Observations with KMOS
As for any instrument in operation at VLT the astronomer has to prepare
observations with KMOS in advance. Due to the special nature of a
multi-object spectrometer in general and the specific KMOS features in
particular some additional constraints apply: To get the maximum benefit from
the granted observation time it is especially crucial to allocate as many
targets as possible to the 24 pick-off arms, thereby taking the target
priorities, restricted arm travel ranges, possible arm collisions and
shadowing effects into account. In addition, it shall be avoided that any arm
will be hit by field objects brighter than a certain limiting magnitude and
thus creating stray light.
Therefore a dedicated preparation tool KARMA (KMOS ARM
Allocation) is being developed, which guides the astronomer through the
different steps of observation preparation. As a plugin into the well-known
astronomy software package Skycat it provides all its catalogue and
image handling capabilities as well as a combined automatic/interactive method
of the target-arm assignment. The figure below shows a
screenshot of the KARMA/Skycat prototype currently under development.
Catching the galaxies: Screenshot of KARMA. All 24 pick-off arms are
allocated to faint red objects within the FORS Deep Field. Arms from the
upper layer are coloured green, those from the lower layer are coloured
blue. The square areas covered by the 24 IFUs can be seen magnified on
the right hand side.
One of the characteristics of any infrared instrument is that data
reduction has to cope with clever sky background subtraction. KMOS
provides two different methods of doing this, corresponding to two different
"Nod to Sky" Mode
In this mode the sky background signal is obtained by
moving ("nodding") the telescope and/or rotating the instrument
between two previously defined positions,
depicted by the two schematic
configurations in the left hand figure.
Therefore each pick-off arm (IFU represented schematically by a
black square) switches between its scientific target
(red bullet) and a corresponding own sky background position (blue
square). The latter one has to be found during preparation. In the
process of nodding only the telescope and/or instrument rotator positions
are altered, the pick-off arm configuration remains unchanged.
Since it is very unlikely that all 24 arms can be allocated at once
to scientific targets within a single telesope/instrument position it
will be possible to assign not used arms to remaining targets in the
second position where the already allocated arms are usually covering
blank sky. This leads to a configuration where in both
telescope/instrument positions (upper and lower figure on the left)
one subset of arms is allocated to scientific targets
and another subset is allocated to sky background.
The characteristic feature of this mode is that the telescope
always points ("stares") to the same position
as well as the instrument rotation
angle doesn't change (upper part of left hand figure).
The sky background then will be obtained by dedicated "sky
arms" which, in contrast to the "Nod to Sky Mode", deliver a signal
which is picked not by the same IFU as for the scientific target. All
pick-off arms allocated to science targets can, however, serve for sky
background subtraction simultaneously if their IFUs contain a sufficient
number of empty sky pixels.
A dedicated sky position where all arms are on blank sky in addition
is necessary only once (lower figure).
In addition, a special observation mode, known as "Mosaic Mode", is
dedicated to the spectroscopy of large contiguous areas. For this purpose a
predefined fixed configuration of pick-off arms (see figure below) will
be used covering an area of 64x43 arcseconds with 16 successive telescope
Not a board game:
The predefined pick-off arm configuration for the Mosaic Mode (upper arms
green, lower arms blue). By
shifting the telescope both horizontally and vertically within 16 steps a
contiguous rectangular field can be covered.
Last modified: 2017-03-08 (Michael Wegner)