HOPS Data Products

HOPS data products are provided in three general formats:

1. FITS format data-cubes for eleven strips of the Galactic plane approximately 10o X 1o (large files ~500MB each).

2. A data cutout server providing smaller files at user-specified coordinates.

3. Catalogues of contiguous emission for each molecular transition.

Data Processing

Data are provided for each of the following spectral lines: water masers; NH3 (1,1), (2,2), (3,3), (6,6); HCCCN (3-2); and H69α. Full details on the data reduction method can be found in HOPS Paper II, but for convenience we outline the method here:

1. Standard livedata and gridzilla packages for Mopra data reduction were used to produce initial data cubes.

2. Duchamp was run on the data cube to find strong emission. Duchamp is a package designed to find real emission in data cubes.

3. The Duchamp output is used to create a mask of where strong emission has been found. If baseline ripples are identified in error as real emission they are manually edited out of the mask file.

4. The original data-cube baselined with a third order polynomial using the advanced routines found in ASAP. During the baselining process the mask produced in step 3 is used to remove any real emission.

5. The resulting baselined data cube is then smoothed to a beam size of 2.5' and spectrally smoothed using a hanning filter of width 5-channels, in order to produce a cube with the most easily detectable emission.

6. Duchamp is run on baselined and smoothed data in order to make a more accurate mask. This mask is applied to the original data cube to blank any pixels with real emission, producing a data cube containing only noise.

7. The line-free data is used to make a pixel-by-pixel noise map by measuring the MADFM statistic along the spectral dimension of the cube. This is quite robust to some emission in the cube and produces an excellent RMS noise map.

8. Each plane in the baselined cube (Step 4) is divided by the rms noise map (Step 7), in order to produce a signal-to-noise cube. This results in a data cube that has the same noise characteristics throughout and can be used to automatically find emission in the entire cube in an unbiased manner.

9. Duchamp is run for a final pass, including the a'trous reconstruction method, on the signal-to-noise map, to identify clumps and their properties.

Last modified 6th June 2017.