WRITING A UKIRT TECHNICAL CASE

The technical case is an important part of your application. UKIRT staff use it to check he feasibility of your observations, to verify that the time request is adequate and to ensure that we can plan our instrumentation schedule to meet your requirements.

Give as complete a target list as you can - put this in the technical case if there is not room in the body of the form. Note that sources above +60 Declination and below -41 declination cannot be accessed by UKIRT. Of course you should also check whether your source RAs are appropriate for the time of year. Software packages are available to help you schedule your own observations, such as UKIRT Rise and OBSERVE for STARLINK users.

UKIRT rise is available at :

uktrise.html

You must include in your case (estimates of) source continuum magnitudes or fluxes, or line fluxes if appropriate. This information is essential for the technical assessor of your programme to check your quoted exposure times. Please quote these values at the intended observing wavelength; a V magnitude is of little use to a technical assessor for a UKIRT proposal. If you are proposing to observe a large sample of sources, make sure that you give the mean and range of values. Please adopt the following guidelines when quoting the brightness of your source.

Extended continuum source	Flux density in W/m2/µm/arcsec or (m)Jy/arcsec2
Point continuum source	Flux density in W/m2/µm, (m)Jy or give a JHKLL'M magnitude.
Extended line source	Flux in W/m2/arcsec2

Extended continuum source	flux density in W/m2/µm/arcsec2 or (m)Jy/arcsec2
Point continuum source	flux density in W/m2/µm, (m)Jy or give a JHKLL'M magnitude.
Extended line source	flux in W/m2/arcsec2
Point source emission line	flux in W/m2

Finally, make sure that you use consistent units across your sample; it has been known in the past for the brightness of four different objects to be quoted in four different units, which makes the task of the technical assessor far more onerous than it needs to be.

You should state your signal-to-noise requirements very clearly (even if all you say is that you require 2-sigma detections). Of course, a science case which fails to explain why these are the requirements is not complete.

The various instrument pages give details of the exposure times required to reach the background limit, if appropriate. In this limit your signal to noise will increase as the square root of the exposure time and in proportion to the source flux. When the observation is not background limited the S/N increases linearly with on-chip integration time. This is of relevance for imaging observations in the I and Z and, especially, for spectroscopy with UIST and with the echelle grating (CGS4) in the J, H and K windows when working between the OH lines (echelle background limit reached in around 600seconds).

You should estimate exposure times; do not leave it to the technical assessor to show that your programme is technically feasible (or not). It is valid to base your requested time on previous experience; however if you do this, it is your responsibility to demonstrate that the objects being observed previously were similar to those now being proposed, and to describe the weather conditions on the previous run.

Finally, if your sources are bright, you should of course check the saturation limits for your instrument.

Allow a proportion of overhead in estimating the total time required for your observations. This includes peak-up time when using the CGS4 echelle spectrometer (this is much reduced with UIST because of the availability of the internal imaging mode), and standard star observations for all programmes. Consult the web pages for your instrument for special cases (for example, CGS4 thermal infrared observations, which are dominated by overheads relating to readout and data transfer time). The following lists typical values.

• Allow 1 minute per object for slews/guide star aquisition.
• UFTI read overhead is currently 13 sec/frame for full array. This adds up if you are asking for 2hrs of on source integration at 60s/frame.
• Allow for darks with UFTI
• FLATS and arcs for CGS4 take about 3 minutes per wavelength change
• CGS4 peak-ups take about 2 min per object (typically; can be much longer if the source is faint).
• Standards (5-10 minutes for a JHK set with UFTI)
• UIST imaging acquisition mode is appropriate for targets fainter than 10th (K). In this case, acquisition for UIST spectroscopy takes of order 1 minute.
  

UKIRT frequently delivers half arcsecond seeing. Sub-half arcsecond seeing is not uncommon, but requires flexible scheduling to take advantage of. For this reason UKIRT is a flexibly-scheduled telescope from semester 03A onwards. You should specify your programme's seeing requirements in the Technical Case, even if this is simply to state that your programme has no particular seeing criteria. Other weather criteria include three-micron transparency (function of water vapour column) and photometric conditions. If you need reasonable transparency but can do without photometric conditions, you should state this clearly and explain how your science will be affected.

A description of the statistical behaviour of CSO tau over the last few years, and a compaison between photometric time and water vapour, is available at:

MK Tau (water vapour) statistics

The echelle grating covers a very small wavelength range in one setting. To ensure that the assessor can check your requirements, be sure to include the resolution and wavelength coverage demanded by your science, and to determine whether more than one grating position is required.

Check filter availability tables, and include filter information in your case.

Check filter availability tables, and include filter information in your case. For FP work check that appropriate narrow-band filters are available.

Notes on Michelle performance and overheads will be posted here before the next period of Michelle availability in Semester 04A.

UIST overheads:

For thermal imaging overheads in array readout can be significant, and depend on camera and readout area, see

sensitivity.html

Other imaging read overheads are given at

readspeed.html

There are additional observational overheads of target acquisition, moving filter wheels and telescope settle time on a jitter - allow an additional ~20% for these overheads.

Spectroscopy with UIST is more efficient than with CGS4 as there is no need to step a grating position to obtain fully sampled spectra, and acquiring faint targets can be done by imaging acquisition as opposed to peaking up through the slit.  A typical flat, arc, standard,target sequence can be done with ~80% efficiency.

When using IRPOL with any of the other instruments (Michelle has its own, built-in waveplate unit), always give the expected polarisation of the source and the polarisation accuracy needed (see the IRPOL web pages for details).