| The symbol to the right indicates students who were supported under the National Science Foundation (NSF)'s Research Experiences for Undergraduates (REU) program. All other students were supported under the NRAO Summer Student Research Assistant Program. Follow this link for a list of all student programs at the NRAO. You can also view the student projects titles from 1991-present in tabular form. |
|
For descriptions of the individual summer programs,
including itineries, pictures, lectures, and joint student projects,
see the previous
Program Reports.
Select a date range, student program, and site below:
2004 Summer Students |
In order to successfully detect the highly red shifted HI associated with the Epoch of Reionization (EoR), a proper understanding of the galactic and extragalactic foreground contaminants is necessary. An optimum method to do that is by radio interferometry. An EoR interferometer requires antenna elements which:
An antenna which suits the above criteria was found to be a "disc dipole" antenna, with a pair of dipoles in quadrature encapsulated by a pair of metallic discs, such that the axis of the each dipole is parallel to the diameter of each of the discs . For optimum performance, the antenna is placed at a specific height from an electrical ground plane covered with wire mesh.
Biswas was involved in the design, simulation, fabrication and deployment of the antenna elements for the prototype "Epoch of Reionization interferometer." The original design of the EoR antennas was a modification of the sleeve dipole done by Dr. Richard Bradley at the NRAO CDL. Biswas evaluated and modified the design of the EoR prototype antenna using the CST Microwave studio package, running simulations of the antenna by varying dimensions of its individual substructures. A set up with the correct dimensions of the antenna substructures was selected, which gives overall optimum antenna performance.
In addition to the design and simulation of the prototype antenna element, Biswas was involved in the design of the antenna mount. This was designed with the intention of providing a stable set up for making the antenna securely fastened to make data collection possible. Two antenna elements and mounts were constructed in Green Bank WV in early August and eventually deployed in the period between Aug 16-22, 2004. Biswas was involved in the deployment of the entire interferometer array including setting up of the east-west interferometer baseline using GPRS technology, linking up each antenna to the back end electronics using coaxial cables, and testing the performance of the array as a whole.
This summer Brian developed software for the rapid reduction of large quantities of VLA continuum and spectral line data. The first component of the software consists of procedures written for the Astronomical Image Processing System (AIPS) which can quickly accomplish basic flagging, robust calibration, and simple imaging of a single, multi-source UV data file. The second piece consists of a python script that drives these AIPS procedures through numerous UV data sets. The script is designed to require almost zero user input while keeping the user informed of progress and possible problems. It accomplishes this task by gathering all of the information relevant for reduction from the raw data header and various VLA calibrator manuals, then delivering it to AIPS. This pipeline is well suited for projects involving large amounts of public data from the NRAO Data Archive. A write-up of this work is available here.
This summer, Charles worked in the Quiet Skies Project testing the prototype small portable RFI detector, calibrating it, and developing a curriculum for its use for inquiry based learning for middle school and high school students.
Pre-protostellar Cores, or PPCs, are cold, dense clumps within nearby star-forming molecular clouds that are detected only through their far-infrared to millimeter emission. Since they have no (or very low) internal luminosity, but high column density, they are plausible candidates for the stages just before the formation of a central protostar. Important information can be gained through the study of molecular tracers in these cores.
Diazenylium, or NNH+, has its J=1-0 hyperfine transition at a frequency of 93.17 GHz. Three starless cores, L1498, L1512, and L1544, have been observed using the Nobeyama Radio Telescope with BEARS (a 5x5 array of SIS receivers). The three starless cores are located in the Taurus molecular cloud complex, and have central densities of 104, 105, and 106 respectively. It is hoped that they represent an evolutionary sequence. Jana reduced and analyzed these data, producing final maps consisting of 100 spectra separated by 20.55".It has been assumed in above the that NNH+ emission is a good tracer of the dust continuum emission, but studies that show this have generally had poor spatial resolution. Jana used the increased resolution of the new data to show that the NNH+ still accurately traces the dust continuum emission for these cores, matching the center position and aspect ratios. She found that the abundances remain constant throughout the core, with no evidence of central depletion. In addition, the velocity dispersion remains constant throughout the core. The spectra showed some evidence of anomalous hyperfine structure and blending, which interfered with the calculation of the optical depth correction. A blue asymmetry was observed in the line in the central regions of L1544. Jana also reduced observations of the NNH+ J=3-4 transition taken at the Caltech Submillimeter Observatory. The line was seen at the central positions of L1544 and L1512, but not observed in L1498. A write-up of this work is available here.
