I'm interested in the structure of the Milky Way Galaxy.
So far I've worked on a couple of aspects of this topic:

HI spiral arms in the outer disk
Leo Blitz, Carl Heiles and I made a map of the spiral structure in the outer Milky Way HI disk. We used
the LAB survey to make a surface density map, and then used the unsharp masking technique on that map.
The resulting figure is shown below. We found well-ordered spiral structure at southern latitudes, and less
distinct structure in the north. None of the arms we found were new, but our map shows the arms better
than any previous map. We were also able to show that the overdense regions are coincident with regions
of reduced gas thickness.


This is a plot of the perturbed surface density in the outer Milky Way disk. The orange-red regions are overdense,
and the greyscale regions are underdense.

This work was the subject of an article in the nytimes (reg required).

Warp in the HI gas in the outer disk
Leo Blitz, Carl Heiles, and I did a quantitative analysis of the structure of the outer Milky Way HI disk.
We constructed a height map from an all-sky HI survey done by the LAB consortium. We used a Lomb
periodogram to reveal the Fourier structure of the warp, and found that a superposition of the first three
modes well describes the major features. These are an m=0 bowl shaped mode, an m=1 S-shaped mode,
and an m=2 saddle shaped mode. Our fit is similar to one published in Binney & Merrifeld 1998. The
similarity of our three mode fit to the data is shown in the following picture:

The left panel is the height map from the data and the right panel is our fit. The labels on the colorbar are
in vertical kpc; color shades are above the plane and grayscale ones are below the plane. The location of
the Sun is marked by the solar symbol. The lines labelled by W1 and W2 represent the lines of maxima for
the m=1 and m=2 modes, respectively.

This work was the subject of articles in the nytimes (reg required) and usatoday.

Scalloping in the outer disk
We then subtracted out our fit to the warp and analyzed the residuals. A global analysis with a Lomb
periodogram demonstrated the strength of the m=10 and m=15 modes; a local wavelet analysis showed that
this power resulted from local disturbances. This answered the long standing question regarding whether the
scalloping went the full 2 pi around the disk.