ADONIS Observing Procedures

Infrared Observing Techniques

Adaptive Optics Terminology

How to take Sky Frames

Observing Overheads

Sumary of Observing Strategy

Data Management

Guiding and Finding Chart

Infrared Observing Techniques

The ADONIS observer should be familiar with the techniques of near infrared observations.

Sky Frames: The sky background is strong and variable in the near infrared. Therefore sky calibration is required using sky frames. Skys can be obtained in chopping mode (using the ON/OFF mirror) or in fixed mode (moving the telescope).
Flat Frames: These frames are required to calibrate pixel-to-pixel sensitivity variations for each lens and wavelength. Flats may be easily obtained by observing the twilight sky.
Dark Frames: These frames are used to remove dark current and exposure time dependent electrical offset effects. Darks should be obtained using exposure times that match those of the flats or of the science exposures.
Standard Frames: Observations of standard stars are efficiently obtained in chopping mode using the ON/OFF mirror. It is advisable not to use "shutter mode" for these frames.

Adonis Adaptive Optics Terminology

The ADONIS observer should be familiar with the following terms:

Science Object

The astronomical object which is under study.

Reference Object

This is the object used for the adaptive optics correction. It must be located less than 30 arcsec away from the science object. Even better, it may be the science object itself, providing its intrinsic angular size is smaller than 3 arcsec. The visual magnitude of the reference star can reach 12.5.

PSF Object

Because residual errors remain on the adaptive optics wavefront sensor, a deconvolution process should be applied. This is done using the point spread function (PSF) determination of a so called PSF star. In order to get the same correction quality with the PSF and the reference star, the PSF should respect a few constraints:

The spectral type of the PSF should be the same as that for the reference star (or as close as possible to it). B-V values should be very close.
The PSF should be as bright or a bit brighter than the reference star.
The PSF should be as close as possible to the reference star (a few degrees on the sky is a maximum). The ideal situation would be for it to be located in the same field as the science object.
The PSF should be observed frequently enough in order to match the changing sky conditions of the science object (airmass, seeing etc).

If the science object is not used as the reference object, then the best deconvolution results are obtained by using a binary for which the binary separation is the same as the separation between the science and reference objects. One binary component will be used as the reference and the other as the PSF. Both components should obey the flux and positional constraints listed above. In practice, you have to be lucky to find a suitable nearby binary.

Closing the loop

When the loop is closed, adaptive optic corrections are taking place. That is, the wavefront sensor is monitoring the difference between the actual atmospheric abberations and the most recent position of the deformable mirror.

There are a number of ways to close the adaptive optics loop. The most frequently used methods are as follows:

Close loop with modal optimization

Close loop with zonal optimization - This is much faster (around 1 minute) than closing with modal optimization. However the final correction is not as good. In zonal control, each zone or segment of the mirror is controlled independently by wavefront signals that are measured for the subaperture corresponding to that zone. If the object is a photometric standard or simply a test than the loop may be closed with zonal optimization. Note: For standards, most observers simply close the loop without optimization (see below).

Close loop without optimization - This method closes the loop without acquiring a new optimization. It requires less than 3 sec and is used, for example, when the telescope returns to a science object after a series of sky frames or when the object is a PSF or standard.

Close loop on the PSF - The correction for the PSF must be kept as near as possible to that made with the science object. To do this the telescope operator must adjust the filters and high tension on the wavefront sensors in order to achieve the same flux and noise as the previously observed science object. This takes about 2 mins. Then the loop is closed without optimization (see above).

The observer need only decide what method he/she would prefer. The telescope operator will execute the commands.

Correction of residual abberations

It is very likely that at some point during the observing night a further correction will have to be made for static abberations. Such abberations can appear when observing a scientific target and can change with airmass. To correct for them a series of adjustments are made (e.g. focus, astigmatism, coma, triangular and spherical) by the telescope operator. This process can take up to 15 minutes. The observer need only decide what method he/she would prefer. The telescope operator will execute the commands.

Image centre

Selecting the centre of your desired Adonis image can be confusing if you don't know the difference between the following centres:

Wave Front Sensor (WFS) centre

Infrared (IR) centre - This is the centre of the camera field of view and is defined by the observer using the ON/OFF mirror with a specified offset in arcsec. Usually this position corresponds to the centre of the desired science image.

The default configuration is for when the reference star is the science object. In this case the IR centre is the same as the WFS centre.

How to take sky frames

Moving the telescope

In this case the on/off (M6) mirror remains fixed. The telescope operator opens the adaptive optics loop and applies an offset to the telescope pointing. Once the sky frame is taken the telescope returns to the original pointing position and the loop is closed without optimization.

They are several reasons to use this mode:

If the source is extended or the field is crowded then a large positional offset is required for the sky frame.
This mode is the best way to remove the instrumental background because the same optical path on the bench is used.
This mode is the best way to take good quality sky frames using SHARP II+ with 100mas/pixel scale.
It may be necessary to use this mode when using the coronograph.
This mode is the only option for faint K-band imaging.

The disadvantage is that this mode has large overheads.