Since their introduction as a new class of active galactic nuclei (AGN) introduced 10 years ago, Compact Symmetric Objects (CSOs) have yielded insights into AGN evolution. They have emission on both sides of the central engine in the form of hotspots and jets. Their small sizes (< 1 kpc) are attributed to their youth (previously accepted as being between 350 and 2000 years old.) These ages are derived from the motions of the hotspots as they move away from each other with time. However, this has only been done for a small sample of CSOs.
Nicole studied 3 epochs of VLBI data at 8.4 GHz for 11 CSO candidate sources from the COINS sample (CSOs Observed in the Northern Sky). Along with Alison Peck (CfA) and Greg Taylor (NRAO) and Marcello Giroletti (IRA), she identified 5 of these as new members of the CSO class due to morphology, spectral index maps, and proper motions. Four sources were rejected on the basis of their core-jet morphologies and spectral indices, leaving 2 candidates. With these and 10 other CSOs from COINS, Nicole measured the proper motions of the hotspots and derived most age estimates between 240 and 3000 years, with one source appearing to be only 20 years old. She also detected polarized flux in two CSOs, a first for this class of objects. A write-up of this work is available here.
New and archival Very Large Array (VLA) data at 6 cm wavelength was combined and analysed for faint radio sources in the central 10 arcmin of the Andromeda galaxy. Kelley compiled a catalog of radio sources and compared them to sources found in previous radio surveys of M31 at 4 cm and 20 cm with the VLA, and to X-ray sources recently found by Chandra. The field was found to contain 3 SNRs and 312 point sources with an integrated flux 4 sigma, 15 microJy, above the noise, including the nuclear black hole candidate M31*. The majority of the sources found at 6 cm did not have multi-wavelength counterparts, however those with detections at other wavelengths will be used to determine the relative astrometry between the X-Ray field by Chandra and the radio field seen by the VLA. Ultimately, this information will be used to identify the X-ray counterpart of M31* from among 3 candidates separated by only a few arcseconds. Kelley and Lorant anticipate the completion of 48 hours worth of real simultaneous observations with Chandra and the VLA this winter. This will improve current 6 cm VLA observations by doubling the total observing time, and allow one to study the variability of sources in the field, specifically M31*, at multiple wavelengths.
Supernova SN1885A lies less than an arcminute from the core of M31. It creates a striking impression in Hubble images, however it had not previously been detected at radio wavelengths. Kelley confirmed no detection of SN1885A in 6 cm at 15 microJy and but it remains unclear at the 4 sigma, 12 microJy, level at 4 cm. Three candidates were detected in the vicinity of SN1885A's optical image, but more analysis must be done before any of these are accepted or rejected as a possible radio counterpart. A write-up of this work is available here.
The instrumentation on the GBT operates at very wide bandwidths and at very low noise levels. It is difficult for ordinary laboratory instrumentation to be used for diagnosis and testing of these sensitive instruments. The NRAO at Green Bank obtained the ALMA Test Correlator hoping to create from it a scale replica of the Spectrometer being used in the GBT control room. This replica would be used to test the instrumentation without having to take observing time away from astronomers. This summer, Brad worked with the Green Bank engineers to modify this Lab Spectrometer to suit the needs of the technicians. This included adding additional cross-correlation modes, integrating the device into the existing system, and creating interfaces emulating the real Spectrometer. The Lab Spectrometer is very near to complete. All that remains is minimal debugging of the computer code and final testing. Brad used and developed his skills in digital design, computer programming, and hands-on digital troubleshooting. At the end of the summer he documentmented the final product. A write-up of this work is available here.
The Green Bank SRBS is divided into three distinct stages of implementation. The first consists of a dipole antenna and a spectrometer which scans from 18 to 70 MHz at intervals of four seconds through out the day. This is the system which is currently being used. The second proposed stage replaces the dipole antenna with the LPA (Log Periodic Array) and digital spectrograph which will scan the frequency range from 30 to 350 MHz and adds the 13.7 m dish that will provide a range from 50 to 800 MHz using the Swiss built instrument Callisto. In the final stage Callisto will be replaced with a broadband digital spectrograph capable of ranges from 300-3000 MHz. Also included in this stage will be a new dipole antenna operating in the 10 to 30 MHz range.
