EUROPEAN SOUTHERN OBSERVATORY

La Silla Observatory

SCIENCE OPERATION DEPARTMENT

Project Title: TIMMI2 user's manual LSO-MAN-ESO-9XXXX-X/X.0

Prepared	Vanessa. Doublier	2003-01-01
Reviewed	M. Billeres, G. Lo Curto, U. Weilenmann, U. Kaufl, M. Sterzick	2003-06-01
Released	Olivier Hainaut	2002-07-31

Revision History

Contents

1. Instrument overview
   1. Optical layout
   2. Imaging modes
   3. Spectroscopic modes
   4. Polarimetric mode
   5. Detector and acquisition system
      
2. Observing at 3, 5, 10 and 20 micron
   1. Atmospheric bands
   2. Background emission
   3. Imaging
   4. Spectroscopy
   5. Influence of the moon
      
3. Observing at the 3p60m
   1. Visitor mode
   2. Service observing
   3. The telescope
      
4. Observing with TIMMI2
   1. Observation software
   2. Real Time Display
   3. Target acquisition
      1. Imaging
      2. Spectroscopy
         
   4. Maximum brightness of observable targets
   5. Calibrations
   6. Chopping
   7. Nodding
      
   8. Pipeline
      
5. Templates cookbook and overheads
   1. Templates general description and summary
   2. Source of overhead
6. Imaging modes
   1. Characteristic
   2. DIT and other Read parameters
      
   3. Pipeline
   4. Performance
      
7. Spectroscopic modes
   1. Characteristic
   2. DIT and other Read parameters
   3. Pipeline
      
8. Polarimetric modes
   1. Characteristic
   2. DIT and other Read parameters
   3. Pipeline
      
9. Filters characteristics
10. Atmospheric coefficients for La Silla in the Mid Infrared
11. Standard stars
12. Appendix
    1. A - Templates description
       
13. Acronyms
14. About this document...

1. Overview

TIMMI2 stands for Thermal infrared Multi-Mode instrument.

• Thermal infrared: it covers the 3.5 to 28 micron wavelength range
• Multi Mode: TIMMI2 allows to observe in spectroscopic and imaging modes, inc. in both cases polarimetric mode.

It is based on cryogenic technology associated with chopping/Nodding technics which allows observatins at these wavelength were thermal background dominate.

TIMMI2 can be considered as the precursor for VISIR and is a perfect complementary instrument since VISIR will have a smaller FOV and will not offer any polarimertry capabilities.

Back to Content

1.1 Optical layout

Fig. 1 shows the optical layout of the instrument

The optical layout of TIMMI2 is very compact, allowing the vessel to be placed at the Cassegrain focus of the 3.60m telescope at La Silla Observatory. The light path is the same in all three observing modes: imaging, spectroscopy and polarimetry at all wavelength. The mode selection is based on positions of the lens wheel and filter wheel which bear the objectives and the filters/grisms respectively. The pixel scale for spectroscopy is fixed as the order sorting filters for spectroscopy are built-in with the lens.

Back to Content

---

1.2 Imaging modes

In imaging mode, we propose 2 different pixel scales: 0.2 arcsec/pixel for both Q and N band imaging, and 0.3 arcsec/pixel for both L and M, and N band imaging.

Scales

	Nominal scale	Measured	full field of view
N band Imaging	0.2"/pix	0.202"/pix	64" x 48"
Q band Imaging	0.2"/pix	0.202"/pix	64" x 48"
N band Imaging	0.3"/pix	0.296"/pix	96" x 72"
L, M band Imaging	0.3"/pix	0.315"/pix	100" x 76"

Available Filters  

Filter Name	Blue cut (mm)	Blue (50%)	Red (50%)	Red cut (mm)
L	3.60	3.73	4.11	4.13
M	4.30	4.36	4.91	4.99
N1	7.79	7.99	9.20	9.46
N2	9.43	9.74	11.33	11.78
N7.9_OCLI	7.18	7.42	8.11	8.62
N8.9_OCLI	7.90	8.29	9.07	9.46
N9.8_OCLI	8.76	9.10	10.02	10.49
N10.4_OCLI	9.46	9.80	10.82	11.21
N11.9_OCLI	10.61	10.99	12.19	12.50
SiC	10.13	10.67	12.93	13.34
N12.9_OCLI	11.54	11.62	12.79	12.98
[NeII]	12.57	12.68	12.90	13.02
[SIV]		10.3752	10.5432
Q1	17.04	17.35	18.15	18.48
Q2 ( 18.75)	18.321			19.176

• Wavelength calibration has been measured with a monochromator in the lab under TIMMI-2 operation conditions

• transmission curves are available here

Wavelength calibration tables are available here

Atmospheric transmission values are available here

Available aperture sizes:

Slit with holes
Mask with holes
Slit Echelle	3"x50"
Slit Echelle	1.2"x50"
Slit	3" x 70"
Slit	1.2" x 70"
Field mask	0.3"/pixel
Field mask	0.2"/pixel

Back to Content

---

1.4 Polarimetry with TIMMI2

Polarimetry observations with TIMMI2 are performed using a wiregrid plate rotating inside the light path. All angles are  virtually possible. The sources are, in standard mode, observed at each of the following angles: 0 90 45 135.  The observation sequence can be: all angles for each NOD cycle, or one angle per NOD cycle. This is mandatory to observe at all 4 anlgle positions to measure the Stokes parameters and hence the degree and position angle of polarisation.

