MAGMA Data Release 1:
CO Data for the LMC
Reference: Wong et al. (2011), ApJS, 197, 16

Data Cubes and Moment Maps

Cubes are provided at two different spatial resolutions. These are gzipped FITS images constructed using the SIN projection in the LSR velocity frame. Note that regions which were not observed have been assigned a blank value. The brightness tempeature scale corresponds to the main beam brightness temperature assuming a Gaussian beam.

CO intensity is determined in two ways: (1) Masking by the 3-sigma contour of a cube that has been smoothed to 90" resolution, then integrating from 200 to 305 km/s; and (2) Identifying the peak of each spectrum and, if at least 4-sigma in intensity, integrating all channels within 10 channels of the peak (window method). The first method was adopted for the ApJS paper. For each intensity image, a corresponding rms image is constructed by scaling the rms noise per channel by the square root of the number of channels that contribute to the intensity map at that position.

Following acceptance of the ApJS paper, we identified a low-level (~10 mK) positive offset in the spectral baseline which led to erroneous fluxes when integrating over large numbers of noise channels. We are still working to determine the cause of this offset, but for the first data release we have subtracted a constant offset of 10 mK from both cubes to bring the spectral baseline close to 0. Note that the rms noise of a spectrum is typically around 300 mK, so the offset is not important when integrating over small (3-D) regions of the cube, and thus does not affect the paper's conclusions.

CPROPS Analysis

These are the results of applying the CPROPS software (Rosolowsky & Leroy 2006) to the LMC_MAGMA_DR1.CO.45as.fits cube above, after truncation to a velocity range of 200-305 km/s VLSR and normalization to a constant rms noise of 0.28 K. Three types of decompositions were run:

• Islands: Regions of connected emission are identified starting from a "core" of 3-sigma significance in two consecutive channels and expanding out to an "edge" of 2-sigma significance that is connected to the core. The resulting regions are required to have a projected area of at least one telescope beam.
• Physical: The islands are decomposed into GMC-like structures using the /PHYSICAL keyword. This identifies local maxima within a moving box of size 5x5 pixels and velocity width of 9 channels. The local maxima must lie at least 1 K above the merge level with any other maximum, otherwise they are removed and their emission may be assigned to other maxima.
• Data-based: The islands are decomposed using the same criteria as for the "physical" catalogue to identify candidate maxima, but this time the maxima are only required to lie at least 2 sigma (0.56 K) above the merge level with any other maximum.

Note that for the physical and data-based methods, not all detected emission is assigned into clouds, and so the total CO flux in the assignment mask will be substantially less than for the islands.

Note also that CPROPS does not number clouds in order of R.A., so a mapping file is given below that gives the catalog number from the ApJS table in column 1 and the CPROPS assignment cube number in column 2.

UPDATE, 21 Jan 2013: The data tables in the ApJS paper incorrectly referred to the line width as the FWHM line width. These columns (Col. 10) actually correspond to the Gaussian sigma. The labels have been fixed in the tables below, and an erratum will be submitted to ApJS. (Thanks to Alberto Bolatto for pointing this out.)