While calfgsa processes each single observation in a stand-alone fashion, ignoring other observations belonging to the same HST visit. calfgsb processes an entire visit’s worth of data as a coherent set, allowing an astrometric “plate” to be produced by tools such as gaussfit.

In a typical Position Mode observing program the astrometer FGS sequentially observes several stars distributed about the FOV. Any temporal variability in telescope pointing will contaminate the measured relative positions of these targets. FGS astrometry is sensitive to HST body jitter and FGS drift. The jitter can be eliminated using the guide star data, whereas the drift is removed by applying a drift model derived from check star data. A check star is a star that is observed multiple times during the visit. Typically the observing strategy should involve at least two check stars observed a minimum of three times each (see the FGS Instrument Handbook for more details on observing strategies).

The calibrations carried out by calfgsb to correct a star’s observed (x, y) position are, in order, for

• geometric distortions (the optical field angle distortion, or OFAD, correction)
• differential velocity aber
• space craft jitter
• drift of FGS field of view.

The data needed to carry out these calibrations are contained in the observation’s *.tab file (provided by calfgsa) and/or in the calfgsb reference library. The calfgsb reference library can be down loaded from the STScI FGS Website.

3.4.1 Preparing to use calfgsb

calfgsb has explicit, well defined expectations of directory structures and their contents that are to be in place before calfgsb is activated. Residing in the directory from which the calfgsb command is executed is the "day.list" file, which contains a list (one entry per row) of the subdirectories containing (calfgsa) astrometry data from various epochs. For example, if the "day.list" file contains

then the sub-directories v01, v02, v03 must also be present. Note that the "/" is not included in the sub-directory specification in day.list, i.e., "v01" is specified, not "v01/". A directory called "pmtdata/", containing the *.tab files generated by calfgsa, is required within each of these sub-directories.

Note that only *.tab files from Position mode observations should be present. calfgsb does not and cannot process Transfer mode data. If Transfer mode *.tab files are present an error results which aborts calfgsb.

calfgsb processes each visit’s data (each entry in the "day.list" file) independently of the other visits. There is a single exception to this claim. Ultimately, the data from various epochs must be combined in order to achieve one’s science goals (for example, to determine an object’s parallax). The analysis tool gaussfit, discussed in detail in Chapter 5, ingests the output products of calfgsb to derive the desired asrtrometry. Each visit is tracked by its "set" number. For convenience, calfgsb increments the value of the set number it assigns to each visit in the "day.list" file. Note that calfgsb prompts the user for the starting value of the set number. It is not necessary to process more than one visit’s worth of data at a time - it’s just more convenient to do so when opportunity arises.

3.4.2 Executing calfgsb

To best way to understand how to execute calfgsb is by working through an example. Suppose there are six visits of HST astrometry data to be processed by calfgsb with the objective of determining the parallax of an astrophysically interesting star. The visits were obtained at times of maximum parallax factors (0, 6, and 12 month intervals), with two HST visits expended at each epoch. The data (*.tab files) reside in the "pmtdata/" directories of the

subdirectories, where v01, v02, etc. correspond to the visit number of the observed data.

The user-created "day.list" file lists the v01 through v06 directories, and should reside in the directory above these sub-directories. From this directory, simply type "calfgsb" from IRAF/STSDAS to execute calfgsb. The following is a screen snap shot initiating such a session. User input is in bold. Entering <ret> causes the default value for the given prompt, shown within the parentheses, to be used in the processing.

After entering responses to these prompts calfgsb automatically processes all the data specified in the "day.list" file. A series of output is sent to the screen but is also captured in a *.log file. A new directory, "pipedir/" is created parallel to the "pmtdata/" directory. All calfgsb products from the processing of a particular visit are placed within the "pipedir/" subdirectory. The rootnames of the output files will, in the example shown above, have "ugem" embedded within. The most important output file will be "ugem_v01_ast" (in "v01/pipedir"). It is this file that is to be edited (as described in Chapter 4 of this document) in preparation for input by gaussfit when, in this example, this visit’s data is to be combined with data from the other five visits for a parallax determination of the target star (U Geminorum).