Sgr B2 is the most massive gas clump in our Galaxy. It's about 100 pc away from Sgr A*, the black hole at the center of the Galaxy. It is forming stars prodigiously, and has extreme properties in many regards.
In many physical aspects - mass, luminosity, velocity dispersion, temperature - the Sgr B2 region is similar to "proto-cluster clumps" seen in other galaxies, e.g. the Antennae galaxies.
There are plenty of mysteries about Sgr B2, including its extraordinarily diverse chemistry and its peculiar kinematics.
I'm raising a sort of new one. There appears to be a hole next to Sgr B2, towards the southeast in Galactic coordinates. It appears as a relatively empty spot in the dust continuum bands. It doesn't seem to be filled in with anything in particular.
Note the slider here: you can view Sgr B2 at any wavelength where I have data. The green circle highlights the major deficiency that is evident in all dust-containing bands. There are other "minor" deficiencies elsewhere.
It is tricky to interpret this: I don't think the absolute flux densities in any of the millimeter observations are perfectly trustworthy because of their proximity to the extremely bright Sgr B2 point sources.
Additionally, this gap is not perfectly evident in NH3 or CO. In the HOPS NH3 maps, there is a relative lack of emission at ~56 km/s - relative to neighboring regions; this is actually the peak of the line profile extracted from the ellipse!
This is in contrast to the central "Gas Hole" which has been noted by others; it exists in CO and NH3. There is almost certainly some NH3 self-absorption involved. The "Gas Hole" coincides with the peak in other gas species and the peak in the dust emission, so it really seems to be a different phenomenon, having to do with line-of-sight confusion and excitation conditions.
The one feature of the "dust gap" that distinguishes this part of Sgr B2 from the other regions surrounding it is that 8 micron emission can readily be observed here. A little 24 micron emission is in the gap as well, but not as much as the 8 micron. The 70 micron emission is quite weak, but not really deficient. Perhaps in this region, very small grains are more common than in its surroundings?
There's a lot more hypothesizing to be done here.
Blogging from IPython
[ Notebook Form of this post ]
There have been many other posts about blogging in ipython, but they all included more overhead than I really wanted to deal with.
Instead, I've gone directly to the source and used nbconvert in ipython 2.0 to convert my notebooks to rst, then put them through pelican.
This post outlines my workflow.
First, and most inconvenient, it is necessary to head any blog post with metadata. I don't have any convenient workflow to deal with that, I just write it in by hand in a raw text box (ctrl-m, t) as you can see above. Daniel Rodriguez took a different approach using the ipython metadata directly and an ipython plugin, but I didn't want to have to worry about installing plugins and I am definitely worried about breaking notebooks by messing with the metadata as the ipy devs give some fairly strict warnings.
Next, the ipynb -> rst step is fairly straightforward. I modified the rst template because I don't want to see the In[#]: and Out[#]: prefixes around my code. I also use .. code-block:: rather than .. code::.
This very simple function below makes sure that my local rst.tbl file is seen before ipython's. Their templates are in IPython/nbconvert/templates. Note that you could normally use Jinja2 templating and "Extend" their template, but I wanted to remove rather than extend.
[ Source ]
from IPython.nbconvert.exporters import RSTExporter
def export(nbname, outfilename=None):
exportRST = RSTExporter()
# exclude default paths
exportRST.template_path = ['.','/Users/adam/repos/blog']
(body,resources) = exportRST.from_filename(nbname)
if outfilename is None:
outfilename = nbname.replace("ipynb","rst")
with open(outfilename,'w') as f:
f.write(body)
return body,resources
I then run the script from the command line: ~/virtual-ipydev/bin/ipython ./nbconverter.py content/GenerateCVExample.ipynb
I'm using a virtual environment with ipython 2 installed because I'm not yet ready to make the jump to the dev version of ipython (though these days, with Travis and Jenkins around, it's probably safe to assume the dev version won't break anything).
With this rst file generated, the only remaining step is pelican, which only requires a make github command in the blog directory. Installing & setting up pelican is reasonably easy, but not the topic of this post.
I've made my own custom pelican theme, so I generally need to update the theme before building:
pelican-themes --upgrade /Users/adam/repos/pelican-themes/mine && make github
Page 1 / 1