Comparison of Results

As mentioned previously, the results from the MSSP were to be compared with actual results from the completed Monash spectrograph. However, the spectrograph is still to be completed so it was impossible to do such comparisons. However, the theoretical data from the MSSP have been compared with other data, both theoretical and actual, from other spectrographs.

Throughputs and Count Rates

A comparison of the theoretical throughput of the Monash spectrograph was made with the theoretical throughput of the Faint Object Red Spectrograph (FORS) on the Anglo-Australian Telescope. Also a comparison was made between the theoretical count-rate of the Monash spectrograph with that of the RGO spectrograph assuming both were situated on the Monash telescope.

As the components of the FORS and the Monash spectrograph have similar overall transmittance, one expects the spectrographs' theoretical peak system efficiencies also to be similar. This was found to be so: the theoretical peak system efficiency of the FORS is 11.4 0.2% (Couch 1989) and the theoretical peak system efficiency of the Monash spectrograph, calculated by the MSSP, is 11.3 0.1%.

We compared the calculated theoretical count-rate for a star of a paticular visual magnitude using the RGO spectrograph, assuming that it is situated on the Monash telescope, with the count-rate predicted by the MSSP for the Monash spectrograph. These count-rates should be similar, because the spectrographs' components are similar. For a star of V = 8 and based on the count-rates given in Sadler, Harrison & Lee (1991) a theoretical count-rate of 42 counts per ö… per second at 5000 ö… for the RGO spectrograph on the Monash telescope was calculated. For a star of this visual magnitude the MSSP predicts 38 counts per ö… per second at 5000 ö… for the Monash spectrograph.

We believe that these comparisons indicate that the calculations by the MSSP are reasonable.

Radial Velocity Measurements

The main purpose of the Monash spectrograph is to determine stellar radial velocities. To determine the precision of expected radial velocity measurements, several simulations of target and comparison stars were made. A sample target star was of type K0V, with a visual magnitude m = 11.0 and a simulated projected rotational velocity () of 50 km . The standard comparison star simulated by the MSSP was a K2V type star with a visual magnitude of m = 4.0 and a simulated projected rotational velocity () of 50 km . Exposure times were 1000 and 100 seconds for the target and comparison star respectively.

To simulate a radial velocity measurement, the MSSP was used to simulate nine separate exposures of the target star and one exposure of the comparison star. Only one exposure of the comparison star was taken, as this star is relatively bright and the variations between exposures were shown to have little effect on the estimation of the radial velocity measurements. The nine exposures of the target star were cross-correlated with the exposure of the comparison star using the FIGARO spectral analysis code (Shortridge 1991). The standard deviation (1 value) of these nine measurements was used to estimate the uncertainty in the radial velocity measurements. For the standard comparison and target stars listed above, the standard deviation of the nine radial velocity measurements is 5.5 km .

As there are no actual data from the Monash spectrograph, we have compared the above results with data for three stars in the Monash observing program, AB Doradus (HD 36705), PZ Telescopium (HD 174429), and CF Octantis (HD 196818), observed by a former Monash Ph D student J. Innis (Innis 1986). The radial velocities of these stars were taken with the Boller and Chivens echelle spectrograph at the Cassegrainian focus of the 1.0-metre telescope at the Siding Spring Observatory, or with the coudö© echelle spectrograph on the 1.9-metre telescope at the Mt. Stromlo Observatory.

The three real stars are all late-type stars similar to the sample target star simulated by the MSSP. The 1 values for the real stars are similar to those for the simulated star. As the real stars were observed with SNRs of around 10:1 to 30:1, and the SNR for the simulated star is approximately 25:1, this would seem to lend some support to the accuracy of the simulated results given here even though the spectra from the three real stars were taken with spectrographs of higher spectral resolution than the Monash spectrograph.

á© Copyright Astronomical Society of Australia 1997