The Stokes parameters are then determined as follows:

Where i() is the intensity of the source which transmits light that is polarised at angle . We have assumed that the rotator is at a position angle of 0 degrees for the first measurement. This needs not be the case. The degree of polarisation and the polarisation angle are given by;     

To derive the correct value of , attention needs to be paid to the signs of and .

This algorithm neglects instrumental polarisation. Preliminary measurements with TIMMI2 are under investigation to determine the primary source of instrumental polarisation. In principle, all the imaging modes are available for Polarimetry.

Back to Content

---

1.5 Detector and acquisition system

TIMMI2 detector is a Raytheon 320x240 pixel array. It operates at 6.5-7.5K, and is controlled by the ESO-IRACE controller. Here follow the characteristics of the array:

Manufacturer	Raytheon
Part	CRC-774, 30183 science grade
Type	Si:As High background impurity band conduction
Format	320x240
Pixel size	50 micron
Wavelength range	2 to 28 micron
mean Q.E	31.6%
Read Noise	922e-
Full well capacity	1.13e7/2.04e7 e- (high/low gain)
Number of output chanels	16
Dark current	<100e-/sec at T=6K

The values are obtained from the manufacturer specifications.

The IRACE acquisition system is used to readout the array and pre-process the data. Amongst other tasks, it co-adds individual DITs, builds differences from the chop-cycles and transmits the INT frames to the instrument workstation. Note that the number of count is normalised to the DIT.
 

DIT and NCREAD

The DIT and NCREAD parameters cannot be defined by the user. They have been determined such as to avoid background saturation under average weather (CLEAR) conditions.

Readouts

For the moment, we only offer one readout mode: Single integration-while-read with chopping: the array is read, reset and read. The readout modes are presently not parameters defined by the users. 

Chopping

Chopping is the default mode of operation for 2-28 micron observations. Chopping is achieved by synchronizing secondary mirror of the telescope (M2) with the detector, and by subtracting the images from the respective beams. The result of a chopped image is therefore a background subtracted image with positive and negative images. We deliver one of the two half cycle frames for each chopped image (ie an ON frame averaged over the number of chop cycles), and the subtracted frame. These data are stored in a cube.

Windowing
Windowing is presently not offered.

Features

TIMMI2 array suffers from several features which can render image analysis difficult.

Like all IR arrays: TIMMI2 detector may suffer from background saturation (clouds, dome...)

In the following image, we show the effect of a bright star: this is called: smear and stripe pattern due to row and column level drop in the on-chip anlog circuitry.

2.1 Atmospheric bands

The atmosphere transmission varies with the wavelength. Fig 2, Fig 3.1 and Fig 3.2 show the transmission bands in the 3-5 micron range and in the 10-25 micron range.

Fig 2. Atmospheric transmission between 3 and 5 micron. In pink: L band, in skyblue M band. On TIMMI2, the L and M band filters have a slightly shorter wavelength coverage.

Fig 3.1: N band atmospheric band: N1 and N2 filters cover the whole band, overlapping at the ozone atmospheric absorption at 9.5 micron.

Fig 3.2: Q band atmospheric transmission: Q1 and Q2 roughly cover the whole band, overlapping at 19 micron. The transmission in this band is extremely varaible, and observations are very delicate. Observing in either Q1 or Q2 requires stable and clear conditions.

Back to Content

2.2 Background emission

The infrared is characterised by two regimes. Below 2.2 micron, the OH lines dominate the spectrum, while beyond 2.2 micron, the emission is dominated by the thermal contribution of the environment: atmosphere, telescope, dome...

In the MIR (7 to 30 micron), the background emission  will vary strongly with ambiant temperature and humidity:

Fig 4: Observed flux limits for TIMMI2 during 2002 with the technical grade (starting May 2002). The sensitivity variations are due to weather conditions. We also observe a large scale variation possibly due to seasonal weather patterns. For a complete set of sensitivity measures see:

/sci/facilities/lasilla/instruments/timmi/html/ImaMoni/TIMMI2_moni.txt

Back to Content

2.3 Imaging

Chopping

Chopping is mandatory at these wavelengths. The method consists of moving the secondary mirror once every second or fraction of seconds. The typical throw is about 20 to 40 arcseconds. Therefore, in some circumstances, the images corresponding to the two beams may overlap. An essential requirement of this technique is to combine chopping with telescope nodding (i.e. offsetting in the direction opposite/parallel to that of the chop). The chopped images usually leave strong residuals on the detector, that are due to the different optical paths of the two beams. These residuals subtract well between two chopped images taken with a telescope nod in between.

Nodding

While chopping involves the secondary mirror, nodding involves telescope offsets. Typically, a nodding throw is between 10 and 30 arcsec. There are two nodding methods: quadratic, i.e. perpendicular to the chopping direction, or triangular i.e. along the chopping direction. In the case of the triangular nodding, the nod throw is exactly the oposite to  the chop throw.

Jittering

Jittering can be used in complement with nodding and chopping. It is better used in case of extended sources, which nodded/chopped images overlap. It allows to create a mosaic of the extended field.

Photometric calibrations

Because the strong atmospheric absorption varies with the water vapor column density and the airmass, absolute photometric calibrations are difficult in the MIR. In order to obtain accurate flux calibration, standard stars have to be observed at similar airmass as the science target and within less than 3 hours. We maintain on the web page a list of both bright and faint MIR ISO standard stars.