Sarah participated in the third stage, creating a prototype for the broadband digital spectrograph that will be used on the LPA in the second stage and for all instruments in the third stage. The broadband digital spectrometer consists of a frequency converter and a PC with a data collection card. Sarah built a filter for the frequency converter and wrote programs to acquire and process the data, and to display a radio spectrum.
The ALMA correlator software needs to be tested during its development. This software is tightly bound to the ALMA correlator hardware, but often the correlator hardware is not available for testing against the software.
For this reason, Carlos worked to develop a correlator simulator which simulates the functional interfaces to the correlator control computer (CCC) and the correlator data processing computer (CDP). The correlator simulator is a Linux PC running a modified real-time Linux kernel which accepts CAN commands from the CCC and returns responses to it in accordance to established correlator/computer interface. The correlator simulator also produces raw lags via a Python script which is the raw input for the CDP. Carlos focused his work on the CCC interface which accepted control commands, validated the protocols, and returned valid responses. Carlos helped to design, develop, test, debug, and document the correlator simulator system within the framework of the existing ALMA correlator software.
This summer, James created a data analysis program to display pulsar information coming from the Green Bank Telescope pulsar spigot card in semi-real time. The program can then be used both for snap-shot images of pulsars, to insure that data quality is good, and to monitor the system power levels. The program was written in the python language to allow ready portability between various software platforms (windows, Linux, etc.).
This summer John completed data reduction of approximately 50 hours of VLA 20 cm observations of the Chandra Deep Field South, which includes the new Hubble Ultra Deep Field, taken February and March 2004. The elevation of the CDFS field at the VLA is very low, compounding problems with ground pickup and interference, as well as limiting the observable time to only six hours per day. To obtain the highest fidelity image possible, John broke the data into small time segments and cleaned each of these separately before combining them in the image plane. This removed time variable effects and allowed much improved images to be made. The deepest image made using the 2004 data had a noise of ~15 uJy in the center of the field, somewhat higher than the theoretical, likely due to dynamic range effects from the bright sources. This was combined with a previous set of observations to produce an image with a noise ~7 uJy. John also began preliminary work on optical and X-ray comparisons with data from Hubble, Chandra and the VLT. The Chandra data contained about 50 obvious X-ray counterparts to radio sources, as well as a systematic offset in both RA and DEC which must be accounted for in future analysis.
Density fluctuations in the galactic neutral hydrogen play a significant role in the dynamics of the ISM and have been observed on a wide range of scales with aperture synthesis telescopes. Our goal was to establish the relationship between the parsec-scale fluctuations detected in this experiment and the observed A.U. scale fluctuations that are seen in the neutral ISM. We aimed at finding the variation of the strength of the HI turbulence with Galactocentric radius. The results would help us determine if the HI fluctuations are a global property of the galactic HI and give us insight on the possible sources of energy for the HI turbulence.
Four lines of sight, towards HD 18537, PSR 0540+23, 220.0+0.0, and 149.5+0.0, were chosen in the outer Galaxy where the velocity-distance relationship is monotonic. Mariana calibrated VLA spectra-line data from five observations at 21 cm. The data were then exported to ASCII files for analysis. Since the spatial power spectrum (SPS) is obtained directly by squaring the amplitude of the complex visibility function, no imaging was required. Mariana developed two IDL programs to (1) estimate the noise power spectrum, and (2) create SPS plots after the noise contribution had been corrected for.
After about 50 years, theoretical and observational research show that the solar chromosphere is not thoroughly understood. Most models have been static. Static models of the solar chromosphere are based on UV/EUV line and IR/mm continuum observations. However, observations indicate a dynamic state. Dynamic behavior is visible in lines and continua observed on the SOHO spacecraft and Ca II resonance lines. This gives rise to conflicting views of the structure of the poorly understood chromosphere. It is believed that acoustic waves generated in the solar atmosphere propagate upward and form shocks, dissipating their energy and heating nearby material. We question whether these waves play a role in the structure of the chromosphere.
In order to examine the dynamics of the chromosphere, we have made VLA observations because of the VLA's ability to make high resolution two-dimensional pictures of the sun relatively quickly and at several frequencies and also because of the fourier transform relationship that exists between an interferometer and the image formed. The image formation is a result of this mathematical relationship between the visibility function and the radio brightness distribution. Observations were made in 7mm, 13mm and 20mm bands over a two day time frame. We have obtained data made in Ca II from Big Bear solar Observatory; all data was taken in 10 or 15 second cadence. Using AIPS (UVPLOT and ISPEC) we reduced data and obtained time sequence images, formed data cubes (compute the brightness variation as a function of time at different locations), created light curves and formed preliminary k-? diagrams. Our preliminary finding is that there are no oscillations in the radio.
