
which will use module Detect Med Filter to produce a
zero-background image.
For the detailed explanation for each of these
options, please refer to http://ssc.spitzer.caltech.edu/postbcd/mopex.html
.
For our specific data set, we choose option (1), which allows
the Detect module to directly run detection on it. With this
option, you can directly fit the PRF to the image to obtain point source
photometry. If you choose option (4), then APEX will need to subtract
the background image. We have tested both options, and found that our
manually background subtracted image produces slightly better (<2%)
fluxes for bright sources than option (4), which uses Detect
Medfilter to subtract the background.
For the shallow xFLS data, option (3), the filtered image, gives good
results in detection, particularly for not producing false detections
near bright sources. Option (4), PSP image, smooths data too
much. A detailed discussion of the shallow data can also been found
for the publicly released 70 micron image and catalogs for the xFLS
data. Please refer to Frayer, D. et al. 2005, AJ, 131, 249. For the
COSMOS data, where source blending is an issue, we use the input image
without background subtraction since the filtered image can smooth out
faint sources near a bright source. The down-side of this choice is
that some false detections near most bright sources can not be
avoided. These false detections can be cleaned up manually. |
| |
use background subtracted image for fitting = no |
| |
use data unc for fitted SNR = no
This option is set to no by default when you use GAUSSNOISE to estimate a better noise map for
fitted SNR calculation. However, if you have a better noise map defined by SIGMA_FILE, you can set this option to yes
for calculting SNR in the PRF fitting procedure. |
Detect Med Filter Settings:
| Window X (Y) |
100 (100) |
| Outliers/Window |
500 |
If the input image to the DETECT module is option (4),
background subtracted image, then you need to run module Detect
MedFilter to set the background level to zero before passing it to
DETECT. Here is the explanation on how to set the parameters,
and what we have tested for the SCOSMOS data set. This module produces
a reasonable background subtracted image if the box size is set
correctly. For MIPS 70 micron mosaic images featuring a pixel size of
4 arcsecond per pixel, one should use a box of 100 pixel by 100 pixel
with the number of outliers set to 500. This rejects about 5% of
bright pixels. This parameter is tuned for the moderate to deep MIPS
70 micron images taken by the SCOSMOS survey. The total integration
time is about 1500 seconds. Boxes smaller than this will produce a
background image with bright patches in regions with clusters of
sources. This causes over-subtraction of the background for bright
sources. This will affect the flux densities of bright sources at a
level of a few percent. Obviously for shallower data with a lower
surface number density of sources, the median box size could be
smaller than what was used for the COSMOS data. Users need to
experiment with their data to find the right parameters.
GAUSS NOISE Settings:
| Window X (Y) |
100 (100) |
| Outliers/Window |
500 |
This module estimates the noise from a distribution of pixel values
measured from a sliding box with a user-defined number of
outliers. This noise image represents spatial pixel-to-pixel noise. By
default, APEX uses the noise estimated from the GAUSSNOISE module to
determine sources above the specified detection threshold as well as
for computing the SNR of the resulting photometry.
DETECT
| Detection_Max_Area |
25
This is a number of pixels within the FWHM of the beam. Here we assume that the pixel
size is 4 arcseconds. It is recommended that you set this value small to avoid
breaking up individual sources.
|
| Detection_Min_Area |
6
We suggest that this parameter is not set to 1 to avoid spurious single pixel
sources. |
| Extended_Object_Area |
10000
You should set this parameter very large to avoid missing bright
nearby sources by mistaking classifying the entire cluster region as
an extended object. |
| Detection_Threshold |
2.5
This threshold depends on the input image
type and use_psp parameter. For the filtered and PSP images, one
needs a larger threshold parameter than used for the raw image to
reach similar source SNRs. We do not recommend digging too deep with
APEX; we recommend only going deep enough to reach the desired depth
in source SNRs such that number of sources in the raw extraction table
is about 1.5--2 times that in the selected extraction table. If you
set the threshold too low (APEX will take a long time) and can assign
real source flux to nearby spurious noise peaks. If you set the
threshold too high, then you can be incomplete at your selected SNR
cutoff (e.g., when the raw table has a similar number of sources as
your selected table). |
| Input_Type |
"image_input"
We use Threshold_Type = 'peak'. This requires a pixel to be larger than the
8 adjacent pixels before declaring a detection (for Peaks_Radius=1).
