(exponentiation), as well as a
special magnitude operator ``&'' for adding numbers in quadrature.
The noise expression may also involve the functions sqrt() and log().
Next: SPEFO---A Simple, Yet Powerful Program for One-Dimensional Spectra Processing
Previous: A Reduction and Analysis Pipeline for ROSAT PSPC Data
Table of Contents --- Search ---
PS reprint
Bernie Simon,1 Richard. A. Shaw
Space Telescope Science Institute, 3700 San Martin Drive,
Baltimore, MD 21218
1Computer Sciences Corporation
Synphot is a synthetic photometry package in STSDAS for modelling and trending the photometric response of the scientific instruments on board HST, as well as for general analysis of photometric and spectrophotometric data. Bushouse & Simon (1994) point out that the data-driven nature of synphot makes it possible to model the response of similar detectors for any observatory, given the appropriate throughput data.
A new sub-package, called simulators, has recently been added to synphot to model the two-dimensional response of both imaging cameras and long-slit or echelle spectrographs. Simulators of this sort are valuable both to the scientists who build or calibrate the instruments on board HST, and to potential Guest Observers who must evaluate whether they can accomplish their proposed scientific goals with a given exposure time. The simulators were built to model the complex data that are anticipated from the next generation of HST science instruments, STIS and NICMOS, but they were designed to model data from the current cameras as well.
There are a number of features that are common to the simulator tasks. They take as input the observing mode (i.e., the combination of instrument, detector, and the grating or filter) and a table specifying the attributes of a list of objects to be simulated. One may also optionally model and add the celestial and/or instrument background, and include or substitute a user-specified background image. The simulators can optionally introduce noise that is appropriate for the specified instrument/detector combination. The calculated noise will be affected by the number of readouts of (or sub-exposures with) the detector, which can be specified through the nread task parameter. The brightness is adjusted by the exposure time parameter exptime, and the output pixel values may be rounded to simulate an A-to-D quantization effect. Warnings will be issued if the brightness in any pixel, or the combined brightness in all pixels, exceeds the local or global limit for the detector. The output image format for these tasks can be any native IRAF image format. Note that the calculation itself can be performed on a finer grid than the detector pixels by setting the parameter nsub to an integer greater than one.
The simimg task computes a simulated image for HST instruments with two-dimensional imaging modes: FOC, WFPC, WFPC2, STIS, and NICMOS. This task will compute the distribution of light on the detector, convolve it with the appropriate point-spread function (PSF), and normalize it to the computed throughput for each object. In practice, the task uses the input PSF image to construct the point sources, and only performs the convolution on the extended targets before incorporating them into the image. The simspec task computes a simulated spectral image for STIS long-slit and echelle modes, given the observing configuration and the object description table. The task parameters cenwave and cenorder specify the wavelength and order that fall on the image center, and the dispersion relation is adjusted accordingly. For each spectral order, the object shapes are convolved with the PSF, masked by the spatial extent of the slit, and (optionally) convolved in the dispersion direction by an appropriate line-spread function; each order is then mapped onto the detector pixels.
Two ancillary tasks have been added as well. The simbackgd task permits the construction of a customized background image, which may be used instead of or in addition to the background computed by simimg or simspec. The calculated background has contributions due to zodiacal light, scattered earthlight, and thermal background. Zodiacal light is a function of the relative position of the sun and the telescope pointing. The telescope pointing is set by task parameters det_ra and det_dec, and the sun position is set by task parameter time which specifies the date of the observation. The earthlight background is calculated from task parameter earthtab, which specifies the maximum earthlight spectrum, and task parameter eshine, which specifies a fraction of the maximum earthlight. The thermal background is calculated from thermtab, which specifies the spectrum of the thermal background.