Chun reduced recent VLBA+Gb+Eb 43GHz observation of M84 where the northern jet has been detected. Poor weather (e.g., snow on GBT and rain on Eb) limited the sensitivity of these measurements. Chun also analyzed VLBA/VLBI observations of M87 taken between 1999 and 2004. He found that a counter-jet exists at 43GHz and 22GHz, and determined that the counter-jet is not free-free absorbed. In addition, the edge-brightened structure and changes in the jet structure are seen. These data should help motivate a proposal to make a movie of M87. Such a movie will hopefully provide clue on how jets are being collimated very close to the supermassive black hole. A write-up of this work is available here.
GigahertzPeaked Spectrum (GPS) sources are typically characterized as young sources representing an early stage in the development of larger, structurally complex radio objects. They are believed to be young radio jets forming near black holes in structures identified with galaxies and quasars. The GPS sources have similar morphologies to older sources, but a more compact and simple structure. They are identified by a characteristic spectrum, which peaks, appropriately, around 1 GHz. Other criteria for selection include sufficient brightness, spectrum width, and overall shape of the spectrum. Further understanding of these sources depends upon the collection of a larger, more complete sample than has been available in the past. Additionally, the ideal sample would be based upon simultaneous broadband measurements, allowing the observer to work with instantaneous spectra. Nonsimultaneous measurements are inadequate because GPS sources tend to be variable. In order to best understand the physics of GPS sources, and to prevent bias in the selection of a sample of GPS sources, one should avoid making conclusions based on observations taken days, months, or years apart.
Ann selected a sample of GPS sources from a sample of 3000 quasars and radio galaxies brighter than 100 mJy observed with the Russian 600m ring radio telescope RATAN. RATAN provides simultaneous broadband measurements in range of 122 GHz, avoided some of the pitfalls mentioned above. Sources for the sample were selected in many intermediate steps. In the first, 348 possibilities were culled from the original 3000 spectra. This selection was based on the overall shape of each spectrum. Ann fit both parabolic and linear models to the spectra in order to estimate interesting parameters such as the spectral indices, the turnover frequency, and the turnover flux density. She further refined the sample according to how well the data fit the predicted selfabsorbed synchrotron model. Sources were examined for changes in flux density over time, in order to investigate a theoretically predicted increase in the flux density of optically thick synchrotron emission. The results of the project included two lists of sources, one of which was composed of definite GPS sources, and the other of possible GPS sources which demand further investigation. Ann identified 146 sources as GPS sources, with another 28 possibile sources.
DEAP (Data Extraction and Analysis Program) is a modular package for the interactive display and editing of scientific data with the goal of producing publication quality plots. DEAP is written in Python and incorporates PGPlot, a well-known plotting package in astronomical applications. DEAP was initially developed by a group of graduate students from the University of Maryland Software Engineering program, which included one NRAO employee. DEAP is also used as a framework for bigger packages such GBT FITS Monitor (GFM), which is used as a "quick look" tool by observers at the GBT to show data as it is generated throughout an observation.
This summer, Dani added a 2-D imaging color display capability to DEAP. PGPlot functions were used to set the color table and draw the image. She also implemented a Y2-axis display in DEAP. The Y2-Axis allows the observer to plot two XY plots in the same window using different scales, meaning one plot will be scaled using the Y1-Axis while the other uses the Y2-Axis. Dani prepared and implemented a full design of both new plotting capabilities, including documentation and examples.
Becca's project was to assemble, align, and debug the experimental
setup for making automated 600-720 GHz receiver noise
measurements. She spent the first half of the summer outfitting the
receiver cryostat with wiring and the appropriate components to test
mixers being made at the University of Virginia's Electrical
Engineering lab. The purpose of this receiver is to measure different
types of Hot Electron
Bolometers (extremely sensitive thermometers that heat up when an
incident photon strikes it). Becca assisted in testing the mixer chips
(which contain the HEB's). The purpose of these tests is to
characterize different types of HEB mixers and decide which will be
best for a terahertz receiver. To do so, she set up a Martin-Puplett
Interferometer (similar to a Michelson Interferometer except that it
aligns the polarizations) which is being used to send the signal into
the receiver. This involves sending a Local Oscillator (LO) signal and
a load (which simulates an astronomical source) through the
interferometer and into the mixer chip. The mixer chip mixes the
signals and reduces them to an Intermediate Frequency (IF). We then
read the power being produced by the system and we can calculate the
system noise. Becca wrote a LabView program to read the power and
control a tunable filter. The filter sweeps through the IF range, and
a reading is taken at each step. A write-up of her final report can be found
here.
