R-band image with ROSAT PSPC contours
(Owen et al. 1999)
|
R-band image with ROSAT PSPC contours (Owen et al. 1999) |
Deep radio observations of the galaxy cluster Abell 2125 provide an opportunity to also observe the distant galaxies behind the cluster members. This ~300 galaxy sample is ideal for studying the distribution of sources with redshift. One question to ask is whether the flux distribution (log N-log S) for the Abell 2125 field matches deep surveys in other areas of the sky. Photometric observations have been added to the radio data in order to derive redshifts for the galaxies, cross checked and supplemented with spectroscopic redshifts. With the addition of redshifts, the radio luminosity function can be derived and star forming galaxies can be statically separated from active galactic nuclei (AGN). Now we can ask whether the faint radio population is dominated by star formation or AGN activity. It may also be possible to measure the star formation rate as a function of redshift. |
Preliminary results suggest that we see two distinct populations clustering at redshifts of 0.6 and 1.2, which may be signatures of the large scale structure extending behind the cluster. In addition, an excess of sources at the faint end of the distribution, while still being investigated for systematic errors, could indicate cluster lensing, providing an even deeper window into the behavior of distant galaxies
HST image of SBS0335-052. Image from Kunth & Ostlin 2000. |
Some of the most important and fundamental processes in astronomy are still poorly understood: one of these is star formation. While widely observed throughout the universe, both the preconditions and triggers to star formation are poorly known. Of particular interest is the initial burst of star formation which takes place in a chemically unenriched, low-mass environment. These are the conditions of the early universe described by hierarchical galaxy formation scenarios but are very difficult to study at high redshift. Such primordial environments can be studied at high resolution and close quarters in a class of galaxies called blue compact dwarfs (BCDs). Radio observations contribute two critical measures to the study of BCDs -- high-resolution maps can constrain the small sizes (tens of parsecs or less) of compact star formation regions, and multifrequency observations enable the separation of thermal and nonthermal emission. Currently, with the collaborators above, I am working on new Very Large Array (VLA) observations of the BCDs SBS0335-052 (Figure 2) and archival data of IZw18, as well as a more extended sample of low-metallicity (<1/10 solar) BCDs, also from the VLA archive. SBS0335-052 is a particularly fascinating source since it appears to be heavily absorbed at 20 cm. |
Single dish observations of supernova remnants
SN 1006 at 20cm with the GBT |
High resolution radio observations of galactic supernova remnants (SNRs) have revealed tremendous detail, especially in conjunction with ROSAT, ASCA and now Chandra x-ray observations. However interferometers such as the VLA lack the information needed to accurately reconstruct the absolute flux of many SNRs, making comparisons between radio and x-ray wavelengths, as well as between 6 cm and 20 cm, unreliable for many important SNRs. While it is well known the the addition of a single dish radio maps can correct this problem, this has seldom been done and calibration between telescopes has been an issue. The commissioning of the Byrd Green Bank Telescope, the worlds largest steerable radio antenna, represents a new opportunity to address this problem. The unique off-axis feed results in a clean response function, producing images with superior fidelity. I have obtained single dish radio maps of several important galactic SNRs with the Byrd Green Bank Telescope, to be combined with VLA observations in order to produce images with accurate fluxes. While this technique has significant impact on many fields in astronomy, two issues of immediate import to SNRs can be addressed: |
1. Testing shock theory with spectral index variations
The synchrotron emission seen in the radio from supernova remnants comes from relativistic electrons accelerated in strong shocks, interacting with the magnetic field. First order Fermi acceleration describes the shock acceleration mechanism, predicting a particle spectrum with an index of 2, leading to a spectral index of -0.5. The overall spectra of shell SNRs, while not identical to, agree with this prediction. It should be possible to go beyond this to probe the shock structure by measuring very small changes +/-0.02-0.03 in the spectral index across the shock. Many observers have looked for these changes, only to be hampered by the lack of true flux measurements. Algorithms designed to find changes under conditions of missing flux have proven unreliable. GBT measurements should allow true flux reconstruction, which along with accurate errors will enable the testing of shock theory.
2. Radio, X-Ray, and H
comparisons of shock location
The historical supernova remnant SN1006 AD has already played a key role
in our understanding of SNR shocks. It appears to be expanding into a
uniform upstream medium and its location off the galactic plane results in
low absorption, allowing both optical and x-ray observations of the shock.
The arcsecond resolution images in H and hard and soft x-rays show
a complex shock structure (Long et al. 2003, Figure 3), especially in the northwest quadrant of the
SNR, which is difficult to see in the VLA image since it is exactly there
that lack of zero spacing flux introduces serious artifacts in the map.
GBT data should produce an arcsecond resolution image with reliable flux
which can be directly compared to the optical and x-ray profiles.
Separating supernova remnants and x-ray binaries in other galaxies
High spatial resolution x-ray observations are revealing a population of
point-like sources in a host of nearby galaxies. These are
normally identified as SNRs or x-ray binaries (XRB). However, in
general these detections do not have enough counts to fully populate a
spectrum, leaving identifications extremely speculative. Using simulated
spectra from standard XRB and SNR models and data from well studied
Galactic remnants and binary systems I am investigating the uncertainties
in fitting low source count spectra. Early results indicate that with
Chandra CCD-type spectral resolution it is not possible to distinguish
between a plerion-dominated SNR spectrum and a standard binary model even
in well sampled data, although other SNR models do have distinguishing
characteristics. I also intend to test hardness ratios which may
distinguish between source types.
Jump to Kristy Dyer's
Homepage