Flat fields

Flat-fielding in ground based MIR does not provide any improvement in photometry accuracy. The effect is of the order of 0.1% overall, which is negligeable with respect to the photometric accuracy (~5-10 %) of the flux calibration. For your information you can have a look at the tests performed on flat-fielding procedures.  Flat fielding is not part of the standard calibration plan, hence it is  the responsability of the user to make the relevent observations (please contact TIMMI2 instrument manager ).

An estimation of the photometry accuracy can be obtained by the method of the 1000 points. The idea is to observe a star at various positions on the detector and construct a "illumination" map. This is easily done using the "jitter" template.

Back to Content

2.4 Spectroscopy

Nodding
The classical technique in spectroscopy is to observe object(s) at two or more positions along the slit. The sky is effectively removed by subtracting one frame from the other, registering the two beams and then subtracting again. This process is sometimes called double subtraction. If the field is crowded, or if the object is extended, then a blank sky may be necessary, and, in this case, the double subtraction is done slightly differently.

Spectrophotometric Calibration
Calibration of spectroscopic data in the IR is a complicated procedure that requires care. It is generally done in three steps. The first step removes telluric features, with what is commonly called a telluric standard; the second step removes the spectral features of the telluric standard that are imprinted onto the science spectrum because of the first step; and the third step sets the absolute scale with what one may call a spectroscopic standard. In general the spectroscopic standard and the telluric standard are the same star, but this does not need to be the case.

The most prominent feature in IR spectra are the telluric lines. Unfortunately, many of the telluric lines do not scale linearly with airmass, so it is necessary to observe a standard at the same airmass and with the same instrument setup as that used for of the science target. Furthermore, the strength of the telluric lines varies with time, so it is also necessary to observe the standard soon after or soon before the science target. On an average La Silla night, standard stars should be observed at interval of 1 hour.

The spectrum of the telluric standard is divided directly into that of the science target. Ideally, the spectrum of the telluric standard should be known, so that features belonging to it can be removed. However, this is never the case, so one has to use standards in which the spectrum is approximately known.

We usually use no later than M type stars with well determine synthetic spectrum. We use mainly the ISO bright star catalog. We also propose the faint star catalog, however, not all the stars have been accurately calibrated. In the near future, we will intergrate the MIDI catalog. The combination of these catalogs allows an acceptable sky coverage such that the user can find a good standard star close to the science target.

For absolute calibration, slit losses have to be estimated. This is usually a difficult task, as the spectroscopic standard (which is usually the telluric standard) and the program object may not be positioned at exactly the same place in the slit.

If the object is a point source, it can be assumed that the slit losses for the standard and the program object are the same. If the program object is not a point source, the slit losses have to be estimated on the basis of its morphology. It is always good practice to observe the spectroscopic standard through both a wide slit (e.g. 3 arcsec, or slitless) and the slit used for the program object, in order to get an estimate of the slit losses by comparing the spectra obtained with the two slit widths.

Another method consists of taking an image in a narrow band filters (11.9_OCLI and/or 12.9_OCLI) and extract the flux along the slit., and compare with the flux of the integrated spectrum with the filter passband.

However, the image quality is diffraction limited with TIMMI2 at the 3p6m telescope at 10 micron, for external seeing below 1.5 arcsec. So the loss is negligeable even with the 1.2" slit under average La Silla conditions.

Alternatively, if the broad-band magnitudes of the object are known, the absolute flux calibration can be derived by convolving the measured spectrum with the broad-band filter curves. In this case, the IR magnitude of the standard is irrelevant, only the spectral type is important. This works reasonably well at Low Resolution.

Locating the spectra and acquisition of faint objects

To ease the location of the spectrum, we recommend to take a thruslit image between the acquisition and the observation. This is possible using the "two in one" template TIMMI2_spec_obs_thruslit. The acqusition of a source brighter than 0.5Jy can be done directly on the RTD. The qcquisition is done therefore only the N1 filter. If the source is an emission line object, we recommand to do a few minutes image in a narrow band filterto deternine the position of the object on the field.

For faint objects, the acquisition cannot be done directly on the RTD. The strategy is to preset to the field, take a image of less than 10 minutes (if more is required, then spectroscopy may not be possible for such a faint target) to determine the position of the target on the detector and compute the "blind" offset to the position of the slit. However, this method is not entirely reliable, hence we advise the observer to prepare observations using reference stars within 3arcmin maximum of the scientific source. The procedure will be as follow:

1- the reference is a MIR source: preset direclty to the reference star (input coordinates in the OB), center the source on the TIMMI2 detector, then manually give the RA and DEC offset to the Telescope operator, then start observing.

2- The reference star is NOT a MIR source: preset to the reference star (input coordinates in the OB), center the star on the GP field (GP in center field position), then manually offset to the scientific object.

In the near future, we will implement the "offset from reference star" in the TIMMI2 acquisitions.

Wavelength claibration and flatfielding

Wavelength calibration with TIMMI2 is straitgh forward. The drift in wavelegth with time of the instrument is negligeable. ( < 1% / 6 month). The wavelength calibration file is available on the web, and is updated everytime TIMMI2 is open. Users cannot do their own wavelength calibrations.

Flat-fields are not provided for TIMMI2. Tests show that the improvement in the accuracy is less than the increase of noise due to the procedure.

