1 Introduction

 1.1 An overview
 1.2 What’s new in this release
 1.3 A little more detail

1.1 An overview

ESP (Extended Surface Photometry) is a package of application programs developed to allow you to determine the photometric properties of galaxies and other extended objects. It has applications that:

ESP processes images stored as NDFs, and therefore you can use it in conjunction with other packages like; KAPPA, CCDPACK, Figaro, PHOTOM, JCMTDR and PISA.

It allows you to define areas to exclude in the analysis using keyword descriptions in text ARD files.

There is now an interface to ESP within GAIA. This supports all the ESP applications except GRAPHS and HISTPEAK: GRAPHS is purely concerned with the display of ESP output, and all the non-graphical functionality of HISTPEAK is provided by the support for HSUB.

The ESP web page is at http://www.astro.gla.ac.uk/users/norman/star/esp/.

1.2 What’s new in this release

ESP has been modified to be aware of the World Coordinate System (WCS) components of NDFs.

The COSYS parameter has been removed from all ESP applications, so that all user input and output is now in the Current co-ordinate system of the NDF. The Current co-ordinate system is a characteristic of the WCS component of the NDF. Thus, instead of setting COSYS to use either ’Data’ or ’World’ co-ordinate sytems, you should now set the Current frame of the NDF’s WCS component (e.g. using KAPPA’s WCSFRAME apllication) to the desired co-ordinate system before running the ESP application.

Output to data files is however in Base (WCS GRID frame) co-ordinates, which are pixel co-ordinates guaranteed to start at (1,1).

For NDFs which have a WCS component with a SKY frame in it, the PSIZE parameter (pixel size in arc seconds) is now determined automatically rather than being solicited from the user. This can be overridded by specifying it on the command line.

ESP applications which generate output NDFs from input NDFs now propagate the WCS component where appropriate.

1.3 A little more detail

ESP contains four major applications for its prime purpose: ELLFOU, ELLPRO, GAUFIT and SECTOR. These allow for the galaxy to be ellipse ‘fitted’ using contour analysis, ellipse ‘fitted’ using intensity analysis, profiled as a 2-D Gaussian distribution or pie slice cross-sectioned, respectively. In addition to these core applications, a number of other applications are available to make the task easier and more reliable.

In the real world, any image of a galaxy is likely to contain other objects either in the foreground or the background. This ‘contamination’ may lead to bumps in, or distortions to, the generated galaxy profiles and are best removed in some way. This may be easily done by using KAPPA’s ARDGEN to create a text file (ARD) that can be read by ELLPRO, ELLFOU and SECTOR to define the duff areas. Alternatively, you can use ARDGEN and ARDMASK in combination so that thereafter all Starlink packages view the pixels described as bad.

One of the most important factors in determining the profile of a galaxy is an accurate value for the background value of the image as a whole. If this is not available, then it will never be possible to accurately calculate the scale length or brightness of any object on the image. To allow you to obtain accurate values for a given image, the ESP application HISTPEAK determines the modal pixel value and, also, a range of other statistical data (mean, median, skewness etc.) from any two dimensional image. Since estimates of the modal value (background) and its associated standard deviation (often SIGMA) are the most important information required, it is determined by four different methods. The standard deviation of the pixel values and the background value standard deviation are also calculated (see §5 ).

The application SKEW allows the input image to be examined to highlight faults in the flatfielding. If the image is far from flat, with large variations in the background value, then the ESP application FASTMED may be employed to reduce the size of such fluctuations. This is a variable size median filter routine which reduces variations on a scale defined by you.

If you are interested in the background value for several galaxies on a given frame, then you may use the ESP application LOBACK which determines the modal pixel values at image locations specified in a simple text file. Such a text file might be easily generated using PISA, KAPPA’s CURSOR, RGASP’s IMAGES, or IRAF’s FOCAS. The background values determined are placed in an output text file and may be easily used in conjunction with the profiling applications ELLPRO and ELLFOU.

The applications ELLPRO, ELLFOU, GAUFIT and SECTOR may be used to examine the profile of galaxies/sources on an image. SECTOR is intended as an interactive application allowing the scale lengths of given objects to be very quickly estimated. ELLPRO and ELLFOU are more substantial applications. They either allow a single object on a frame to be identified interactively and subsequently profiled or, alternatively, examine a number of galaxies (image co-ordinates supplied by you in a text file) and generates profiles for each of them. The profile information generated includes position angle, ellipticity, brightness and Fourier descriptors (up to fourth order). GAUFIT by comparison, generates values for the location, brightness, orientation and widths of several galaxies/sources on an image in terms of 2-D Gaussian functions. It will prove especially useful for JCMT users (see also JCMTDR).

If ELLPRO, ELLFOU or SECTOR have been used to examine a galaxy image, the results, stored by the application in a text file, may be examined using the application GRAPHS. This displays the profile information generated in a variety of forms (i.e. intensity versus radius, position angle versus logarithmic radius etc.) The scale length of the galaxy profile can be determined simply by choosing the radius range of the data points to be considered.

As an aid to identifying faint diffuse galaxies in an image, the applications CORR and SELFCW may be used to correlate the image with a template galaxy of a given scale length. The significance of the detection can be estimated using the application MIXUP to generate a noise equivalent image.

The application TOPPED may be employed to remove bright pixels and their immediate neighbours from an image. This can reduce the influence of cosmic rays and also saturated regions on CCD frames. The neighbouring pixels are considered to be those within a defined radius of the bright pixel.

Most of you will be content to use the application HISTPEAK as it stands, however, the application HSUB is provided in which HISTPEAK is called as a subroutine. This is supplied so that the more adventurous of you may easily include the functionality of HISTPEAK into your own applications as required.