Spectra shown in the Formaldehyde Observation Summary below have been minimally processed with only baseline removal and 10-channel boxcar smoothing.
Spectra shown in the Results farther below have been minimally processed with only baseline removal and 20-channel boxcar smoothing.
Temperature scale is Ta.
Data Reduction Steps Used for 06B Measurements (Paper I):
Inspected all scans for bad baselines, correlator lag problems (rare), and band suckouts (even rarer). See comments in results table below.
In a few cases, where the results were surprising, I dug a bit deeper into the measurements and analyzed each individual 10 second sample for problems (none found).
Summed spectra for each source/IF in files exgalh2co-N.fits, where N = session number.
Summed all spectra for each H2CO frequency and resolution in file exgalh2co-sum.fits.
Spectra shown in the Formaldehyde Observation Summary below have been minimally processed with only baseline removal and 10-channel boxcar smoothing.
Spectra shown in the Results farther below have been minimally processed with only baseline removal and 20-channel boxcar smoothing.
Temperature scale is Ta.
Data Reduction Steps Used for Final All-Measurements Analysis (Paper II):
Inspected all scans for bad baselines, correlator lag problems (rare), and band suckouts (even rarer). See comments in results table below.
Summed all spectra for each H2CO and H111alpha frequency and resolution in file exgalh2co-sum-all.fits.
Analysis branched at this point to accommodate need for varying spectral resolution as a function of source:
Path A:
Averaged all "like" spectra. This effectively means that I average all 4.8 GHz spectra by first smoothing to a common spectral resolution.
Smooth all averaged spectra to 10 km/s spectral resolution using gaussian smoothing (macro smoothit.pro).
Input File(s) = exgalh2co-sum-all.fits
Output File = exgalh2co-combsmo10-all.fits
Perused each spectrum and experimented with various baseline fits before determining which was appropriate. Used knowledge of other molecular emission (H2CO, NH3, CO, HCN, etc.) to guide baseline region setting.
Input File = exgalh2co-combsmo10-all.fits
Output File = exgalh2co-base10-all.fits
Path B:
Average all "like" spectra. This effectively means that I average all 4.8 GHz spectra by first smoothing to a common spectral resolution.
Smooth all averaged spectra to 1 km/s spectral resolution using gaussian smoothing (macro smoothit.pro). This output spectral resolution will allow for later smoothing on a source-by-source basis.
Input File(s) = exgalh2co-sum-all.fits
Output File = exgalh2co-combsmo1-all.fits
Used knowledge gained from Path A above as guide to analysis of each source, and specifically which ones are detections and which are not. From this point on only analyzed detected sources. Use 10 km/s smoothing analysis for nondetection results.
The following is a repeat of the Path A procedure above, but only for detected galaxies and using a variable, source-catered spectral resolution for the "final" results.
Perused each spectrum and experimented with various baseline fits before determining which was appropriate. Used knowledge of other molecular emission (H2CO, NH3, CO, HCN, etc.) to guide baseline region setting.
Input File = exgalh2co-combsmo1-all.fits or exgalh2co-base10-all.fits