Back to Content

2.5 Influence of the moon

The Moon is not a problem in the MIR, moonlight does not significantly affect the background. There is therefore no point in requesting grey or dark time. However, the scattered light from the moon in the optical may be an issue for guiding. There is no problem if the pointing is farther than 30 degrees from the moon. Since the observations will be scheduled during full moon, please make sure that when receiving your observations dates that your favorite object is NOT behind the moon. If this is the case, please contact visas@eso.org as soon as possible, for re-scheduling. 

Back to Content

3. Observing at the 3p60m telescope

3.1 Visitor mode

Visitors arrive on La Sillal 1 day ahead of their observing run (36 hours) and receive support from La Silla Science Operations (ls-sciops) to prepare their OBs. Users are requested to read the P2PP and TIMMI2 web pages (inc. Last updates) before arriving. During the night, users do not have direct interaction with the instrument and the telescope. The execution of their OBs is undertaken by the Telescope Instrument Operator.

Dealing with the moon
If the object is very close to the moon (less than 20 degrees away), moonlight can prevent the telescope guiding system from working efficiently. The effect is difficult to predict and quantify as it depends on too many parameters. Visitors are encouraged to carefully check their target positions with respect to the Moon at the time of their scheduled observations (http://www.eso.org/observing/proposals/skycalc.html). Backup targets are recommended whenever possible, and users are encouraged to contact ESO in case of severe conflict (i.e. when the distance to the Moon is smaller than 20 deg).

Back to Content

3.2 Service Observing

Service observing at the 3p6 telescope with TIMMI2 is, for the moment only offered as "delegated observing". For details about the procedures and HOWTO: see the La Silla service observing web pages. No observations will be performed unless the procedures have been followed.

Back to Content

3.3 The 3p6m Telescope

Telescope control
Most interactions with the telescope consist of telescope presets for acquisition, telescope offsets during observations, and M2 chopping. Offsets (i.e. less than 1 arcminute) are usually completed in 4-5 seconds of time.

The guide star is used to recover the object exact position on the detector, which is mandatory for proper co-addition of the frames  between nodding cycles.  

Other important facts are:

• Defocusing the telescope is not recommended.

• Scanning along the slit in spectroscopy during one exposure can be done by setting an additional velocity to the telescope tracking at the level of the OB.

• The telescope pointing accuracy is of the order of 5-10 arcsec within the entire sky. However, to increase the accuracy, as the telescope offsets with very good accuracy (depending on the distance of the reference star to the obejct), we advise the users to select a reference star within a distance of 10 arcmin fromthe scientific target. However , precise astrometry (better that 0.5 arcsec) cannot be achieved at the 3p6m telescope.
  

Guide Stars
Guide stars are searched within the GP FOV (10 x 10 arcminutes) by the Telescope/Instrument Operator. As MIR observation most of the time involved looking at dark regions in the sky, it may be impossible to find a proper guide star. Please, before requesting preset to a target, make sure that a suitable guide star can be found (brighter than 12th magnitude). If observations have to be done without guiding, please be aware that the 3p6m telescope drfits by 1"/30min. Guide stars will have to be brighter than 12th magnitiude in R. If the MIR is bright enough in the optical, it will be used as a guide star through the dichroic (3x3 arcminutes) .

Flexures and tracking stability
The flexure of TIMMI2 at the detector plane is negligeable. In most circumstances, the image stability of both telescope and instrument is so good that there is usually no need to reacquire the target during long integrations (up to two hours) in spectroscopy. 

Chopping and Rotation

Chopping angle: is not yet operational. Il will be implemented as from P72.

Rotation on sky (imaging and slit position): not yet implemented.

Back to Content

4. Observing with TIMMI2

4.1 Observing Software

OS is the high level software controlling the instrument. It has its own GUI which allows one to access all instrument parameters. Figure 4.1 shows the TIMMI2 OS GUI. The users only use templates to control the instrument, and therefore have no direct interaction with OS. However, the OS GUI is useful for the visitors as a status display panel, displaying all information from instrument, detector and telescope.

Fig. 4.1: Observing Software Graphical User Interface: this panel is passive. It displays all the necessary information for both the visitor and the TIO. The status for the telescope/Instrument and Detector are shown as well as a few informations about the system.

Fig 4.2: BOB is  the Broker for Observation Blocks. Its function is to gather the observation informations and relay them via the OS. Its versatility allows to stop during the execution of an OB between the templates, as well as to skip some templates.

Fig 4.3: P2PP is the Phase II Proposal Preparation. At the telescope, its function is to help the observer define his/her next observation. It is very versatile and allows realtime decisions. It contains: Preset to the target, acquisition of the target and observation of the target.

Back to Content

4.2 The Real Time Display (RTD)

The Real Time Display is central to observing with TIMMI2. Like a video camera, every frame taken by the detector is continuously displayed on the RTD. It is important to realise that the continuous display of images on the RTD is not related to saving the data to disk. An image is stored to disk only if the adequate action is taken to do so, i.e. when a `Start Exposure' is sent. This is what the templates do.

The RTD provides a number of tools for measuring statistics, measuring the position and FWHM of objects in the field, and for storing an image to be subtracted from incoming images. This latter tool is referred to as store a fixed pattern, and is very frequently used during acquisition, quality control, etc.

Fig. 5: Real Time Display.

Note: The INT frames do not indicate when saturation of the array occurs. To check on the background level: select HCYCLE1 or HCYCLE2, or check on the oscilloscope.