Lin worked on the chopper software for ALMA Band 6 mixer and cartridge measurement systems. The goals for the summer were to modify the chopper circuit, to write a LabView program, to integrate it with the BAND6DAQ measurement software, and to test different input parameters to optimize the chopper speed. The program synchronizes the spinning chopper wheel to a continuous record of hot load and cold load measurement using the power meter, calculating the y - factor by dividing the hot load noise power by cold load noise power. Lin restructured the algorithm to make it more understandable, simplified the architecture for easier code maintenance, and most importantly, resolved the synchronization problem the previous chopper routine had when it missed one trigger signal. The student also successfully sped up the chopper, shortening the full bandwidth test by 16 minutes, resulting in a 22% reduction in test time. The software development process was documented to meet the design standard, which involved changes to the hardware, old commands being discarded and new functions being created in the software, and goals being achieved on both test systems. A write-up of his final report can be found here.
The discovery of radio emission from brown dwarfs (accomplished a few years ago by a group of NRAO summer students!) surprised the stellar community, and raised deep questions about the continuance of activity into ultracool stellar objects.
Lynnae worked with Dr. Rachel Osten in studying the characteristics of radio emission from a sample of sources identified as brown dwarfs from the Two Micron All Sky Survey (2MASS) and the Sloan Digital Sky Survey (SDSS). These sources were recently observed with the VLA in X-band (3.6 cm). Lynnae calibrated and imaged these data in AIPS, and measured the flux density of any associated radio emission. In all, Lynnae reduced and analyzed data for nine brown dwarfs (five L dwarfs and four T dwarfs), eight of which were 2MASS sources and one SDSS source. Radio emission was detected from three of the 2MASS objects: 2MASS 0652+472, at a distance of 11.1pc and of spectral type L4.5; 2M0559-1404 at a distance of 10.24 pc and of spectral type of T5; and 2M0857+5708 at a distance of 11pc and of spectral type of L8. A write-up of Lynnae's final report can be found here.
Image analysis, such as component fitting of radio interferometric images has traditionally been based on likelihood techniques applied to deconvolved images. The analysis usually ignores the uncertainty arising from the process of deconvolution, a non-linear inverse problem. One would therefore like to estimate the properties of components representing the entire emission present in the raw, dirty image. In practice, this is not feasible given the large dimensionality of the parameter space in a pixel based method. A scale sensitive representation of an image as a collection of elliptical gaussian components can control this problem.
Urvashi's project involved the design and implementation of an algorithm that performs a Bayesian image analysis to fit elliptical gaussian components to sub-regions of the dirty image, taking full account of the point spread function. Prior to the fitting of emission inside a given region, the emission outside the region is removed by subtracting a previously deconvolved image. This method produces samples of the posterior distributions for the number and parameters of elliptical gaussian components within the region of interest. Information about relative probabilities and uncertainties associated with the component parameters can augment the process of object detection and characterization. Urvashi compared the performance of this approach to the standard methods. She also tested the algorithm on simulated as well as real data sets, and implemented it as an aips++ glish tool. A write-up of this work is available here.
This summer, Steve worked with the Radio Frequency Mitigation group at Green Bank to investigate the mitigation issues in and around Green Bank. He field tested a small portable RFI detection unit being developed to determine its limitations, calibration, and usability by middle school and high school students for inquiry based learning projects.
SS433 is one of the best known precessing jet sources, and the only jet whose composition, temperature, and density are known from X-ray and optical atomic emission lines. It is also a lovely radio source, as the precessing relativistic jets trace out loops on the sky which can be followed for many arcseconds (corresponding to a few years in the lives of the radio ejecta). We have recently completed a major imaging campaign, covering over a quarter-cycle of the 164-day precession period with daily VLBA and weekly VLA observations to trace the moving ejecta.
