Description of SUSI2 setup for astrometry within the program ============================================================ 'Trigonometric Parallaxes for Member Stars of the Young Nearby TW Hya ===================================================================== Association' ============ Sabine Reffert, December 2004 contact information ------------------- email for questions etc.: sabine@strw.leidenuniv.nl for contacts during the night (please do not hesitate to call if something is unclear, even in the middle of the night!): Leiden, office: +31-71-527 8411 Leiden, home: +31-71-576 2746 Heidelberg, home: +49-6221-401394 target list ----------- There are 10 science targets (TTS in TW Hya association) plus a cluster (NGC 2571) which is used for calibrations and observed at the beginning and at the end of the half nights. So a typical observing sequence would look like this, where the stars are arranged according to right ascension, and the cluster is added at the beginning and the end: --------------------------------------------- object RA (J2000) DEC priority --------------------------------------------- NGC 2571 08 18 53 -29 44 42 very high TWA 6 10 18 27 -31 50 02 low TWA 7 10 42 28 -33 40 17 very high TWA 1 11 01 50 -34 42 17 very high TWA 14 11 13 25 -45 23 43 high TWA 12 11 21 04 -38 45 16 high TWA 9 11 48 22 -37 28 49 very high TWA 15 12 34 19 -48 15 15 low TWA 10 12 35 02 -41 36 39 high TWA 17 13 20 43 -46 11 38 high TWA 18 13 21 35 -44 21 53 high NGC 2571 08 18 53 -29 44 42 high --------------------------------------------- The priority column gives an indication how important a particular observation is, for bad weather, e.g., where not all targets can be observed. charts ------ Finding charts are available for all objects. It is important to place the target on a certain pixel, with good precision. More details are given below. filters ------- On the filter list sheet, which has to be handed in to the support astronomer in advance, 3 filters should be selected: the B, V and OIII/cont filters. B: #811 V: #812 OIII/cont: #883 calibrations ------------ The most important calibrations are skyflats in the OIII-continuum filter (ESO #883), which is probably non-standard and should be requested in advance of the run. It would be rather good to have both evening and morning skyflats. 8 flats, or however many are possible, are fine. In addition, domeflats should be taken in the evening and in the morning, as a safety precaution, just in case the skyflats do not work. Also, domeflats are needed in additional filters (B + V, plus OIII/c). Five flats in the evening in each filter, and another five in the morning are fine. More flats are always better! In addition, for the photometric calibrations, we have also been taking bias observations, five in the evening and five in the morning. Note: Sometimes it is possible to share calibrations with the program that is scheduled for the other half night. However, many programs use binning of the CCD's, whereas we do not, so it is good to make sure we have all calibrations without any binning of the CCD's if something is going to be shared! The OIII/cont filter is also non-standard, so it is unlikely that the skyflats could be shared. observing conditions -------------------- This program benefits from good seeing conditions. However, even if the seeing is bad, there is nothing else that could be done except for continuing with the normal observing program. The same basically holds for thin cirrus. In cloudy conditions, the observations would not be of much use (we have specified 'CLR' clear observing conditions in the proposal, and seeing better than 1.2", although the seeing is not a hard limit). setup ----- It is very very important to not bin the observations, i.e. the CCD binning should be set to 1. Otherwise we loose half of our accuracy! We use the full field that is covered by the 2 CCD's, so no windowing. The astrometric observations are all done in the OIII/cont. filter. A sequence of 8 consecutive observations is specified for each night, and again another set of 8 observations in the second half night. Both 16 observations will be averaged and used together as one astrometric observation. So in principle more observations are better (if there is extra time), but a few less (like 5 instead of 8 per set) are also okay if really pressed for time. For the target stars, the integration time in the OIII/cont filter is 2 minutes. In normal conditions, it should be just possible to get 8 observations for all 10 stars during a half night, plus the cluster in the beginning and the end. The cluster, for calibration purposes, is also observed in B and V filters (just one observation each), and a sequence of 5 observations in the OIII/cont filter. The OIII/cont observations of the cluster will also be averaged. Integration are times are shorter for the cluster: just 1 second in the B and V filters, and 15 seconds in the OIII/cont filter. It is also important to place the same stars on the same pixels every time. So the move_to_pixel template is executed first, and one has to select - by hand - the star which should be placed on the specified target pixel. The target pixel we have used for the previous runs is '2650 2048'. I have modified the coordinates of the target stars a little bit to reflect this offset (from the middle gap), but the offset is not perfect yet. But it's reasonably close. It is a good idea to look at the finding chart each time and make sure one selects the right object. The target star is almost always the bright star in the center of the finding chart, but there are also cases where there are other bright stars in the field. One quick look at the finding chart, and the right object is selected. One might have to rotate the display of the acquisition image to match the orientation of the finding charts. For the cluster, the target star is HD70058, the one that is more to the west of the two bright stars, and the one in the center of the finding chart. The target pixel for the cluster is different; it is '2319 1585' (for historical reasons). It might be good to print out the finding charts, so that they are available at the telescope. We do not use jittering; that would impact the astrometric performance. material that should be attached: --------------------------------- finding charts OB's original proposal