Back to Content

4.3 Osciloscope

• Allows to check for saturation of the array in non-optimum sky condition
• chop throw monitoring

Back to Content

4.4 Target Acquisition

4.4.1 Imaging

The pointing accuracy of the 3.60m telescope is good, but  usually a blind preset to the field is not sufficient in imaging if the source is not bright enough to be seen on the guide probe FOV or on TIMMI2.  

For a fine pointing, so as to position an object in a particular region of the detector, they should use TIMMI2_img_acq_MoveToPixel. This template provides an interactive tool to define telescope offsets.

For blind acquisition (faint sources), the user should use the simple preset acquisition: TIMMI2_img_acq followed by a short imaging sequence (< 10 min) to determine the position of the source on the detector and re-align if necessary by offsetting the telescope. If the source is too faint, then the user should provide a reference star that should be seen at least on the GP FOV and through the dichroic (source brighter than 11 mag) , and then give the telescope operator the offset to the science target.

Observers in service mode shall provide, together with their OBs, all necessary information regarding the centering of the field if they have special requirements.

Back to Content

4.4.2 Spectroscopy

Blind centering of objects in the slits based on coordinates is not supported. Neither the pointing accuracy of the telescope, nor the coordinate accuracy of most targets would guarantee that the objects go straight into the slit.

It is mandatory to use the TIMMI2_spec_acq_MoveToSlit acquisition template in all spectroscopic program OBs and to use the same slit in both the acquisition and observing templates. If a different slit setting is required, then a NEW acquisition (without preset) should be done.

This template provides interactive tools to make telescope offsets to center objects into the selected slit. 

 Acquisition can be done either on the target itself or on reference targets. The object used for acquisition has to be brighter than approximately 500 mJy when the acquisition is done with the MIR Broad Band filter N1. Exceptions will be tolerated for moving targets, and special situations will be evaluated on a case by case basis.

If the object is bright enough so that an image can be obtain within 10 minutes of exposure time, the user should image the object first using: TIMMI2_img_acq and TIMMI2_img_classic templates. Once the source is detected, it can be centered on the slit manually using telescope offsets.  

When the science target is fainter, the procedure for acquisition should rely on reference objects which are brighter than this limit. These reference objects will be used for initial centering on slit, followed by a blind offset to move the target into the slit. 

These reference objects should be stars or point-like objects. Blind offsets from a reference object should be limited to approximately 5 arcminute.  To minimise offset errors, users should use reference objects that are as close as possible to the target, rather than trying to use the brightest reference objects.

In service mode, it is mandatory that all the relevant information is given in advance to the operation staff. This information should consist of:

• Finding charts, with clear definition of the field orientation and of the scale

• Overlay of the slit

• Clear identification of the object

• Clear identification of the reference object to be used for preliminary slit centering, if applicable.

• The offset to be applied between the reference and the target, if applicable. 

Thruslit image

For sources that are bright enough such that a detection can be obtained, it is possible to get a thruslit image using TIMMI2_img_thruslit.  The template includes the spectroscopic science template.

See Phase II Proposal Preparation: P2PP for more detailed information on the format of the finding charts and README files to be provided at the time of OB submission. Should this detailed information be missing, the observations will not be scheduled.

Back to Content

4.5 Maximum Brightness of Observable Targets

Artefact caused by a bright source in imaging mode. This a due to column/row levels drops in the on-chip analog circuity. The minimum integration times are fixed and will not be modified. Hence the users should be careful in the choice of their target.

In general, any sources brighter than 2000Jy will produce strong residual artefact. These artefact can be removed following the procedure given by: Hony et al., 2001 Astronomy and Astrophysics Vol. 377 page 1.

Back to Content

4.6 Calibrations

We do not provide a specific calibration plan.

1. Flat fielding procedures: /sci/facilities/lasilla/instruments/T2_Flats/timmi_flat.html 

2. Wavelength calibrations: TIMMI2 wavelength calibrations are stable and are only updated after TIMMI2 has been opened: /sci/facilities/lasilla/instruments/timmi/html/calib_waves.txt

3. Standard starts: we provide a list of bright standard stars compiled from the ISO standard stars list. At the telescope, IDL scripts allow to compute the sensitivity as function of FILTER in imaging, and as function of WAVELENGTH for spectroscopy.

Back to Content

4.7 Chopping

Chopping is the only offered mode. The basic characteristics and definitions of chopping are:

• The chopping throw is the distance between the two beams. The maximum chop throw is 100 arcsec.

• The chopping angle cannot be changed yet. It will be available as from P72.

• Guiding is done on either side of the beam for the TIMMI2_img_obs_classic and TIMMI2_spec_obs templates. For the other observing templates, guiding is performed on only the central (ON) beam. For chopping amplitude larger than 40 arcsec, overheads increase because more DIT have to be skipped. The transistion time increases from 20ms to 30ms. Thus we generally recommend that the chopping throw be kept as small as possible.

• The chopping frequency is automatically defined in the templates and is based on the instrument mode in use. It typically is around 6Hz.

• Chopping is associated with nodding in the opposite direction of the chop for TIMMI2_img_obs_classic and TIMMI2_spec_obs. In all templates, the nodding frequency is also automatically defined in the templates to give optimum performance for each instrument mode.

• The M2 chopping is synchronised in a transparent way with the detector acquisition system.

