I had an invitation from the Gemini South Associate Director Phil Puxley to join the Science Staff of the observatory in semester 2001B, to help in several aspects:
 

Commissioning of the telescope;

Commissioning of some instruments, the Acquisition Camera, Abu and Flamingos-I;

Processing of near infrared Abu data;

Phase II planning on scheduled programs;

Acquisition and processing of data for the queue Flamingos-I programs.

 

Telescope Commissioning

During the commissioning of the telescope I worked on several tests, whose main goals are to put all the telescope systems working properly and reach optimal performance. Some of these tests are:
 

• Tracking test - We take several short exposures of bright stars to test tracking with and without fast guiding, which is the secondary mirror tip/tilt corrections. The stars must have different zenith distances and azimuths, so that the test is performed with various tracking speeds. The tracking shall be within 1 arcsec without tip/tilt and 0.1 arcsec with tip/tilt. A critical time happens when the star is near transit, specially when it is coming down in elevation.
• Pointing test - We take one short exposure for each star in a large list. Typically around 40 objects, but some tests with around 80 stars were performed. This is to check the pointing accuracy.
• Probe mapping - The Gemini telescopes are provided with 2 Peripheral Wave Front Sensors (PWFS) probes, assigned to guide stars, and responsible for fast guiding, active optics (aO) and/or adaptive optics (AO). The probe mapping test verifies the position angles of these 2 probes, and their behavior, when the science object and guide stars lie in many different positions in the focal plane.
• Wave front sensing test - The Gemini telescopes also have a High Resolution Wave Front Sensor (HRWFS) which provides a more detailed response than the PWFS. This sensor is used to set up the primary mirror figure in the beginning of each night. However, the HRWFS is located on-axis (while the 2 PWFS are off-axis), so that it will not be used while doing science observations. It is necessary to check and compare how the HRWFS and the PWFS are correcting the primary mirror figure through the actuators. Both must apply the same correction in order to maximize image quality when doing science. Thus, it is necessary to model the response of the PWFS, compared to the response of the HRWFS, for different elevations, probe arm angles and guide star magnitudes, for instance. This modeling uses the Zernike polynomials, and the astigmatism correction was one of the main concerns. This is a very important test, since only with the WFS working properly Gemini will provide excellent image quality, even comparable to the image quality attained in HST observations, with the advantage of a very much larger mirror ground based.

Instrument Commissioning
 

• Acquisition Camera: several tests were performed during the commissioning of the AcqCam. Some of these are:

 
• dark stability - several dark frames are taken with exposure times varying from 0.005 to 600 seconds, and CCD temperature varying from -31C to -29C. This is to check for drift and light leak.
• blank sky flat fields - to check fringe correction.
• saturation behavior - exposures of a bright (V~5) star are taken with integration time varying from 0.1 to 60 seconds to get the whole CCD saturated. Then 0.01 second exposures of a fainter star (V~7) are taken until the image looks ok.
• world coordinate system determination - imaging of dense astrometric fields.
• sky brightness monitoring - 2 or 3 images in each filter (UBVRI) of standard stars are taken at different air masses and lunar phase.
• linearity - a standard star is used to take individual images with increasing integration time until the whole core of the PSF saturates.
• charge transfer efficiency - using a standard star, images are taken with the star positioned at the center of the CCD and close to each corner. The integration time is varied to have a sequence of images with around 10000 counts and other sequence with around 100 counts.
• photometric calibration - 2 or 3 images of a standard star field are taken at several zenith distances in all filters, with and without the neutral density filters.

• Abu: the nights we have worked with Abu were not properly a commissioning time. Actually, this instrument was used to test the telescope systems and the acquisition and reduction of near-infrared data, as well as to imaging of several interesting objects for the Gemini South Telescope Dedication (see below). By the way, this instrument is a NOAO infrared imager that was used in observations in the South Pole.

 
• Flamingos-I: since it is not a facility instrument, but a visiting instrument, Flamingos-I commissioning started from scratch. A set of broken lenses during shipment, problems with the array detector, and a not so lucky week of bad weather, contributed to the delay of both the commissioning and science observations. Part of the work done now follows:

 
• determination of the science fold position;
• pupil alignment;
• check of the field orientation;
• check of instrument control, acquisition and dithering scripts;
• check of dark current at several exposure times and detector temperatures;
• check of image quality variations along the detector;
• determination of the instrument focus offset with PWFS1 and PWFS2.

Processing of Near Infrared Data

With Abu, several objects were imaged in the near infrared bands J, H, K, L, Br-alpha and Br-gamma. Among the very interesting objects there are star forming regions, galaxies, globular and open stellar clusters, and planetary nebulae. All these images were reduced under the Gemini IRAF packages QUIRC and GEMTOOLS. Reduction procedures include:
 

• determination of bad pixel masks;
• flatfield division with both dome and sky flatfields;
• sky subtraction;
• co-adding of dithered images.

