Area
The VMC survey covers a total area of 170 deg sq. distributed as follows: 105 deg sq. in the LMC, 42 deg sq. in the SMC, 21 deg sq. in the Bridge and 3 deg sq. in the Stream (Fig. 1). These regions are covered by 68, 13, 27 and 2 VISTA tiles, respectively. The list of tile centres and their position angles is available here. Tiles are 1.5x1.18 deg sq. in size and they overlap by 0.1 deg in the vertical direction and by 0.016 deg in the horizontal direction. The latter is the same as between adjacent pawprints re-constructing a tile. Detector #16 has a 200 bad pixel area but the lack of sources due to this region is much smaller than the lack of sources in the line of sight of bright Galactic stars. Associated to each tile there is a confidence map carrying the information on the affected pixels.
|
|
Fig. 1 Distribution of VISTA tiles across the Magellanic System
(LMC, Bridge, SMC and Stream). Underlying small dots indicate the
distribution of carbon stars (black), stellar clusters (blue) and
associations (red) while thick dots show the location of
observations to be performed with the VLT Survey Telescope (VST) in the optical
domain. |
The VMC survey aims to obtain data at three near-infrared filters: Y, J and Ks detecting sources with a S/N=10 at Vega magnitudes of Y=21.1, J=21.3 and Ks=20.7, respectively. Upper magnitude limits, corresponding to Y=12.9, J=12.7 and Ks=11.4, have been estimated assuming linearity until about 105 electrons which corresponds to a DIT time factor to peak saturation of about 1.5.
Strategy
A single VMC observing block (OB) observes one tile at one filter. The time necessary to reach the nominal survey depth is accumulated by repetitively observing a tile 3 times in Y and J and 12 times in the Ks.
These multi-epoch observations are obtained as follows: - one epoch per wave band as close as possible in time with pairs at half exposure during the same night, - the remaining epochs during the same year or as close as possible after that with observation preferably following a time interval of +1, +3, +5 and +17 days for all remaining epochs at Ks but without such a schedule in the Y and J bands. The first point minimises variability effects on colours, which are more important for bright objects, and at the same time guarantees more or less homogeneous observing conditions among different bands for a given epoch. The second point is important to derive average Ks magnitudes for RR Lyrae and Cepheid stars with sufficient accuracy (a few hundredths of mag) to unveil the 3D structure features of the Magellanic system.
Sky
The average requested seeing of the VMC data is 0.9 arcsec. This guarantees homogeneity of the data sensitivity and a good point-source extraction. The most crowded regions (e.g. the LMC bar), which already approach saturation with this condition, are observed with a seeing of 0.8 arcsec while the outer (un-crowded) regions are observed with a seeing of 1.0 arcsec.
The moon brightness is never a problem because it is always at least 60 deg away from the Magellanic system.
Observations with thin cirrus add an extra extinction of about 0.08 mag in Ks (much lower in J and Y) which decreases by 10% the S/N of the data. A similar extinction is caused by observing down to airmass=2. The airmass limit for VMC observations is 1.7.
Exposures
The estimated exposure time has been calculated using the VISTA ETC assuming a blackbody (T=5000 K) flux distribution, an aperture diameter of 1.6 arcsec and 0.8 arcsec seeing at airmass 1.5 in Ks-band (0.9 arcsec in J and 1.0 arcsec in Y) which corresponds to a DIMM seeing of also 0.8 arcsec (this occurs about 45% of the time at Paranal). The on-source time necessary to satisfy the sensitivity criteria is estimated to be: 50m, 40m and 3h per Y, J and Ks band. These integration times also reach 1 mag fainter sources at S/N=4 which is well below the main sequence turn-off of the oldest stellar population in the LMC.
The total survey time results in 1889h (1056h for exposure time and 833h of overheads as detailed below).
The observation of a full tile in a given filter are obtained by mosaicing 6 pawprints to cover the empty space among the 16 VISTA detectors. Jittering along a pattern of 5 is adopted to improve image quality (a jitter move occurs after 80-100s depending on filter, which is well above the minimum of 30s for the LOWFS to work in basic mode).
By observing adjacent tiles in the same filter the tile-change overheads are reduced from a few tens of seconds to at most 10s per tile. A slew/preset from the LMC centre to the SMC centre requires ~25s. However, if VMC observation follow those at a distant sky position the presetting overheads can be of the order of 2m or larger. Assuming that 75% of the time presetting overheads will be small (15s) and 25% large (120s) we will spend on average ~41s for presetting.
When VMC observations are interleaved with those from other surveys at other filters, it will take 25s to change between adjacent filters and 60s to change between the most distant filters. The same criterion as above results in an average ~34s for filter change.
Summarising, the gross time to observe a single tile in three filters (3 consecutive OBs) is 2h42m, including 50s to move twice to adjacent filters, 41s for presetting and 1s for acquisition, while individual OBs require on average a gross time of 56m which includes 34s for filter change, 41s for presetting and 1s for acquisition.