We selected all the lines present in the FTS200 spectra with a signal-to-noise-ratio larger than four. We disgarded blended lines, lines from the image side band and “ghost” lines in the E230 setup arising from mixing with higher order harmonics of the local oscillator frequency. For each selected line, the intensity was integrated over a fixed interval of 29 km/s, equal to the typical linewidth of the source, after subtraction of a baseline of first order in a region of three times the linewidth. The relative difference of integrated line intensities between MIRA and MRTCAL-calibrated spectra (i.e., the difference of integrated intensities divided by the mean integrated intensity) as a function of frequency is shown in Fig. 7.
The mean of the relative difference over all lines and scans is consistent
with zero difference for all frequency setups (0.7
0.7% for setup
E150/POINT2MM, 1
18% for setup E150/L170250, 0.0
2.2% for setup
E230, and
1.2
1.8% for setup E330). This indicates an on-average
good agreement of line intensities between MIRA- and
MRTCAL-calibrated spectra. For the E230 setup the maximum relative
difference for individual lines is 12%, for the E330 and E150/POINT2MM
setup it does not exceed 5%, compatible with the typical expected
calibration accuracy1. Significant deviations are
noted at the high frequency end of the E150/L170250 setup, close to the
edge of the 2 mm band atmospheric window, with integrated line intensities
in MRTCAL-calibrated spectra up to a factor of two higher than those
from MIRA-calibrated spectra. This particular case will be considered in
detail in the following section.