• One chop cycle corresponds to one ON-OFF cycle, i.e. one period of the M2 chopping motion. Both the first half cycle and the subtraction frames (ON-OFF) are saved in the resulting frame.

• Several chop cycles are averaged by the pre-processor to deliver only one image: it is the video frame.  For the observing templates, this parameter is automatically set.

• The data are stored in a cube. There is one cube, containing on half cycle frame on the first plane, and the subtracted frame on the second plane. This change is transparent at the template level for the user. The pipeline has been updated accordingly.

  The subtracted ON-OFF image can still be displayed on the RTD,as well as the half cycle frames. Storing the half-cycle frames allows checking of the sky background levels to operate the detector at the optimum level ranges, and makes it easier to apply any non-linearity corrections during data reduction. 

• DIT and NDIT are not parameters of the templates, as they are automatically set to the optimal values imposed by the chopping frequency and saturation levels.

Back to Content

4.8 Nodding

Nodding should be associated with chopping. The chopped imaged are left with a strong sky background residual which are due to the difference in light path between the two beams. Usually, the best way to reduce the residuals is to NOD perpendicular to the nodding direction (Figure 2). However, because sources in the the MID infrared are faint, it is also possible to NOD parallely to the chopping direction: this is used by the spectroscopic template and the classic imaging template (Figure 1).

Figure 1: Chopping/nodding pattern AB for both classic and spectroscopic templates. The NOD amplitude is the same as the CHOP amplitude and the telescope nods opposite to the chopping direction of the M2. After pipeline reduction, the positive image contains half of the signal while the side negative images contain only a quarter of the signal each. Hence to recover all the signal, one has to extract all three images and a combine them. Only the chopping amplitude is user defined.

The classic template is recommended usually foe faint and/or isolated small objects.

Figure 2: Chooping/nodding pattern of the smallsource template. The nodding is perpendicular to the chopping direction. The amplitude is user defined. After pipeline reduction, each of the positive and negative images contains a quarter of the signal.

The Small source mode is recommended for bright and/or large objects. It is also more sensitive for intrermediate brightness objects.

For Extended regions, we recommend to use the mode jitter which allows to scan across the extended regions.

Back to Content

4.9 Pipeline

TIMMI2 pipeline at the telescope provides to the user an excelent quality RealTime quick look tool. At the moment, only simple classic/smallsource imaging and spectroscopy templates are supported. Real time decision can be taken on the fly and  with confidence.

The first window, pre-processor, displays individual video frames. The second window, post-processor-1, displays the co-averaged video frames for each NOD position as the video frame are coming, and the third window, post-processor-2, displays the subtraction between the NOD positions. The resulting reduced image is of excelent quality (detections are real), and the users may use the frames reduced by the pipeline at the telescope for scientific analysis. However, the pipeline will not produce a final frame (post2) if the template has been aborted. The intermediate frame can be retrieved in the directory timmidata/tmp/post2. This subdirectory is not saved during the backup operations, since by default we do not guaranty REDUCED files, only the RAW data. Here follows what to do:

In an xterm on w3p6tnt:

inmidas -p xa

midas> chan/dir /home/observer/timmidata/tmp/post2
midas> $ls -ltr

- check the last file created with the.bdf as extension (exemple: 307012344.bdf)

midas> outdisk/fits 307012344.bdf 307012344.fits
midas> bye

observer> cd timmidata/tmp/post2
observer> mv 307012344.fits../../final/cube/result/

This will allow the file to be saved in the reduced data backup.

The signal in each reduced frames is (like in the individual video frames) co-averaged, hence corresponds to a DIT. The S/N however shall be computed using the total integration time.

Back to Content

5. Templates cookbook and overheads

5.1 Templates general description and summary

The instrument, detector and telescope are controlled by Observing Blocks (OBs), which are made up of templates. Templates are divided into three categories: acquisition, observation and calibration.

Usually, OBs consist of an acquisition template and one or more observation templates for science frames, and one or more calibration templates for calibration frames.

One (and only one) acquisition template is allowed in an OB, and therefore only one preset on sky. It is not possible, for example, to group in the same OB, observation templates on the science object and calibration templates on a standard star.

Table 1 provides a short summary of the templates offered for period P71. These templates should cover most needs.

Should observers who have observing time with TIMMI2 consider that these templates do not cover their needs, they can contact the LS-INFRARED group (ls-infrared@eso.org) well before the observations start.

The template parameters are extensively described in appendix A.

Back to Content

5.2 Overheads

Overheads

Overheads should be included in the requested observing time section of the proposals. Hence, special care should therefore be taken when estimating them.

The execution times reported by P2PP are not accurate, and users should refer to this manual for calculating the overheads. Users, especially those in service mode, should check and make sure that the overheads have been taken into account.

Preset and acquisition
The overhead for preset and acquisition depends on the template, as shown in Table 2. These times include the telescope preset, acquisition of the guide star, and acquirering the target if necessary.

Slit acquisition in spectroscopy
This depends on the brightness of the object to be centered on slit. The details are given in Table 2.

Detector overheads
For the templates that involve chopping, it is currently not possible to provide clear guidelines as to how to compute the overheads, since these templates are still subject to slight modifications. One should assume that all the templates have overheads of 50%.

Table 3 gives the overheads per operations. They are given assuming that the source is bright enough to be seen on the TIMMI2 detector, or that blind  pointing from reference objects is done.