In preparation for her project, which was to analyze the sequence of images, Kirstin learned to reduce VLA and VLBA data in AIPS, and attended the 9th Synthesis Imaging Summer School. She then undertook a study of the proper motions in the jet of SS~433, which have been detected optically. She found that while individual radio components travel with steady velocities, these can deviate significantly from the optical Doppler shifts. Ultimately this means that the velocity of the individual bright radio-emitting features differs greatly from the bulk velocity of the underlying baryons that emit the optical lines. The ballistic model however, fits the optical lines and the radio morphology extremely well. This is the first iron-clad proof that the radio components display pattern motion (phase-velocity effects) rather than bulk motion - in no other case is the material velocity known.
Kirstin also analysed the so-called anomalous emission, slower outflows of low surface brightness, oriented perpendicular to the jets, and seen in no other source. Apart from confirming the reality of these outflows, quantitative evaluation of their spectral index and outflow speed proved to be difficult to extract from the present dataset.
Kirstin's work formed part of the analysis of SS433, presented by Amy Mioduszewski at the 2004 High Energy Astrophysics Division meeting in New Orleans, Sept 2004; and will be presented ay Kirstin at the January 2005 AAS. A write-up of her final report can be found here.
Transient radio sources are evidence of unique phenomena and are important probes of fundamental physics. A number of groups have detected both galactic and extragalactic transient sources. These observations have led to predictions that long term observations will yield a background of transient events, in imaging surveys.
Sarah's project was to develop a technique for the detection of transient radio sources using Very Large Array archive data on M81 and A2192. These multi-epoch datasets allow for the detection of transients on timescales from < 1 day to 1-17 years. To find long-term transients, Sarah developed modified a program to compile the fluxes and location of all sources in a field. These were compared to the NRAO VLA Sky Survey (NVSS). Sources found in one dataset but not the other were considered candidate long-term transient. To search for short-term transients, she generated image cubes with each plane consisting of 30 seconds worth of data. An average of all plans was subtracted out, leaving only noise, interference, and variable sources.
Two long-term transients were detected. Many possible short-term transients below the accurate detection level of 10mJy and 10sigma were detected, but no reliable candidates above this level were found. A write-up of these results is available here, and a webpage is available at http://www.gb.nrao.edu/~glangsto/m81/.
The fundamental goal of scale sensitive deconvolution algorithms is to decompose the image in a set of scale sensitive basis functions (called Adaptive Scale Pixon or Asp components). Since such a basis is potentially non-orthogonal, the algorithms are computationally expensive. This expense is further increased for interferometric imaging due to the fact that the point spread function (PSF) couples widely separated pixels in the image. For complex astronomical images with large number of components, search for the optimal set of scales for these Asp components becomes computationally expensive. The computational expense of the scale sensitive Asp-Clean algorithm (Bhatnagar & Cornwell, A&A, 2004, in press), is of order (N2 log(N)) per optimization step per Asp-component. Wavelet transform decomposes the image into spatially localized but orthogonal functions at different scales in O(N2) computations. It can thus provide information about a number of scales faster, compared to the computational load of the Asp-Clean method.
In this project, Anand explored the use of a wavelet transform to design a quicker method to simultaneously estimate the number of dominant scales in the image. Toward this goal, Anand implemented a wavelet server in AIPS++ using a fast wavelet transform scheme (the Lifting scheme). This is now available as a standard tool in the AIPS++ package. He then explored algorithms related to the technique of thresholding in the wavelet space (ForWaRD algorithm, Neelamani et al., IEEE Sig. Proc, 2003) and implemented a simple scheme of thresholding the wavelet transform of the residual image to estimate the dominant scales. He found that hard thresholding at various scales did provide a crude but quicker estimate of the local scales in the residual image, and outlined an iterative scheme to construct a model image. Further work to feed this information into the Asp-Clean algorithm is in progress.
Water masers are bright, compact objects that emit strong radiation at 22.2 GHz, and are perfect Very Long Baseline Interferometry (VLBI) targets to trace the motions of the gas where they are found. In 2003 Dr. Claussen observed a strong red-shifted water maser toward Class I young stellar object (YSO) YLW 16A with the VLBA. YLW 16A is located in a very active star formation region of the rho Ophiuchus cloud, which is only 140 pc distant. Several VLBA proper motion studies have been done toward Class 0 YSOs, and in general have found that water masers trace proper motions along the axes of large-scale gas outflows. These are the first VLBA observations toward a Class I YSO.
