|
SPICAOverviewAlthough it is possible to characterize imaging performance with a single sensitivity figure (e.g., 5 sigma in 1 hour) because virtually all observations with modern detectors on large telescope are background-limited in reasonable integration times, the same is not true for spectroscopy. Spectroscopic observations are similar to simultaneous imaging in many filters (one for each pixel in the spectrum) and the dominant source of noise may vary with wavelength, exposure time and/or spectral resolution. Exposure time calculators for spectrographs are therefore important to properly estimate the success of any proposed observations and, due to the large numbers of pixels, graphical output is preferred. The philosophy behind SPICA, the SPectroscopic Integration Calculator Applet, is to produce a fully interactive tool for simulating observations of relatively complex source spectra. Of necessity, SPICA is written in JavaTM and Swing, and requires at least Java 1.1.2 compatibility. Running SPICAThis page should have launched in a separate browser window, with SPICA running in the original window. If SPICA is not running, please refer to Sun's JavaTM Tutorial page on running Swing applets. You may need to speak to your local sysadmin to have the required plug-in installed on your system if you cannot run Sun's example Swing applet. As an alternative to running the applet in a browser, you may have appletviewer installed on your system, which will almost certainly work. How to use SPICASPICA has four separate display areas, labelled Model, Instrument, Messages, and Results. We will deal with these one at a time. ModelAt the top of the model display area are controls to set the seeing and spectroscopic extraction aperture (in arcseconds), together with the geometry of the source. The options are:
Changes to the source geometry do not take effect until the Observe button (in the Instrument window) is pressed. The spectral model can include up to 10 components. After constructing a component from one of the five types listed below, click one of the Add Before, Replace, or Add After buttons to include it in the model. To add further components, you must first highlight a component by clicking on it in the model display area, to indicate where in the list of components the new component will go (the order may be important). Updates to the model and simulated observation occur as soon as one of these buttons is clicked.
Redshifts are assumed to be recession velocities in km/s if >100, and fractions of the speed of light if smaller than this (but still positive). InstrumentThe buttons and menus in this area are specific to each instrument, and you should view the appropriate web pages for that instrument to understand their meanings. Changes made to the instrument settings will not take effect until you click the Observe button. MessagesThis area indicates any difficulties the ETC has in parsing input (e.g., non-numerical responses have been provided when numbers are required), as well as informing the user of the peak signal (in terms of the detector's full-well depth) and the fraction of the total source flux in the extraction aperture. Note that for a uniform source, this will be more than 100% if the extraction aperture is larger than 1 square arcsecond. ResultsThis area contains three buttons, an area for displaying graphical output, and two slider controls. The upper slider controls the central wavelength of the plot, while the lower one controls how much spectrum is shown. The ordinate is autoscaled. The Recentre spectrum button will reset these so that the entire spectrum is shown in the graph. The leftmost button controls which aspect of the simulated observation is shown. Initially, it is the model, but the simulated Data can be displayed, or the signal-to-noise ratio, or the Sky background. The central button indicates whether the ordinate is flux density in Snu or Slambda units. Comments or problemsPlease note that at the present time, SPICA is still a work in progress. In particular, some of the error-trapping may be incomplete. Suggestions for improvements, or problems encountered while running the applet, should be sent to Chris Simpson. Chris Simpson, 20 March 2001 |