WFPC2 Associations User Interface

HOWTO browse through the associations on the web

The User Interface for the WFPC2 Associations makes use of the web. A part from the regular HST archive "science" screen to browse through all HST observations (not only WFPC2, but also including the WFPC2 associations),

it is possible to browse through the associations making use of a dedicated "WFPC2 Associations" interface (click&search!).


Similar interfaces are also available at the Canadian Astronomy Data Centre (CADC) and at the Space Telescope Science Institute (STScI).


Via all these interfaces the user can request the enhanced WFPC2 associations products, which have been generated using the Artificial Skepticism stacking algorithm (Stetson 1989, V Advanced School of Astrophysics [Univerisidade de Sao Paulo], p.1). Such products have also been corrected in their WCS information, by using the USNO2 stars in the field of view of the camera (when available).

The user interested in time variability can now make use of the time span field to easily select images of the same field observed at different epochs.

Archive Researchers already used to our web interface of the HST Mission Log (the so called science screen) will get easily acquainted with the concept of WFPC2 associations.

The User is presented with a list of associations, e.g., of the selected target, instead of getting a long list of otherwise non-related WFPC2 exposures.

In this way it is much easier to understand the observing strategy of any WFPC2 program.

Via the asn_id hyperlink, any association can be expanded in its members, leading to a new web screen: the association members screen.

Archive Researchers already used to our web interface of the HST Mission Log (the so called science screen) are already acquainted with the concept of WFPC2 associations (type A).

The user interface hasn't changed much for associations of Type B. The only difference is in the so-called science_members screen.
In the type A case, the displayed shifts were always only the shifts computed via the jitter pointing information.
In the type B case, the shifts are computed with three different methods (WCS, jitter, cross-correlation). All the shifts are displayed.

The association members screen not only contain the list of datasets belonging to the selected association, but also shows thirteen important fields:

  • Dx & Dy (4 couples):

    The offsets in WF4 pixels of the current exposure wrt. the so called leader (the first deepest observation) of the association.
    You can also see (Dx, Dy) as the position of the current exposure's WF4 origin as seen in the association leader's WF4 reference frame.

    There are 4 couples of offsets:

    • USED DX & DY
      The shifts which are going to be used by the archive pipeline to build a images.immatch.imcombine product.
      These shifts are the best of the three shifts to follow.

    • W-DX & W-DY (WCS)
      The shifts computed comparing the WCS information available in the science header, with the WCS of the association leader.

    • J-DX & J-DY (Jitter)
      The shifts computed using the Jitter information, available from the Observatory Monitoring Support Jitter files.
      The quality of the jitter is also evaluated, see "jitter flag" below
      These offsets were used in the previous associations (associations of type "A").

    • C-DX & C-DY (Cross correlation)
      The shifts computed using the cross-correlation method. This method, which actually measures the shifts onto the images, is supposed to offer the most reliable offsets.

    The value used for the WF4 pixel scale along the X axis is: 0.099539 (or twice as much if the image was binned).
    To compute the shift magnitude in arcsecs you have to multiply the given Dx by 0.099539


    The value used for the WF4 pixel scale along the Y axis is: 0.099635 (or twice as much if the image was binned).
    To compute the shift magnitude in arcsecs you have to multiply the given Dy by 0.099635

    To shift the WF4 chip of a given exposure onto the WF4 chip of the leader (typically Dx=0, Dy=0) you have to shift by the following amounts:

    imshift iraf task:

    imshift: Xout := Xin + Dx
    i.e., use imshift passing to it the shifts as given in this table:

    imshift IMin IMout Dx Dy

    drizzle iraf task:

    drizzle: Xout := Xin + Dx
    i.e., use drizzle passing to it the shifts as given in this table:

    drizzle ( xsh = Dx ) ( ysh = Dy )

    More complicated formula occours for the other (WF) chips.
    You can use the shiftChip cgi to compute the DX and DY relative to other chips.

    To know more, please refer to the pointing section of the FAQs.

  • JFLAG - Jitter Flag:

    The quality of jitter information is assessed by looking at some telemetry keywords in the header of the jitter files and by looking at the RA, DEC and ROLL standard deviations.

    The so called jitter flag can assume three values:

    P (processable), for reliable jitter information,
    G (groupable), not completely reliable pointing information
    B (bad), pointing information not reliable at all or missing

    Please refer to the Jitter Quality Table for more details.
  • C-ERR - Cross correlation error:

    The cross-correlation process provides also the error of the measurement (in WF4 pixels); if the error is too high, it will be highlighted with a red colour. In such case, the USED values will be chosen between the Jitter and the WCS shifts. Priority is give to the jitter values, provided that the jitter flag is equal to "P"

    The cross-correlation error is not the only test performed to ensure the correctness of the shifts. Since the shifts of each individual WFPC2 chip are computed separately, a voting system is introduced to make sure that the intra-chip shifts agrees.

  • Shift Type:

    The type of offsets which are going to be used by the archive stacking pipeline (see USED Dx & Dy).
    This value is automatically assigned by evaluating the quality of the Jitter and cross-correlation shifts.
    It can assume three values:

    • X: for cross-correlation shifts

    • J: for jitter shifts

    • W: for WCS shifts (in some cases, the value "N" might appear instead)

  • Elapsed Time

    This field shows:

    for the leader
    The start time of the observation.
    for each member
    The elapsed time from the leader, in units of 1 HST orbit (96 minutes), or equivalently (depending on the service), in units of days (1 day = 15 HST orbits).
    It is useful to evaluate the jitter shifts quality, since a jitter zero point drift has been discovered.
    Rule of Thumb: if the elapsed time is less than 10 orbits the drift is negligible.
  • Shift Quality

    This field is not stored in the database; instead, it is computed on the fly. It applies some rules to flag potential problems with the provided shifts (See Dx and Dy later) to the users.
    The icon indicates that there the USED shifts are to be trusted.

    The icon indicates that the given exposure might have associated non-fully certified offsets. A possible scenario where this could happen is the following: the used offsets are the WCS ones and the observing date was prior April 1996 (in that case the dithering pattern was not reflected in WCS). Still, this doesn't mean that a red icon indicates a real problem. In fact, as in the example above, a CR-SPLIT observation (2 or more perfectly aligned WFPC2 observations) observed before April 1996 will be still marked red, even though the WCS gives the correct offsets.

    It is left to the user to understand what kind of problems could arise by using the provided offsets, or the products automatically produced by the archive retrieval system.

    For a list of potential problems, please refer to the Idiosyncrasies & Warnings section of the FAQ.

    See the Frequently Asked Questions document to find out more.

    If you have suggestions, you are very welcome in sending them to Alberto Micol (Alberto.Micol@eso.org).