Searching for Pulsar Wind Nebulae at the Kavli Institute for Particle Astrophysics and Cosmology14 Feb 2013
Pulsar Wind Nebulae
I got a PhD from Stanford University in Physics. In graduate school I studied the Gamma-ray emission from pulsar wind nebulae. A pulsar is a rapidly rotating neutron star. Pulsars were first discovered in 1967 by Jocelyn Bell Burnell. Pulsars are typically powered by the energy released when a neutron star slows down. Much of this energy is released as a wind of electrons, which interacts with the interstellar medium. This forms a diffuse cloud called a pulsar wind nebulae that can be observed in gamma-rays.
The Fermi Gamma-ray Space Telescope
During my thesis, I observed pulsar wind nebulae with the Large Area Telescope (LAT), the main scientific instrument on board the Fermi Gamma-ray Space Telescope (Fermi). Fermi, pictured on the right, is a pair-conversion gamma-ray telescope. Cosmic gamma-rays are interesting to study because they originate in the most extreme and energetic astrophysical environments. Here is a good description of the design of the instrument. Fermi was launched in June of 2008 with a designed mission length of 5 years and a goal of operating for 10 years. Here is an artist rendering of the mission launch.
The first detection of gamma-ray emission from a pulsar wind nebula was the crab nebulae, detected by a previous gamma-ray detector called the Energetic Gamma Ray Experiment Telescope (EGRET) in 1991. Since then, the Vela X and MSH 15-52 pulsar wind nebulae were detected by Fermi in 2010.
The first paper I was involved with was a search for new pulsar wind nebulae that emit gamma-ray radiation. In this paper, we selected 54 pulsars which had been previously detected by Fermi. Because the emission from pulsars is pulsating in time, we could select and remove the pulsed emission. In this so-called off-pulse window, we performed a search for gamma-ray emission that could come from pulsar wind nebulae.
In this analysis, we discovered a new pulsar wind nebula associated with the Westerlund 2 region (see above). In addition, we performed a population study of four gamma-ray emitting pulsar wind nebula. We observed that young and highly energetic pulsars power gamma-ray emitting pulsar wind nebulae.
The full text of the paper can be read for free at the arXiv.
The next paper I worked on focused on developed new methods to study the spatial structure of sources detected by Fermi. Before Fermi, previous gamma-ray telescopes had relatively low statistics and a notoriously-poor angular resolution. Given its improved angular resolution and better statistics, Fermi was the first gamma-ray detector capable of spatially resolving a large number of gamma-ray sources.
Studying the spatial structure of a Fermi source is important because it is often difficult to find a counterpart for a Fermi source observed at other wavelengths (for example, radio, optical, or x-ray). The second Fermi source catalog detected 1,873 sources, but 575 of them could not be associated to a known source counterpart. In addition, often there are several potential counterparts to a Fermi source. In many situations, the spatial shape of the observed source can be compared to potential counterparts to uniquely classify the source.
In the paper, we developed a new method to study spatially-extended Fermi sources. We used the Fermi data fitting-fitting package called pointlike (described here) and added functionality to the program to fit the shape of an assumed spatially-extended Fermi source. We then defined a test for the statistical significance of the detection of extension and validated this test with an involved monte-carlo simulation. Next, we computed the sensitivity of Fermi to detected spatially-extended sources. We then validated this test by applying it to a active galactic nucleus, Fermi sources which are known to be pointlike.
Finally, we took this new method and applied it to two years of Fermi data to search for new spatially-extended Fermi sources. In our search, we discovered seven new spatially-extended sources, bringing the total number of extended sources to 21. In particular, we detected the source RX J1713.7-3946 to be spatially-extended (see the picture above) and the extended sources HESS J1616-508 and HESS J1615-518 (pictures on the left). The location of the extended Fermi sources is overlaid on a map of the gamma-ray sky below. For these sources, this spatial analysis was important for identifying the nature of the gamma-ray emission. The full text of the paper can be read for free at the arXiv.
The pointlike maximum-likelihood package
During my PhD, I have been involved with the development of a maximum-likelihood package to fit LAT data. In the process, I have learned about software management, code refactoring, interface design, version control, and issue tracking. I have written about the program here.
The slides from my PhD talk are on slideshare.