Exemple for "60 minutes on source" imaging classic, object centered using reference star.

Preset: 6 minutes
instrument setup: 2 minutes
NOD time: 3 minutes
each "NOD" is divided in video frames of DIT*NCREAD*NCHOP*2 (in N1 band: 20.83ms, 3, 60)
overhead per video frame: 1*DIT
number of NOD cycles: 10
number of telescope offsets: 10
=> 3 min = 7.5s * 24 frames.
total time: 1 NOD: ((0.02083*(3+1)*60*2)*24)*1.4 = 336sec

Total Telescope time: 10*2*336sec + 6min + 2 min = 120min = 2hrs

Back to Content

6. Imaging modes

6.1 Characteristics

See section on Imaging Modes for a description of this mode.

Chopping is essential for observations with wavelengths longer than 3 micron, and is used for all observations. Chopping will always give a better sky subtraction. For more information about chopping see section about chopping. The chopping templates produce a series of data cubes for each nod position called video frames, which contains one half-cycle frame and the subtracted frame.

Back to Content

6.2 DIT and other Read parameters

In TIMMI2, reads paramters are as follow:

1. DIT: detector integration time
2. NCREAD: number of reads during one half chopping cycle, it is similar to the NDIT of the other ESO NIR instruments: SOFI/ISAAC/NACO
3. NCDUMMY: because the chopper needs to stabilise, we drop a number of DIT. They are called dummy reads
4. NCYCLE: number of chopping cycles

Hence the total ellapsed time for 1 video frame:

DIT*(NCREAD+NCDUMMY)*NCYCLE*2

The effective exposure time:

DIT*NCREAD*NCYCLE*2

These READ parameters, though are not user defined. They have been determined as function of the instrument setups to avoid sky saturation and un-reasonably large overheads.

Back to Content

6.3 Pipeline

Imaging templates are all supported by the TIMMI2 pipeline. Data fro aborted templates are not fully reduced. Intermediate reduced frame can be found locally (at the telescope) in the timmidata/tmp/post2 directory in MIDAS bdf format.

Back to Content

6.4 Performance

The user should refer to the ETC for estimating the performance of this mode: /sci/facilities/lasilla/instruments/timmi/ExpMeter/timmi2exp.html

Back to Content

7. Spectroscopic modes

7.1 Characteristics

See section on Spectroscopic Modes for a description of this mode.

Chopping and nodding are essential for observations with wavelengths longer than 3 micron, and is used for all observations. Chopping/Nodding will always give a better sky subtraction. Nodding and chopping are both done along the slit orientation: i.e. North/South.Nodding is done opposite to the chopping position, and is of the same amplitue. The templates produce a series of data cubes for each nod position: called video frames, which contains one half-cycle frame and the subtracted frame.

Back to Content

7.2 DIT and other Read parameters

In TIMMI2, reads parameters are as follow:

1. DIT: detector integration time
2. NCREAD: number of reads during one half chopping cycle, it is similar to the NDIT of the other ESO NIR instruments: SOFI/ISAAC/NACO
3. NCDUMMY: because the chopper needs to stabilise, we drop a number of DIT. They are called dummy reads
4. NCYCLE: number of chopping cycles

Hence the total ellapsed time for 1 video frame:

DIT*(NCREAD+NCDUMMY)*NCYCLE*2

The effective exposure time:

DIT*NCREAD*NCYCLE*2

These READ parameters, though are not user defined. They have been determined as function of the instrument setups to avoid sky saturation and un-reasonably large overheads.

Back to Content

7.3 Pipeline

All spectrocopic templates are supported by the TIMMI2 pipeline.

8. Polarimetry

8.1 Characteristics

See section on Polarimetric Modes for a description of this mode.

Chopping and nodding are essential for observations with wavelengths longer than 3 micron, and is used for all observations. Chopping/Nodding will always give a better sky subtraction. For more information, see section about chopping/nodding. The templates produce a series of data cubes for each nod position called video frames, which contains one half-cycle frame and the subtracted frame.

There are two different templates proposed. Both templates allow to observe at the 4 linear angles: 0, 90, 45 and 135 degrees. The difference is how the "nodding" is done with respect to the polariser angle. Either the nodding (always perpendicular to the chopping direction) is done once all angles have been observed at one position, or the nodding is done for each angles.

For the time being, polarimetry is offered only for imaging modes. Spectro-polarimetry is foreseen for P73.

Back to Content

8.2 DIT and other Read parameters

In TIMMI2, reads paramters are as follow:

1. DIT: detector integration time
2. NCREAD: number of reads during one half chopping cycle, it is similar to the NDIT of the other ESO NIR instruments: SOFI/ISAAC/NACO
3. NCDUMMY: because the chopper needs to stabilise, we drop a number of DIT. They are called dummy reads
4. NCYCLE: number of chopping cycles

Hence the total ellapsed time for 1 video frame:

DIT*(NCREAD+NCDUMMY)*NCYCLE*2

The effective exposure time:

DIT*NCREAD*NCYCLE*2

These READ parameters, though are not user defined. They have been determined as function of the instrument setups to avoid sky saturation and un-reasonably large overheads.

Not all filters have been defined yet. Only the most "user requested" are available and for which Polarimetric standard stars have been measured.

Back to Content

8.3 Pipeline

Polarimetric templates are not yet fully supported by the pipeline. However, it will treat the coming images are normal imaging frames, hence the resulting frame can be used for comparision of observation without the polariser in the same band.