Some of the images are:
 
 

The Galactic Center (at left) in Br-alpha, showing emission from hot gas, which will probably form stars or feed the Milky Way's supermassive black hole, and the Planetary Nebula NGC 6369 (at right) in K, result of the death of a star with a mass close to that of the Sun, ending its life as a red giant. The remaining star probably lies at the center as a white dwarf. Image quality is around 0.35 arcseconds in both pictures (Gemini Observatory/NOAO/AURA/NSF).

The Seyfert 1 galaxy NGC 1097 in J, imaged in a not so good night, with wind gusts making it a hard work for the telescope systems. But still, image quality is 0.5 arcseconds, which means 41 parsecs (H₀=75 km s^-1 kpc^-1). The size of the image above is around 30 arcseconds, which is only 5% of the galaxy size. The very bright and active nucleus can be easily seen, as well as its secondary bar, which could help feeding the nucleus activity (Gemini Observatory/NOAO/AURA/NSF).

Another very interesting Abu image (of the Circinus galaxy) can be found here.
 
 

The Phase II Process

After one submit an observing proposal (a process called Phase I), and before the observations take place, a series of processes occur, which are globally called Phase II. In Phase II, all the details concerning the observations of an approved observing program are revised by the contact scientist. Among these details, one can cite:
 

• exposure times and total number of exposures;
• dithering pattern;
• coordinates of the source;
• observing constraints (image quality, sky background, cloud cover, water vapor content and zenith distance);
• filters;
• slit width and position angle;
• wave front sensors stars;
• etc!

All this information must also be checked to be in agreement with what the PI (principal investigator) wants to do with the data acquired, what the PI is expecting from the data, and what is the science proposed to do with it. I had the opportunity to help in several approved programs with the Acquisition Camera and with Flamingos-I, including imaging and spectroscopy, and programs related to targets of opportunity.

One important thing to do in Phase II is use the Observing Tool (OT) to verify whether or not certain parameters are correct. The OT simulates an observation, as well as a whole observational program. The figure at left is a snapshot of one of the windows from the OT. It shows how the image of NGC 1097 above was acquired. At right, one can see the PWFS1 probe arm, the guide star used and the vignetting the arm produces. One must be sure that the vignetting does not affect the instrument's field of view. The square in the center is the field of view of the Acquisition Camera (2 x 2 arcminutes). The outer circle shows the outer limits for guide stars.
 
 

Data Acquisition and Processing for Flamingos-I Programs

That was the main reason for me to come help the science staff, namely, to acquire and reduce data for the approved programs for queue observations with Flamingos-I in semester 2001B. As said above, a number of unhappy facts prevented us to start the queued programs as scheduled, but finally these were started in the first week of October. Since in the commissioning only the imaging mode could be verified, all spectroscopy programs were excluded, which means four programs, since, with the shortened schedule, it was realized that only programs for bands 1 and 2 could be started.

In the observing runs I have participated, we have imaged distant galaxies (one at redshift 3!), dwarf galaxies, quasars, Lyman break galaxies, very faint stars in star forming regions, and halo white dwarfs. In the first runs we chose the most simple programs for us to be more familiarized with all the processes. In two or three nights we were already able to start more complex programs, involving complicated dithering modes and/or a high number of different objects and observations. Then we chose programs from their ranking band, observing constraints, and whether or not they have been started, need more observations, or have already being applied for the total allocated time.

Data were reduced on-line, either at the control room in Cerro Pachón, or at the computer room in La Serena. In the last case we were connected with Cerro Pachón through video conference the whole night. Obviously, since we were doing that for the first time for Gemini South, data processors could not finish to reduce one night's data at the end of the night! The pipeline reduction is still being developed, the tasks and scripts still being written, so there are still many things to be implemented until data processing can occur in a most efficient way. Data processors have four tasks to accomplish during the night:

• Keep track of the weather, by saving satellite infra-red images of northern Chile at the beginning, middle and at the end of the night. Also by saving a summary of the environmental conditions provided by CTIO (Cerro Tololo InterAmerican Observatory).
• Reduce baseline calibration and science data.
• Keep track of the image headers, correcting errors when they appear.
• Update the database, regarding observed programs, i.e., indicate what programs have been started, which observations were done, and whether or not data is usable.

Acknowledgments

I would like to thank Dr. Phil Puxley and the Brazilian National Gemini Office for this splendid opportunity, and all the Gemini South Science Staff for making my stay in La Serena a most fruitful and enjoyable time.