added 19 Apr 1996


The Processing Path for Faint Object Spectrograph (FOS) Data

The Hubble Space Telescope has a number of astronomical instruments on board. All of these have some way of measuring the spectrum of an object, that is, the amount of light that comes from the object as a function of wavelength or color. The Wide Field/Planetary Cameras (WFPC and WFPC-2) that have produced the stunning images seen in papers aand magazines provide spectral information by using filters. As a result, the images show the amount of light produced over a small range of wavelengths for a number of objects. The Faint Object Spectrograph (FOS) built by the University of California, San Diego and the Goddard High Resolution Spectrograph (GHRS) provide the information as a spectrum or rainbow. Most of this rainbow is not apparent to our eyes however because it is formed from light in the ultra-violet, light which is absorbed by the Earth's atmosphere. It can be detected with ultra-violet sensitive detectors, such as those in the FOS and GHRS. These spectral images, usually of just one object at a time, are not as exciting to the average person as images but convey a great deal of information to the astronomer. From spectra, we can determine the composition of objects, their relative speeds, their rotation velocities, indications of their activity (or lack of), their temperatures and many other properties. As well, we may be able to derive some information about the material between us and the object which is observed.

Information obtained from the instruments on board the Hubble Space Telescope (HST) undergoes a number of changes before the final result is delivered to the scientist. Some of the steps in the FOS reduction procedure are outlined below.

The Exposure

The brightness of the target determines the number of counts registered by the elements in the detector during the exposure duration. The exposure builds just like a photographic exposure would (more time means more exposure), but unlike the photographic image on film, the FOS image can be accessed or read out a number of times before the exposure time is completed. The first readout is quite noisy (that is the variations from one detector element to another along the spectrum are quite large and most of the fine detail is missing). For this quasar target, the first readout represents the results after a relatively short exposure. The final readout for this target has a much lower noise level. A comparison of the two spectra shows that much more detail is present in the longer exposure.

Exposure Corrections

For various reasons, the components which measure the light (diodes in this case) in the FOS detector may become unreliable. These diodes are disabled by commands from the ground. The lack of signal from the disabled diodes produces characteristic depressions in the unprocessed readouts. (We have a way of moving the target's spectrum across the detector so that we don't have gaps in our final spectrum.) The first correction adjusts for these disabled diodes (in the figure, the regions affected by disabled diodes have hat-shaped symbols above them). The readout has been divided by the exposure time to convert it to a count rate.

Spectra taken with the same instrument settings can now be compared relative to one another.

Non-Target Contamination

An exposure records light from both the target and the region of the sky around it. Diffuse galactic light, zodiacal light (from our own solar system) and light from the Earth's outer atmosphere all contribute to the sky background light. Except for a couple of emission lines from the Earth's outer atmosphere, this effect is negligible in the wavelength region used for most FOS observations (the ultra-violet between 1100 Angstroms and 3200 Angstroms). Since the emission lines occur around 1250 Angstroms, they do not appear on this plot.

High energy particles, called cosmic rays, trapped in the Earth's magnetic field generate light, called Cerenkov radiation, in the detector itself. This particle background is about the same across the detector and usually quite small relative to the signal from the object. This background level varies, however, as the HST moves in its orbit, being higher as the telescope gets closer to the Earth's magnetic poles.

During this step, the contributions from the sky and the high energy particles are removed.

The remaining signal should depend only on the target.

Instrument Effects

Not all paths through the detector have the same response to light of a given wavelength (color). At this point, corrections are made for the differences in response. The instrument is not equally sensitive to light at all wavelengths. Here, the sensitivity is taken into account. A wavelength scale is added to aid interpretation.

The spectrum is now in a form that can be compared to or combined with spectra from other instruments or epochs. This is the final result.

Ron Lyons CASS, UCSD 0424, 9500 Gilman Drive, La Jolla CA 92093-0424

Back to the CASS Home Page