ING Banner
Home > Public Information > ING Newsletter > No. 6, October 2002 > Science

ING Newsletter No. 6, October 2002

Previous: Quasar Redshifts from S-CAM Observations: Direct Colour Determination of ~12 Gyr-Old Photons | Up: Table of Contents | Next: The ING Telescopes in a Changing Landscape

Other available formats: PDF


R. G. Sharp, R. G. McMahon, S. Hodgkin and C. D. Mackay (Institute of Astronomy, University of Cambridge)

The Isaac Newton Telescope has been used in conjunction with the Cambridge InfraRed Survey Instrument (CIRSI), to undertake a wide area deep IR survey in the J and H bands. This article gives a brief introduction to the survey and presents some initial results. In the spirit of the INT Wide Field Camera Survey program (Walton et al., 2001;; we are making reduced data products publicly available. A preliminary data release is planned for April 2002 with a complete release planned in the Summer 2002. The survey observations have been used in conjunction with optical CCD data from the INT Wide Angle Survey to undertake a survey for low and intermediate redshift quasars (z<3) free from the potential biasing effect of dust absorption. The results of these observations are reported to illustrate the utility of the survey data for combined optical-IR survey projects.


With a field of 4×7.80'×7.80' at the prime focus of the 2.5m Isaac Newton Telescope, the Cambridge InfraRed Survey Instrument, CIRSI (Mackay et al., 2000), is currently the largest field of view IR imager in operation. The camera, a mosaic of 4 Rockwell HgCdTe HAWAII IR arrays, is capable of observing in the J and H bands at the INT. The physical construction of the detector arrays prevents them being butted together in close proximity as is normal for optical CCD mosaic cameras. There is a 90% spacing between the elements of the mosaic. Sequential observations are offset to fill the gaps in the mosaic. A 4 pointing tile of observations covers an area of 29.6'×29.6'. Figure 2 demonstrates the camera layout.

Figure 1cirsi1.jpg
Figure 2
Figure 1. CIRSI at the prime focus of the 2.5m INT. [ JPEG | TIFF ]
Figure 2. The image to the left is a 29.6'×29.6' mosaic of four pointings of the CIRSI camera. To the right the weight map associated with the mosaic is shown. [ JPEG | TIFF ]

The nominal survey depths attained are J<20.0, H<19.0 roughly three magnitudes deeper than the level attained by the 2MASS project. Details of the fields observed are given in Table 1. The survey observing strategy consists of ~350sec exposures. Depending on sky brightness conditions during observations the exposure is built up from a sequence of 4 (or 5) dither positions with 4 (or 3) exposures of 22s (or 30s) duration at each position. Approximately 5Gb of data are obtained a night. Data processing is performed using pipeline processing software developed in Cambridge (Sabbey et al., 2001).

RA Dec (J2000)   

16:10 +54:30
J<20.0, H<19.0
16:37 +41:16
J<20.0, H<19.0   
14:52–14:58 +00:00
22:00–22:16 +00:00

Table 1. Survey regions and observational data available. The North and South Galactic Cap regions (NGC/SGC) are coincident with the recently released SDSS zero declination strip observations.


The fields targeted by the CIRSI-INT survey program are chosen to be coincident with observations from the INT Wide Angle Survey project (McMahon et al., 2001). The INT WAS fields surrounding the ISO ELAIS N1 and N2 fields at 1610+54 and 1637+41 have been extensively observed in the J and H bands and observations have also been undertaken in the J band of the zero declination strips at RAs of 14:55 and 22:08, coincident with the recent data release from the Sloan Digital Sky Survey (SDSS).

Pixel scale (arcsec/pixel) 0.46
FOV per chip 7.8' × 7.8'
FOV for 2×2 mosaic 29.6' × 29.6'
Frames/deg2 16
J, 5min 5σ in 1.2'' seeing 20.0
H, 5min 5σ in 1.2'' seeing 19.0
Table 2. CIRSI Characteristics on 2.5-m Isaac Newton Telescope.

Example Science: Reddening Independent Quasar Selection

It has long been known that absorption by dust, if present either along the line of sight to quasars or within the quasar host galaxies themselves, could bias samples of quasars selected primarily on the bases of ultraviolet excess (UVX). Figure 3 demonstrates the predicted effects of dust reddening on UVX selection of quasars. Two of the quasars identified in the CIRSI-INT sample for which U band observations are currently available do not show a UV excess. There are a number of reasons to endeavor to construct a quasar sample free from a bias against dusty quasars. Two examples are the following:
Figure 3
Figure 3. A comparison is shown of a sample of previously known quasars identified in the SDSS. The new quasars identified by gzH selection for which ugr data is available from the INT WAS are marked with filled circles. The horizontal line at u–g= –0.3 represents a UVX selection boundary analogous to that used in the 2dF quasar survey (2QZ Boyle et al., 2000). The stellar locus, computed from a spectral atlas, is shown along with stellar objects from a field of the INT WAS. The locus of quasar colour as a function of redshift is indicated by the solid line. The onset of absorption in the g band, due to the Lyman-a forest, is evident for quasars with z>3. Reddening vectors for dust models based on the Galaxy and the Small Magellanic Cloud are shown over a range of redshifts. The 2175Å feature in the galactic law passes through the g band over the range 1<z<2. [ JPEG | TIFF ]

Figure 4 illustrates the principle of the gzH colour selection technique developed for defining quasar candidates. The principle is similar to the VJK selection method proposed by Warren, Hewett & Foltz (2000), however, observations are required in only one IR band. A preliminary quasar identification has been undertaken by Sharp et al. (2002). 68 candidate quasar are identified in data taken from a subset of 0.7deg2 of the available survey area. Spectroscopic observations of 32 targets have been obtained confirming 22 quasars, a success rate of 65%. Observations are currently available across the ugrizJ and H bands for a sub set of the quasars identified. The ugr colour diagram, analogous to the traditional UVX quasar colour selection scheme, is shown in Figure 3. Two of the newly discovered quasars show no UV excess at all.

Figure 4
Figure 4. The candidate selection diagram for a 0.17deg2 region of the survey is shown with the full quasar sample overlaid. The selection boundary, chosen based on model quasar colours, is indicated by the broken line offset from the linear fit to the stellar locus. [ JPEG | TIFF ]


Email contact: Robert Sharp (

Top | Back

Contact:  (Public Relations Officer)
Last modified: 13 December 2010