For close blends, the peaks of the fainter source(s) in the blend(s)
may not be found. Also, APEX factors in the coverage map in
determining whether a pixel is a peak (in rare cases the peak image
pixel is not identified due to local variation of coverage). |
|
|
| Minimum Coverage |
4 |
EXTRACT MEDFILTER
| The parameters in this module are set with the same values as in
Detect Med Filter |
SOURCESTIMATE
| Fitting_area_x/y |
5
The units are pixels. This is set to be similar to the FWHM, so that APEX performs the fitting over a
significant fraction of the area around the peak. If this parameter
is set too low, there will not be enough pixels and APEX may derive a
position off the real peak. |
| Max_Number_PS |
1
This is set to 1 so that APEX fits only one source to each peak.
|
APERTURE PHOTOMETRY:
Number of Apertures
|
3
Number of circular apertures to draw around each source, between 1 and 5.
|
Aperture Radius 1, 2, 3
|
8.0, 10, 12.0
Radius of aperture N in pixels.
|
| Inner and Outer Radius |
15, 30
|
Here we provide the exact parameters that were used for the COSMOS
MIPS 70 micron mosaicked image. This set of parameters has produced a
reasonable photometry catalog.
compute_uncertainties_internally = 0
have_uncertainties = 1
run_detect_medfilter = 0
run_gaussnoise = 1
run_pointsourceprob = 0
run_bright_detect = 0
run_detect = 1
run_select_detect = 0
run_extract_medfilter = 0
run_fit_radius = 0
run_sourcestimate = 1
run_aperture = 1
run_select = 1
OUTPUT_DIR = mips70/apex_output
INPUT_FILE_NAME = mips70/sub_mips_70_go2_sci_10_0bck.fits
SIGMA_FILE_NAME = mips70/sub_mips_70_go2_unc_10.fits
COVERAGE_MAP = mips70/sub_mips_70_go2_cov_10.fits
PRF_file_name = /Applications/mopex/cal/mips70_prf_mosaic_4.0_4x.fits
use_refined_pointing = 0
use_data_unc_for_fitted_SNR = 0
use_background_subtracted_image_for_fitting = 0
PMask_Fatal_BitPattern = 0
use_background_subtracted_image_for_aperture = 0
use_psp_to_detect = -1
use_extract_table_for_aperture = 1
RMask_Fatal_BitPattern = 0
DCE_Status_Mask_Fatal_BitPattern = 0
select_conditions = "SNR > 4.0 and deblend ! NO and deblend ! PO and deblend !
AO and deblend ! PAO"
select_columns =
"srcid,detid,RA,delta_RA,Dec,delta_Dec,delta_RAD,x,delta_x,y,delta_y,delta_xy,fl
ux,delta_flux,chi2/dof,ps_chi2/dof,status,SNR,deblend,aperture1,aperture2,apertu
re3,bad_pix1,bad_pix2,bad_pix3,ap_unc1,ap_unc2,ap_unc3"
use_bright_object_mask = 0
PROBABILITY_THRESHOLD = 0.0
&SNESTIMATORIN (THIS MODULE IS TURNED OFF)
&END
&DETECT_MEDFILTER
Window_Y = 100,
N_Outliers_Per_Window = 500,
Window_X = 100,
Max_Bad_Pixels_OutputImage = 1.0,
Min_GoodNeighbors_Number = 4,
Min_Good_Pixels_In_Window = 9,
&END
&GAUSSNOISE
Window_Y = 100,
Max_BadPixels_OutputImage = 1.0,
N_Outliers_Per_Window = 500,
Window_X = 100,
Min_GoodNeighbors_Number = 4,
Min_Good_Pixels_In_Window = 9,
&END
&POINTSOURCEPROB
PRF_ResampleY_Factor = 4,
Apriori_Probability = 0.1,
PRF_Xsize = 11,
PRF_Ysize = 11,
PRF_ResampleX_Factor = 4,
&END
&BRIGHT_DETECT (THIS MODULE IS TURNED OFF)
&END
&DETECT
Detection_Threshold = 2.5,
Input_Type = 'image_input',
Detection_Max_Area = 25,
Min_Coverage = 4.0,
Detection_Min_Area = 6,
Threshold_Type = 'peak',
Extended_Object_Area = 10000,
&END
&EXTRACT_MEDFILTER (This modulel is turned off for this setting. If the initial input image has none zero background, this module needs to be on)
&END
&FIT_RADIUS
&END
&SOURCESTIMATE
MinimizeFtolSuccess = 1.0E-4,
PRF_ResampleX_Factor = 4,
Max_Number_PS = 1,
Chi_Threshold = 3.0,
Background_Fit = 0,
Max_N_Iteration = 50000,
Fitting_Area_Y = 5,
InputType = 'image_list',
PRF_ResampleY_Factor = 4,
DitherPixelFraction = 0.1,
Max_N_Success_Iteration = 1000,
Chi2_Improvement = 1.0,
DitherFluxFraction = 0.8,
MinimizeFtol = 1.0E-4,
Fitting_Area_X = 5,
DeblendDitherPixelFraction = 1.0,
N_Edge = 4,
Random_Fit = 0,
&END
&APERTURE
Annulus_Compute_Type = 'median',
N_Apertures = 3,
Use_Annulus = 1,
Aperture_Radius_1 = 8.0,
Aperture_Radius_2 = 10.0,
Aperture_Radius_3 = 12.0,
Min_Number_Pixels = 10,
Inner_Radius = 15.0,
Outer_Radius = 30.0,
&END
&SELECT
&END
#END
|
