Beyond The Peak

Data Details

The data consist of spectra from 194-671 μm (1550 - 450 GHz) taken with the SPIRE Fourier Transform Spectrometer (FTS) onboard the Herschel Space Telescope. FTS consists of two detector arrays, the SLW at long wavelengths (303-671 μm) and the SSW at short wavelengths (194-313 μm). The arrays consist of 19 SLW (37 SSW) hexagonally-arranged detectors. We make use of the 7 (19) unvignetted portion of the field-of-view, as the vingetted detectors provide significantly degraded signal. The size of the detectors on the sky is a function of wavelength, ranging from 17-21" in SSW (fairly proportional to wavelength) and from 29-42" in SLW, with a complicated wavelength dependence due to the multi-moded feedhorns.

For the Beyond the Peak Project, we mapped at intermediate sampling, which moves the detector array around so that the detectors sample at 1/2 the Nyquist frequency. For our galaxies, we have used the intermediate sampling patter with 4-jiggles (dither). This is projected onto a 6 × 6 grid, resulting in a mapped area of 2 × 2 arcmin for each galaxy center or extra-nuclear region.

Raw data to spectra

The raw data from FTS are timelines from each detector as the optical path difference between the two split radiation beams is scanned over a range determined by the spectral resolution required. These observations were taken in high-spectral resolution mode (approx. 300, 500, 1000 km s-1 FWHM at 200, 300, 650μm). A forward and backwards scan is always taken; the BtP data consist of 55, 32 and 16 such pairs per jiggle position depending on .

Using HIPE v12 with SPIRE-FTS calibration v12.2, cosmic ray removal and a non-linearity correction from signal voltage to source power are performed. Then the interferograms are made and converted to spectra by the inverse Fourier transform.

Flux calibration and spatial sampling

Line fitting and fluxes

Emission lines in the SLW and SSW data were fit separately in HIPE using a variable width sinc profile. The underlying continuum is fit simultaneously with a 5th order polynomial. In each Galaxy the line-list-library is redshifted according to the galaxies characteristic velocity. Line fluxes are listed in the data tables, along with the line velocity, with respect to systemic and errors on both of these quantities.

Errors for the line fluxes were determined by measuring the root-mean-square error in the surrounding continuum for each line, then multiplying by the square root of three, as there are three spectral pixels per unresolved line. The errors on the velocities are taken from the uncertainty in fitted central position of the line.