Change page style: 

Mid-IR Astrometry

Mid-IR observing programmes would often benefit from absolute astrometry with subarcsecond accuracy, for example for comparing mid-IR images of galaxy nuclei or star-forming regions with images at other wavelengths. In this page we guide the user through the steps necessary to set up accurate absolute astrometry in the OT. The procedure at mid-IR wavelengths differs from that used in the optical because of low probability that two stars close enough to use the same guide star will be both bright at 10 or 20 microns. As well as the following description, an example of a User 1 setup is shown in detail in the T-ReCS OT library, and the setup is essentially the same for Michelle.




User 1 Astrometry

Two types of astrometry or blind offsetting are available (see also the generic Gemini target acquisition scenarios). In the preferred case, termed the "User 1" case, an optical reference star (the "User 1" star) and the science target share the same guide star. Astrometry using this method can be accurate to 0.2” for a User 1 star located within 50” of the source. The astrometry becomes less accurate the further away the reference star (precise measurements are still being gathered by Gemini staff), so if no suitable star is located within 50” of the source one as close as possible to the science target should be chosen.

The procedure involves slewing first to a bright 10 micron star (the "hotspot" star). This star is placed at the centre of the Acquisition Camera, and then (known) offsets applied to move it to the Michelle or T-ReCS "hotspot" (i.e., to a location at or close to the instrument array centre but off-centre on the Acquisition Camera). An image is then taken with the mid-IR instrument to record the location of the star on the array. Next, the telescope is slewed to the User 1 star which is then moved onto the instrument hotspot. Once this star is accurately positioned on the hotspot (as judged by its position on the Acquisition Camera), the telescope is offset to the coordinates of the science target, which should then appear at the same position on the array as the bright 10 micron star.

STEP 1: Select a “hotspot” star, a bright mid-IR star within 20 degrees of your source, and set up a short (e.g. 30 sec) imaging observation (Fig. 1). A Cohen standard or other star of spectral type K or M and Vmag<8 should be suitable. This observation will establish the expected position of the science target on the array and is charged to the PI (please be sure to include extra time in your proposal).



Fig. 1 (click for high-resolution version)


STEP 2: Select a "User 1" star. This is a star close to the science target (preferably <50”) that is bright enough to be easily detected by the optical acquisition camera (Vmag <18), and has coordinates that are accurate and in the same system as those of the science target. The Second U.S. Naval Observatory CCD Astrograph Catalog (UCAC2) from CDS has been used effectively in the past. The UCAC2 is the second release of the ongoing UCAC project, and contains stars from about R magnitude 7.5 to 16 at declinations mostly from –90 to +40. The positional errors are about 20 mas for the stars at R~10 to 14, and about 70 mas at the limiting magnitude of R~16.

A User 1 star can be selected in a similar manner to a guide star by using the “Position Editor” (Fig. 2). Simply load a catalogue (e.g. UCAC2) and select a star within 50" of the target that can use the same guide star as the science target. The User 1 star must be in the PWFS FOV (i.e., not farther than 7 arcminutes away from the guide star), and must be far enough away from the guide star (~4 arcminutes) to avoid its being vignetted by the probe arm. This can be measured by holding down the right mouse button while on the guide star and dragging the mouse to the position of the proposed User 1 star. The distance will be shown in arcminutes and arcseconds.

There is no specific button on the position editor for selecting a User 1 target. However, since T-ReCS and Michelle use the PWFS2 (in yellow) one may use the "add PWFS1" button to pick a User 1 target (Fig. 3). Vignetting with PWFS1 is not an issue in this case since it is used simply to select a target and not for any actual guiding.

Fig. 2 & 3 (click for high-resolution version)


Once the User 1 target has been selected using PWFS1, go to the science target in the OT, select the PWFS1 star and at the top in the “Tag” drop-down menu, change it to “User” (Fig. 4).


>

Fig. 4 (click for high-resolution version)


Once this is done, the PWFS1 probe arm area (in white) displayed in the position editor will disappear. Currently there is no specific graphic to show the User 1 target in the position editor, but the coordinates of the User 1 target can be seen in the "target environment" display in the OT (Fig. 5). Once the User 1 star has been selected, the rest of the observation is set up in the usual manner (see the Michelle and T-ReCS OT libraries for examples).



Fig. 5 (click for high-resolution version)


NON-User 1 Astrometry

In the case where no User 1 star is available, the telescope may use the guide star as an astrometric reference. This case is not as accurate as User 1 and Gemini staff is still gathering precise measurements of its accuracy. As this case is still under testing it should be used with caution. The steps for this procedure are outlined below.

STEP 1: As above select a “hotspot” star, i.e., a mid-IR-bright (and optically-visible!) star within 20 degrees of your source. This short imaging observation is to calibrate the optical-to-mid-IR centering and is charged to the PI.

STEP2: Select a guide star for your target with a coordinate system that is accurate and the same as that of the science target, so that the telescope control system will determine accurate offsets between the two positions. As above the UCAC2 catalogue seems to work well.

STEP 3: In order to highlight that this is not a User 1 sequence that is simply missing a User 1 target, put a note in the sequence explicitly confirming that no User 1 target is available and that this is a non-User 1 case.



Last update 2006 August 24; Rachel Mason (based on original page by James Radomski)