The simnoise task permits the incorporation of customized noise into
an image, optionally following a user-specified prescription. The noise
may be applied to the output image from simimg/simspec, or used to
generate an input noise image that can be used instead of or in addition to
the noise computed by these tasks. If the user does not specify a prescription
for the instrument noise, it will instead be calculated from a NOISE parameter
stored in the throughput table headers. The NOISE parameter contains a
FORTRAN-like expression which is used to compute the mean of a Poisson
random process. The distribution is sampled and the random noise minus its
mean is added to each pixel. If more than one throughput table
contains a NOISE parameter, the results will be added in quadrature.
The noise expression can contain constants and the three variables t,
n, and x, which represent the exposure time, the number of reads,
and the pixel flux, respectively. The operators include the standard
arithmetic operators: (exponentiation), as well as a
special magnitude operator ``&'' for adding numbers in quadrature.
The noise expression may also involve the functions sqrt() and log().
The input table describes the objects to be viewed, and contains one row for each object. There are five fields for each object, specifying the right ascension, declination, magnitude, spectrum, and source shape. The RA and Dec determine the position of the center of the object on the image: objects are mapped into the coordinate frame of the detector, which is specified by the three task parameters det_ra, det_dec, and det_ang (measured from north through east in degrees). Any object whose center does not lie within the detector field of view is excluded from the output. The magnitude is used to scale the integrated flux of the object, the units for which are specified through the task parameter magform. The spectrum is specified as an expression, the syntax for which described in the Synphot User's Guide (Bushouse 1995). The spectrum is used to compute the object flux as a function of wavelength. The flux is renormalized to the object magnitude over the magnitude passband. The object shape specifies the functional form and extent of non-point source objects (see below).
The input table can either be an STSDAS binary table or an ASCII text table. If it is a text table certain restrictions apply to the column order, and units of hours and degrees are assumed for the object RA and Dec, respectively. The spectrum and shape fields are optional in the input table. If the spectrum field for any object is omitted, the spectrum specified by the task parameter spectrum is used in its place. If the shape field is omitted the object is assumed to be a point source.
The shape specification is very much like a function call---i.e., the type of shape is followed by a parenthesized list of function arguments. Table 1 provides a list of the available shape functions. Most shapes take three arguments, the first of which is the radius of a circle (or the semi-major axis of an ellipse), in arcseconds, which contains half the flux of the object. The second argument is the axial ratio, or the ratio of the semi-major to the semi-minor axis of an ellipse. The third argument is the position angle: the angle in degrees between the positive (detector) x-axis and the semi-major axis. Note that the Image Template function provides a means to scale and include any image in the simulation, provided that image has world coordinate information in its header. This can be extremely useful for modelling complex backgrounds, such as H II regions, etc.
The objects are convolved with an appropriate PSF, which is drawn from a user-specified catalog of PSFs. The catalog may consist of a single PSF that is calculated to match the detector/bandpass in question or it may be drawn from multiple PSFs, each of which being appropriate over some narrow range in wavelength. In the latter case, the tasks will compute an average of the PSFs, weighted by the effective bandpass. The PSFs may be oversampled---i.e., computed on a finer spatial grid than the detector pixels. Pixels in each PSF image are assumed to be square, and the size of the pixels and the location of the center of the PSF are calculated from the world coordinate information and other keywords in the image header.
Other ancillary files are accessed as well, although they should rarely need to be reset by the user. The task parameter zodtab specifies a file containing a tabulation of the zodiacal light flux, which is a function of ecliptic latitude and heliocentric longitude taken from Levasseur-Regourd & Dumont (1980). The task parameter earthtab specifies the earthlight spectrum at its maximum value, and thermtab specifies the thermal background spectrum. Note that setting any of these file names to a null string will have the effect of omitting that contribution from the background calculation.
Work on the simulators software is essentially complete, and is undergoing testing. The data files that describe the instrument characteristics, the PSF library, etc., are still being constructed. We anticipate releasing this package to the community in mid-1996 to support HST Guest Observer proposal preparation for the observing cycle following the 1997 servicing mission.
Bushouse, H., ed. 1995, Synphot User's Guide (Baltimore: Space Telescope Science Institute)
Levasseur-Regourd, A. C., & Dumont, R. 1980, A&A, 84, 277