Christine reduced and imaged six epochs of phase referenced VLBA data spread over two months in 2003. These data were part of the first phase referenced observing program of water masers. She determine absolute positions for the maser spots via phase referencing, and determine the masers' three-dimensional velocities and angles of inclination to the sky. She found that these masers are moving much slower than masers toward Class 0 YSOs for which similar studies have been done. She also found that the masers are traveling close to perpendicular to the plane of the sky, which is consistent with 12CO maps of the large-scale molecular outflows toward this source. A write-up of this work is available here.
An ortho-mode transducer is the part of a radio frequency receiver that converts electromagnetic waves into electric currents. A new OMT is needed for the GBT receivers. Gain and noise temperature stability must be improved by using fewer connectors and better physical temperature stability. The rf passband must be made smoother in both gain and noise temperature through the elimination of trapped mode resonances, the reduction in path lengths, and the reduction in impedance mismatches. A lower noise temperature could be obtained by reducing losses and by improving the impedance match to the low noise amplifier and waveguide. The active balun approach, where four low-noise amplifiers are integrated into a new balanced OMT design, shows promise in achieving the above goals while also increasing the operational bandwidth
Dave's project was to provided reliable performance expectations for a newly designed OMT. Dave used a computer aided design program (Computer Simulation Technology's, "Microwave Studio") to evaluate and and improve different designs of the OMTs. According to the simulations, this new OMT will be significantly better than the one it is intended to replace, which currently in use in several receivers of the GBT. A write-up of Dave's final report can be found here.
The Cosmic Background Imager (CBI) is a small 13-element interferometer that operates in Chile at the ALMA site in the Andes. Since 1999, the CBI has been observing the cosmic microwave background (CMB) and the Sunyaev-Zeldovich Effect (SZE) in clusters of galaxies (at radio wavelengths clusters appear dark against the CMB due to Compton upscattering of CMB photons by cluster gas). The CBI team has been observing a large sample of clusters.
Using both radio and previously-published X-ray data, Adrienne studied the SZE in galaxy clusters Abell 3558 and Abell 3266 in depth this summer in order to derive both information about the gas structure in the clusters as well as the Hubble constant. Because foreground sources can contaminate the signal in visibilities, they must be subtracted. Adrienne used VLA data to determine the flux and the location of these sources, and subtracted them from the CBI data in the uv-plane. She wrote a program in MATLAB to model the residual visibilities from the equation derived for the SZE intensity on the sky. Finally, she found the best fit model for the data by using two methods: maximum likelihood and Markov Chain Monte Carlo distributions. From the parameter models (one of which included the angular distances to the clusters), she was able to calculate the gas distribution in the clusters in addition to the Hubble constant. For the two clusters we obtain a value of 67.4 ± 14.7 km/s/Mpc, which is consistent with the currently-accepted value of 72 ± 5 km/s/Mpc measured by WMAP. However, this error includes only statistical errors from the CBI data and the published X-ray models. We hope to continue the project into fall 2004, adding in other factors like a noise term due to the CMB. We would also like to improve our methods of dealing with foreground source contamination and of computing best-fit models. A write-up of this work is available here.
The supernova remnant CTB 80 and the associated radio pulsar B1951+32 form a fascinating system to study the interaction of a neutron star with its environment, with the pulsar traveling toward the edge of an inner core within the larger remnant. For his summer project, Ben combined 1.4 GHz archival VLA data taken in 1989 and 2003 with the array in its A-, C- and D-array configurations. He produced images of the nebula as it appeared at both epochs in order to study its structural evolution. The 14.5 yr time baseline allowed an improved determination of the pulsar proper motion: Ben measured a proper motion of -29.4+-1.3 mas/yr in right ascension and -16.7+-1.3 mas/yr in declination, an improvement in precision by a factor of two over previous measurements. Between the two epochs, Ben found that the stand-off distance from the bow shock to the pulsar decreased by 44 mas, indicating a change in the density of the ambient medium through which the pulsar is traveling and providing support for the models of CTB80's inner core as a pulsar wind-driven nebula.
In the future, images created as part of this project can be compared with x-ray and optical counterparts for correspondence. Further analysis of the evolution of the inner core and continued observations of the pulsar will also improve knowledge of the system. Using the reference sources found in this study, VLBI measurements of the proper motion are also possible. A write-up of this work is available here.