next up previous contents index
Next: Local and Map Source Up: 5.1 Examples of Typical Previous: Source Detection on PSPC

Source Detection on HRI data

 

In this section an example for source detection on HRI data     is given. In comparison to the previous section all detection steps will be done with the single command DETECT/SOURCES. There is no need to edit parameter files any more. The price for that is that the user has to rely on the predefined defaults which may not always be the ideal ones for his data. A few preparatory steps have to be performed before the   source detection can be started.

  1. An image must be created. Images appropriate for source detection must have a size of 512 tex2html_wrap_inline17489 512 pixels and the image pixel size is 10 sky pixels (5 arcsec) in the default case for ROSAT HRI. There is an automatic command to bin an image fulfilling these requirements and it is highly recommended to use this command:

    Midas 001> create/source_detect_image testdata31:events 0-15 ? ?  
               testdata31:eventrates
     

    An image image1.bdf will be created in the raw amplitude channel channel range 0 to 15 from the photon event table events.tbl in the testdata area testdata31. The lifetime correction factor is calculated and written to the descriptor LIF_TIM.

  2. A mask will be created to screen out the outer edge of the detector. This is done by a shift and a rotation of an existing mask image EXSAS_CAL:mask0_hri.bdf. The rotation and the translation of the mask image is done by determining a mean rotation angle and a mean translation from the attitude table with the command ADJUST/MASK.

    Midas 002> adjust/mask image1 EXSAS_CAL:mask0_hri ? ? ? 
               testdata31:attitude load
     

    An adjusted mask called mask.bdf is created. It is strongly recommended to control at this point the correct overlay of the mask on the image (use the default load display option as done in the example before).

Now the local, map and maximum likelihood source detection can be performed by the command

Midas 003> detect/sources image1 testdata31:events mask displ 10
  A table solst.tbl containing all sources accepted by the maximum likelihood technique and having a likelihood tex2html_wrap_inline17491 10. is created. We read the table, which contains 3 sources.

Midas 004> read/tab solst

  Table : solst

 Sequence INDEX IND_MP TOTNUM X_POS           Y_POS           CTS_MP   BG_MP
 -------- ----- ------ ------ --------------- --------------- -------- ------
        1     1      5   1985           78.56          226.02  2428.31   0.08
        2     2      8     13          239.33          353.67     6.88   0.07
        3     3      9    279           85.55          383.72   219.00   0.05
 -------- ----- ------ ------ --------------- --------------- -------- ------
 Sequence ML_MP X_IMA  Y_IMA  CTS      EXI_ML  XERR   YERR   CERR  EXT
 -------- ----- ------ ------ -------- ------- ------ ------ ----- ------
        1  1529  79.65 226.35  1839.23  9831.7   0.03   0.03  43.1   0.42
        2   375 239.45 353.55     6.25    10.7   0.32   0.31   2.6   0.00
        3   108  85.45 384.05   175.31   480.1   0.19   0.19  13.6   1.16
 -------- ----- ------ ------ -------- ------- ------ ------ ----- ------
 Sequence EXTERR EXT_ML FA     PRIO VIG_COR RATE      ERATE
 -------- ------ ------ ------ ---- ------- --------- ---------
        1  0.231    6.3  1.000   48   1.046   0.90532   0.02120
        2  0.000    0.0  1.000   32   1.021   0.00301   0.00127
        3  0.694    5.3  1.000   48   1.071   0.08836   0.00686
 -------- ------ ------ ------ ---- ------- --------- ---------
The parameters of the source tables are described in Table 5.29. We convert the image pixel coordinates to equatorial, ecliptic and galactic coordinates for equinox 2000 with the command

Midas 005> transform/coord solst solst_2000 2000.0
and we read the table solst_2000

Midas 006> read/table solst_2000 

  Table : solst_2000

 Sequence INDEX  X_IMA       Y_IMA       X_SKY            Y_SKY
 -------- ------ ----------- ----------- ---------------- ----------------
        1      1       79.65      226.35         -1769.00          -302.00
        2      2      239.45      353.55          -171.00           970.00
        3      3       85.45      384.05         -1711.00          1275.00
 -------- ------ ----------- ----------- ---------------- ----------------
 Sequence RA_DEG    DEC_DEG   RA_HMS       DEC_DMS      LAMDA     BETA
 -------- --------- --------- ------------ ------------ --------- ---------
        1 332.17203  45.74117 22H08M41.28S +45D44M28.2S 358.69839  51.87713
        2 331.85391  45.56504 22H07M24.93S +45D33M54.1S 358.23993  51.85877
        3 332.15915  45.52217 22H08M38.19S +45D31M19.8S 358.48508  51.70212
 -------- --------- --------- ------------ ------------ --------- ---------
 Sequence LII       BII       OFFAX      FWHM       DIST
 -------- --------- --------- ---------- ---------- ----------------
        1  95.55840  -8.30333      14.95      62.51          1562.19
        2  95.27196  -8.31495       8.21      22.51          1554.03
        3  95.42119  -8.47541      17.78      86.81          1553.84
 -------- --------- --------- ---------- ---------- ----------------
The parameters are described in Table 5.17. We load the image image1.bdf and display the source positions into the overlay channel.

Midas 007> draw/source solst 1 circle white
The result is shown in Figure 5.6.

  
Figure 5.6: ROSAT HRI image (left upper figure). Cheesed image (right upper figure) created by the command CREATE/BG_IMAGE. This is the image with the removed source positions, found by the command DETECT/LOCAL in circles with a radius of 2 tex2html_wrap_inline17493 the FWHM of the point spread function. Binned cheesed image before bicubic spline fit has been applied (left lower figure). Background image (right lower figure) created by the command CREATE/BG_IMAGE. It is a bicubic spline-fit to the binned cheesed image.

  
Figure 5.7: Source positions determined from the maximum likelihood detection (command DETECT/MAXLIK) on HRI data. The circles drawn with the command DRAW/SOURCE_POSITIONS are centered at the source positions and the radii of the circles are equal to the FWHM of the point spread function. Sequence numbers are displayed for the detected sources.


next up previous contents index
Next: Local and Map Source Up: 5.1 Examples of Typical Previous: Source Detection on PSPC

If you have problems/suggestions please send mail to rosat_svc@mpe-garching.mpg.de