STScI Nov. 21, 1997 Version 2.1 (Updated to reflect FITS changes in IRAF) David Soderblom (revised version of 1995 instructions by Palle Moller) How to obtain target coordinates from the DSS. ============================================== To ensure successful completion of your HST program, it is critical that you provide accurate and precise target coordinates, and that they be in the HST Guide Star reference frame. Several HST observations fail each cycle, and often the cause can be traced back to inaccurate target coordinates provided by the observer. In such cases the proposer must bear the responsibility for the failure, and the observations will not be repeated. These instructions are intended to help HST proposers obtain accurate coordinates for their targets, thereby minimizing a major source of observation failures. European observers who use the ESO-MIDAS system and do not have access to IRAF should contact the ST-ECF in Garching (e-mail: stdesk@eso.org) where they can obtain software tools adapted to work on ESO-MIDAS format images as well as instructions and assistance. Observers with access to IRAF and STSDAS should refer to the instructions given below. These come in two parts: Part A: Short summary. Gives a short list of things you need to do to obtain the coordinates. If you are familiar with accessing the Digitized Sky Survey (DSS), and with using IRAF, it should be enough for you to read this. Part B: Detailed "how to do it". This gives a complete step by step description of all you need to do to obtain the coordinates. If you are not familiar with accessing DSS, or with using IRAF, you should read Part B before you begin, or you can read it as you go along. ======================================================================== ************************************************** * For observers without access to STSDAS in IRAF * ************************************************** Contact the ST-ECF in Garching, Germany where software and instructions can be obtained (e-mail: stdesk@eso.org). ======================================================================== *********************************************** * For observers with access to STSDAS in IRAF * *********************************************** Part A --- Summary reference sheet. =================================== 1. Obtain an image of your field from the DSS: http://stdatu.stsci.edu/dss/dss_form_phase2.html IMPORTANT NOTE: You *must* use the http address given above, not the normal DSS or Starview access. The address given above automatically defaults to the correct Guide Star Reference Frame scans, other methods of accessing the DSS may use different scans. 2. Dump a small image containing the target on your disk as a FITS file. 3. Display the IRAF-format image and identify your target; saoimage works well. 4. Measure (x,y) coordinates of your target using the IRAF command "imexamine". 5. Run IRAF command "xyeq" in the stsdas.gasp package. The output is in J2000 coordinates. 6. Obtain the PLATE-ID from the header of the DSS image. The coordinates and corresponding PLATE-ID are to be inserted in your Phase II proposal. 7. To check that you got it right, there is an example attached at the bottom of this sheet. Try to see if you can reproduce the correct coordinates of this object. Part B --- Detailed instruction sheet. ====================================== In the detailed instructions below "pr>", "cl>" etc. at the beginning of a line is always your prompt (hence you DON'T type this), <RTN> means that you must press return. 1. Obtain an image of your field from the DSS. You can do this using Netscape (do *not* use Starview, for parts of the sky Starview is accessing different scans). Go to: http://stdatu.stsci.edu/dss/dss_form_phase2.html Activate the "help" pointer at the top of the page, and read the instructions. Enter your best guess for the coordinates of the object (choose J2000 or B1950 coordinates), set the size of the image to be as small as possible but large enough that you can identify the object. If you are on a very slow network you may want to set the "Compression" to one of the two compression options. Then activate the "Retrieve Image" button. NOTE: Choosing B1950 or J2000 as input coordinates has no effect on your final output coordinates (see below), they are always J2000. 2. When the data are ready, a new window will appear. Enter in it a name for your FITS file (e.g. DSSim.fits) and activate the "OK" button. The FITS image has now been created and is ready on your disk. 3. Now enter IRAF the normal way (probably through an xterm window). If you are not familiar with IRAF, or you are not sure if you have access to IRAF, you must ask your local system manager. Once in IRAF, you get a "cl>" prompt. You must now load the "stsdas" package by typing cl> stsdas <RTN> which gives you an "st>" prompt. 4. In your console window open a display package such as saoimage: saoimage & <RTN> 5. Display the image using the display command and identify your object. Example: -------- st> display DSSim <RTN> z1=2496. z2=4487.092 st> 6. Use the command "imexamine" to measure the (x,y) coordinates. You do this by typing st> imexamine <RTN> place the cursor on top of your object and press "a" on the keyboard (press "a" only, no <RTN>), IRAF will then calculate the best center of the object under the cursor, and return its (x,y) coordinates. If you press "r" rather than "a" you will also get a radial plot of the object profile. This is useful to verify that the centering algorithm was successful, and that the object is not badly saturated. (If your object is badly saturated (hence bright) you should use the coordinates listed in the Guide Star Catalogue. For instructions and access to the GSC, go to http://www-gsss.stsci.edu/support/phase2.html). If you have more objects to determine positions of in this image, then just continue to the next one and type "a" again. When you are done, type "q" to leave the imexamine command. You now have a normal prompt "st>" again. The numbers listed under "COL" and "LINE" are the x and y coordinates respectively. Example: -------- st> imexam <RTN> # COL LINE RMAG FLUX SKY N PA PEAK FWHM 88.91 89.04 18.37 44832.0 4244.00 75 -46.7 10899.37 2.03 125.52 158.41 16.90 173211.0 3008.00 78 11.2 11312.93 3.91 st> In the above example the (x,y) coordinates of two objects at positions (88.91, 89.04) and (125.52, 158.41) were determined. 