Observations

Spatial distribution of the sample of galaxies. The red circles and blue triangles represent dwarf galaxies of the Milky Way and of Andromeda, respectively. Isolated dwarfs are shown by green squares and GCs are shown by brown lozenges. The violet hexagons are nearby giant galaxies and the gold pentagons represent other nearby dwarf galaxies that are studied in our group but not in INT project.

INT Observations:

Over a period of three years (June 2015 to February 2018), we used the Wide Field Camera (WFC) from 2.5-m Isaac Newton Telescope (INT) of the Observatorio del Roque de los Muchachos (La Palma) to survey the majority of dwarf galaxies in the LG. The observing time for this survey was primarily pro-vided by the Iranian National Observatory (INO), complemented by UK-PATT allocation of time to pro-grammes I/2016B/09 and I/2017B/04 (PI: J. van Loon). The WFC is an optical mosaic camera that it con-sists of four 2048 × 4096 CCDs, with a pixel size of 0.33 arcsec/pixel. Observations were taken in the WFC Sloan i and Harris V filters (except the first epoch in which the WFC RGO I filter was used). Some of the observational properties of our sample targets are summarized in Table A1.

The images were processed using the THELI (Transforming HEavenly Light into Image) code, an image processing pipeline for optical images taken by multi-chip (mosaic) CCD cameras. The THELI pipeline is optimized to perform precise astrometry and it is ideally suited to our goal. The Daophot/Allstar software (Stetson 1987) was used to obtain photometry for all stars in our crowded stellar fields by employing a Point Spread Function (PSF) fitting method.

The calibration process was performed in several steps. First, aperture corrections to the PSF-fitting photometry were made. Next, in order to perform photometric calibration, we determined zero points for each frame based on the standard star field observations. Finally, relative photometry between epochs was performed to correctly separate the variable from non-variable sources.

UKIRT Observations:

Observations were made with three of UK Infrared Telescope (UKIRT)’s imagers: UIST, UFTI and WFCAM. Observations with UIST were made in the K band (UKIRT filter K98) over the period from 2003 October to 2007 July. Each image has 1024 × 1024 pixels of 0.12 arcsec, and the combined mosaic covers approximately 4 × 4 arcmin2 – a square kiloparsec at the distance of M33. On three consecutive nights in 2005 August the UFTI imager was used instead, in the K band (UKIRT filter K98) only. The UFTI camera provides 1024 × 1024 pixels of 0.09 arcsec, and an area slightly larger than that covered with UIST was mapped. All UFTI observations were combined to create one mosaic. Observations were made in the K band (UKIRT filter K98) over the period 2005 September–2007 October with WFCAM. On some occasions, observations were made also in the J band and/or H band (UKIRT filters J98 and H98, respectively) to provide colour information. Some of the observational are summarized in Table 1.

The images were processed using the WFCAM pipeline by the Cambridge Astronomy Survey Unit (CASU; http://casu.ast.cam.ac.uk/). Photometry was obtained for all stars within each frame of UIST and UFTI by the automated fitting of a model of the point spread function (PSF), using the DAOPHOT/ALLSTAR software suite (Stetson 1987). Depending on the conditions, we used either a constant or a quadratically varying PSF. The individual images were aligned using the DAOMASTER routine, which computes the astrometric transformation equation co-efficient from the DAOPHOT/ALLSTAR results. We combined the individual images using the MONTAGE2 routine (Stetson 1994) and then coadded these three J, H and K-band images with the IMARITH task in IRAF to create a master mosaic of the central 4 × 4 arcmin2 of M33. The large pixel scale of WFCAM means that the point spread function (PSF) is not always adequately sampled for the crowded fields in M33 under modal observing conditions; to remedy this, we employed a 3×3 micro stepping scheme to improve the sampling of the PSF.

Photometric calibration was then performed in a three-step manner. First, aperture corrections to the PSF-fitting photometry were determined using the DAOGROW routine (Stetson 1990) to construct growth curves for each frame from which all stars had been subtracted except the PSF stars. In the second step, the standard star measurements were used to calibrate the frames on those nights. In the third step, we calibrated the frames relative to one another.