9. Filters characteristics

Here are the different filter characteristics.

Ambient Temperature	77K
7.9mu	7.9mu
8mu - N1 filter
8.9mu	8.9mu
9.8mu	9.8mu
10.4mu	10.4mu
10.6mu - N2 filter
10.9mu - N filter
11.9mu	11.9mu
12.9mu	12.9mu

20 mu Order disperser	20mu Filter
BaF2	CaF2

Back to Content

10. Atmospheric Coefficients in the MIR at La Silla

N band atmospheric coefficient tables at:

Zenith, 30 deg., 60deg.

Q band atmospheric coefficient tables at:

Zenith, 30 deg., 60 deg.

Back to Content

11. Standard stars

We propose two sets of standard stars derived from the ISO catalogs.

1. ISO bright stars catalog: Selection of calibration stars. Parameters RA, Dec, visual magnitude (V mag) and spectral type have been taken from SIMBAD.  IRAS 12mm fluxes are to be used as a guide only. For more accurate photometry see Table 2 below. The selection criteria are:
• Flux at 12mm (IRAS) > 5Jy
• No M spectral types
• No Variable stars
• Photometry should be fitted by a model spectral energy distribution to within 3 percent

In some case, the selectrion criteria have been weakened so to obtain a relatively good RA coverage.

HD	RA(2000)	Dec(2000)	V mag	IRAS 12 um (Jy)	Spectral Type	Other names	Notes
1522	00: 19:25.674	-08: 49:26.14	3.56	13.09	K2III	-	-
4128	00: 43: 35.372	-17: 59:11.77	2.04	39.31	K0III	-	V?
6805	01: 08: 35.396	-10: 10:56.17	3.44	13.55	K2III	-	V
12929	02: 07: 10.407	+23: 27: 44.72	2.00	77.8	K2III	Alpha Ari	E, V
29139	04: 35: 55.240	+16: 30: 33.49	0.85	699.7	K5III	Alpha Tau	V
29291	04: 35: 33.039	-30: 33: 44.43	3.82	6.76	G8III	-	-
32887	05: 05: 27.660	-22: 22: 15.72	3.2	56.8	K4III	Epsilon Lep	V
37160	05: 36: 54.388	+09: 17: 26.42	4.09	6.70	G8III-IV	-	-
47105	06: 37: 42.701	+16: 23: 57.31	1.90	5.04	A0IV	-	-
81420	09: 25: 24.036	-05: 07: 02.62	5.60	7.41	K5III	-	-
81797	09: 27: 35.240	-08: 39: 31.00	2.00	157.6	K3III	Alpha Hya	E, V
110458	12: 42: 35.452	-48: 48: 47.19	4.67	5.83	K0III	-	E
108903	12: 31: 09.959	-57: 06: 47.56	1.63	865.4	M3.5V	Gamma Crux	V
123139	14: 06: 40.949	-36: 22: 11.84	2.06	38.52	K0IIIb	-	-
124897	14: 15: 39.672	+19: 10: 56.77	-0.04	521.04	K1.5III	Alpha Bootes	-
133774	15: 06: 37.600	-16: 15: 24.54	5.20	8.01	K5III	-	-
156277	17: 21: 59.478	-67: 46: 14.41	4.78	5.28	K2III	-	-
169916	18: 27: 58.240	-25: 25: 18.12	2.83	20.97	K1IIIb	-	-
175775	18: 57: 43.802	-21: 06: 23.96	3.53	12.85	G8/K0II/III	-	-
178345	19: 10: 01.757	-39: 20: 26.87	4.10	8.33	K0II/IIIC	-	-
187642	19: 50: 47.002	+08: 52: 05.96	0.77	21.86	A7IV-V	-	-
189831	20: 03: 33.459	-37: 56: 26.51	4.77	8.16	K5III	-	-
196171	20: 37: 34.032	-47: 17: 29.41	3.12	15.08	K0III	-	-

 

1. ISO faint stars catalog:Selection criteria:
   • Hipparcos coordinates
   • V < 7 mag
   • Dec < +10 deg.
   • -0.1 < B-V < 0.44 (B8 to F8 spectral types)
   • Flux (11.9mm) > 0.2 Jy

The complete list can be found here

Back to Content

12. Templates description

Detailled Template and Parameter Description

Please refer to the TIMMI2 status page for updates on offered modes before your run or contact the 3p6 team.  ALL chopping amplitude have been set by default to 10 arcsec. We defined the "smallsource" template as the standard imaging template: this is due to the fact that at very high chopping frequency, there is a drop in the sensitivity. In addition, because of the distorted guide star, the image produced is double. Hence "classic" should be used only with great care.

TIMMI2_img_acq

13. Acronyms

ASM Astronomical Site Monitor

BB Broad Band

BOB Broker for Observation Blocks

CCD Charge Coupled Device

DCR Double-Correlated Read

DIT Detector Integration Time

ETC Exposure Time Calculator

GUI Graphical User Interface

ISAAC Infrared Spectrometer and Array Camera

NB Narrow Band

NDIT Number of Detector Integration Time

NINT Number of integrations

NTT New Technology Telescope

OB Observation Block

OS Observation Software

P2PP Phase 2 Proposal Preparation

RTD Real Time Display

SOFI Son Of Isaac

Back to Content

14. About this document

This document was created under Dreamweaver 4.01 under windows XP

Back to Content