7. You are now ready to convert the (x,y) coordinates to (RA, dec). First you must load the "gasp" package (which is within "analysis") by typing st> analysis <RTN> an> gasp <RTN> which gives you a "ga>" prompt. To convert the coordinates you must execute the "xyeq" command. It is strongly recommended that you first type ga> help xyeq <RTN> and read the documentation carefully. This command critically needs a number of parameters to be set correctly. If the parameters are set incorrectly the command may not execute, or worse, it may return wrong coordinates. If you are not an experienced IRAF user, we recommend that you follow this method: a) Enter the (x,y) coordinates into an ascii file named "xydat". In the above example the file xydat should simply contain: 88.91 89.04 125.52 158.41 b) Change the parameter file for the "xyeq" command. When you enter IRAF this is the default parameter list you will have for xyeq: iminfo = Is there an input image image = Input image name coeffile = Input astrometic solution file xyfile = Input (x,y) file name pix_center = "iraf" X,Y origin: iraf(1,1); dss(1.5,1.5); cos(0.5 xcolnum = X pixel column number ycolnum = Y pixel column number nskip = Number of lines to skip in xy file (original = no) Use original plate solution? (new = yes) Use new plate solution? (cdmatx = no) Use CD matrix values? (ra_hours = no) Display RA in h:m:s? (ra_format = ) Output format for RA (dec_format = ) Output format for DEC (mode = "al") Now edit this parameter file for the xyeq command by typing ga> epar xyeq <RTN> You should edit the following parameters: iminfo = yes Is there an input image image = "DSSim.fits" Input image name xyfile = STDIN Input (x,y) file name pix_center = "iraf" X,Y origin: iraf(1,1); dss(1.5,1.5); cos(0.5 xcolnum = 1 X pixel column number ycolnum = 2 Y pixel column number nskip = 0 Number of lines to skip in xy file (original = yes) Use original plate solution? (new = no) Use new plate solution? (ra_hours = yes) Display RA in h:m:s? (ra_format = %12.2h) Output format for RA (dec_format = %12.2h) Output format for DEC c) Now you execute "xyeq" simply by typing ga> xyeq <RTN> Example: -------- ga> xyeq <RTN> Is there an input image (yes): <RTN> Input image name (DSSim.hhh): <RTN> Input (x,y) file name (xydat): <RTN> X,Y origin: iraf(1,1); dss(1.5,1.5); (iraf|dss|cos) (iraf): <RTN> X pixel column number (1): <RTN> Y pixel column number (2): <RTN> Number of lines to skip in xy file (0): <RTN> Output format for RA: <RTN> Output format for DEC: <RTN> X Y RA_osol DEC_osol (hrs) (deg) 88.91 89.04 1:04:36.07 41:16:00.13 125.52 158.41 1:04:30.46 41:17:57.57 ga> You have now obtained the J2000 coordinates (RA_osol, DEC_osol) of your targets in this field. You can stay in the "gasp" package and repeat all of the above steps for your next field. 8. Obtain the PLATE-ID from the header of the DSS image. It is important that the PLATE-ID is obtained from the same image as the coordinates, because it defines the coordinate system. Header information is most easily obtained with the "imheader" command. However, "imheader" will list the entire header. To select only the "PLATEID" keyword, pipe the output of the command through the UNIX command "grep" (this works only under a UNIX based IRAF). Type ga> imheader DSSim lo+ | grep PLATEID <RTN> Thu Dec 10 15:14:29 EST 1995 PLATEID = '07WO ' /GSSS Plate ID Enter coordinates and PLATE-ID in your Phase II proposal. In case you run VMS based IRAS you may pipe through the "page" command and look for the PLATEID keyword. In case proper motion is important for your object, you must also obtain the epoch (header keyword named "EPOCH") of the plate and enter it into your proposal. 9. To check that you understand the method, here is an example you can test. First retrieve a 3' by 3' image centred on the B1950 coordinates (RA = 5 28 5) and (dec = -25 6). Convert and display the image. Determine (x,y) coordinates of the quasar (roughly 18 mag) seen a bit above, and slightly left of, the centre. Convert to J2000 coordinates: Example: -------- ga> display DSSim <RTN> frame to be written into (1:4) (1): <RTN> z1=3406. z2=6048.274 ga> imexamine <RTN> [move cursor to object, hit "a" key] # COL LINE RMAG FLUX SKY N PA PEAK FWHM 52.91 63.03 18.37 44832.0 4244.00 75 -46.6 10895.86 2.03 [with cursor in saoimage window, hit "q"] ga> xyeq <RTN> Is there an input image (yes): <RTN> Input image name (DSSim.hhh): <RTN> Input (x,y) file name (xydat): <RTN> X,Y origin: iraf(1,1); dss(1.5,1.5); (iraf|dss|cos) (iraf): <RTN> [if in manual mode, enter x and y coordinates here, then <RTN>] X pixel column number (1): <RTN> Y pixel column number (2): <RTN> Number of lines to skip in xy file (0): <RTN> Output format for RA: <RTN> Output format for DEC: <RTN> X Y RA_osol DEC_osol (hrs) (deg) 52.91 63.03 5:30:07.94 -25:03:30.1 ga> imheader DSSim lo+ | grep PLATEID <RTN> Mon Dec 11 16:11:39 EST 1995 PLATEID = '04NM ' /GSSS Plate ID ga> imheader DSSim lo+ | grep EPOCH <RTN> Mon Dec 11 16:12:12 EST 1995 EPOCH = 1.9801049804688E+03 /Epoch of plate ga> ======================================================================== If you have further questions, please contact your Program Coordinator or Contact Scientist.