Moving the M6 on/off mirror

In this case the on/off mirror chops so that the camera field of view can be centred on different positions within the available field of view. Double or triple chopping is possible. The observer defines the frequency and amplitude of the chopping via the ADOCAM software.

They are several reasons to use this mode:

The adaptive optics system will keep working.
Shorter intervals between sky frames are possible.
Data management is better
ADOCAM's automated sequences and batches can be utilised.

The disadvantage is that the sky positional offset must be less than 60" from the reference source. Furthermore, the subtraction of thermal emission from the bench is poor, particularly in K-Band.

Example 1: The scientific target is the reference star for the Wave Front Sensor (WFS).

Example 2: The scientific target is distinct from the reference star.

Observing overheads

In order to plan correctly the time needed for an observation, the astronomer should be aware of the following overheads:

15 minutes at the very beginning of the night for correction of residual abberations.
5 minutes overhead time between each object. This includes the time needed to point to a new object and to center it on the wavefront sensor.
Time to close the loop. For time estimates see the section on closing the loop.
5 minutes overhead time for changing wavefronts sensors. It is best to keep the changes to a minimum (i.e. group science objects according to their brightness).
Time for sky frames. This will depend on what method you use (i.e. with the on/off mirror or by moving the telescope) and on how stable the sky is. Our experience shows that on average the sky is stable over 40 minutes in J- and H- band and 15 minutes in K-band. Of course these numbers may be much smaller if the sky conditions are not stable (i.e. bad seeing, cloud, rapid transit of source).
Time for observations of PSF stars.
Time for observations of standard stars.

Summary of Observing Strategy

The image quality is limited by various parameters that do not follow the same behaviour.

In J band, the background photon flux and noise is negligible over a time-scale of 1 min, but the PSF is highly variable due to partial AO correction in this band.
In K band the opposite occurs: the PSF is much more stable but the background emission is higher and induces shorter time sequences.
In L and M band the strehl ratio might be as high as 60-70% leading to a very stable PSF. However the integration time cannot be longer than 1.4 or 0.36 sec (for L or M repectively) because of background radiation on a very short time scale.

For more information see the article The ESO/ADONIS system: 3 years experience in observing methods by Mignant et al.

Data Management

Image Format

Data are written to one of two disks (/s3 or /h1) on the ADOCAM computer.

Data do not entirely follow the FITS standard. The dimensions of each image are 256x257 pixels. The extra 257th column is used to write the UT time of each individual frame. To reformat the ADONIS FITS to a standard FITS, you can use the ECLIPSE refits command.

The output images are cubes. Each cube contains data from 1 observing cycle, with each plane corresponding to an individual frame.

Here is an example of an Adonis FITS header.

Image Name

At the beginning of the night the observer must enter a name and a nickname which is 3 letters long (e.g. xxx). A directory named xxx_yymmdd_hhmmss.DIR will be created (where y, m, d, h... stand for year, month, day, hour...). This directory will contain a logfile (xxx_yymmdd_hhmmss.LOG) and the raw data files (xxx_yymmdd_hhmmss.FITS).

Observing Logs

For each cube, all the settings of the detector are logged automatically in a file. The settings include integration time, lens scale, filter, observing mode, sequences and ON/OFF position. The user has the option to include the target name and coordinates.

The detector has no communication with the telescope and so does not know what the airmass, seeing or coordinates are. Therefore we suggest that the observer fills out an Adonis log sheet (provided).

The telescope operator will keep a log of the adaptive optic corrections made for each observed source. This log includes the flux and noise measured on the wavefront sensor, the estimated strehl ratio and seeing, type of optimization used and the parameters r0 and lcor. If you would like a copy of this log then please inform the operator.

Data Backups

Each day the operations staff write the previous night's data to dat tape using WDAT. Two copies are made. One copy is for the observer and the other is for ESO.

At the end of each night the observer must list (on a form provided) what directories are to be backed up. If this is not done then the operations staff will not know what directories to backup.

Once the observer receives their dat tape it is essential that they verify its contents as soon as possible. In most cases their data will be deleted from disk before the start of the next night. The dat tapes must be read with WDAT. This is part of the ECLIPSE data reduction package and is available on all machines in the Observer's User Room.
Examples:
With a DAT driver called /dev/nrst0
To write in the file "listing.lis" the content of the DAT tape:
wdat -d /dev/nrst0 -l= listing.lis

Note: Under no circumstances can data be transferred from the ADOCAM computer to another work station during data acquisition. This will cause ADOCAM to crash and halt observing.

Finding chart & Source List

The observer should have finding charts for each science object as well as for any PSF or standard stars. The telescope has a centre field camera with a field of view of 2.5 x 2.5 arcmin. The display is orientated North (down) and East (left).

It is possible for the telescope software to use a source catalogue file. The file must have the extension.cat and the following format:
e.g.
source 033603.2 162802.6 2000.0 2000.0 0.0 0.0 m
where "source" is the object name, "033603.2" is the RA, "162802.6" is the Dec and "2000.0" is the epoch. The line must end with "0.0 0.0 m" which sets the proper motion offsets to zero. The separations must be spaces and not tabs.

This file can easily by ftped to the telescope computer at the beginning of the night.