The NOAO Data Lab aims to provide infrastructure to maximize community use of the high-value survey datasets now being collected with NOAO telescopes and instruments. As a science exploration framework, the Data Lab allow users to access and search databases containing large (i.e. terabyte-scale) catalogs, visualize, analyze, and store the results of these searches, combine search results with data from other archives or facilities, and share these results with collaborators using a shared workspace and/or data publication service. In the process of implementing the needed tools and services, specific science cases are used to guide development of the system framework and tools. The result is a Year-1 capability demonstration that (fully or partially) implements each of the major architecture components in the context of a real-world science use-case. In this paper, we discuss how this model of science-driven development helped us to build a fully functional system capable of executing the chosen science case, and how we plan to scale this system to support general use in the next phase of the project.
KEYWORDS: Data storage, Clouds, Distributed computing, Astronomy, Databases, Data storage, Clouds, Data conversion, Observatories, Data archive systems, Data storage servers, Java
Collaborative research/computing environments are essential for working with the next generations of large astronomical data sets. A key component of them is a distributed storage system to enable data hosting, sharing, and publication. VOSpace1 is a lightweight interface providing network access to arbitrary backend storage solutions and endorsed by the International Virtual Observatory Alliance (IVOA). Although similar APIs exist, such as Amazon S3, WebDav, and Dropbox, VOSpace is designed to be protocol agnostic, focusing on data control operations, and supports asynchronous and third-party data transfers, thereby minimizing unnecessary data transfers. It also allows arbitrary computations to be triggered as a result of a transfer operation: for example, a file can be automatically ingested into a database when put into an active directory or a data reduction task, such as Sextractor, can be run on it. In this paper, we shall describe the VOSpace implementations that we have developed for the NOAO Data Lab. These offer both dedicated remote storage, accessible as a local file system via FUSE, and a local VOSpace service to easily enable data synchronization.
Many astronomical image-analysis programs are based on algorithms that can be described as being embarrassingly
parallel, where the analysis of one subimage generally does not affect the analysis of another subimage. Yet
few parallel-processing astrophysical image-analysis programs exist that can easily take full advantage of todays
fast multi-core servers costing a few thousands of dollars. A major reason for the shortage of state-of-the-art
parallel-processing astrophysical image-analysis codes is that the writing of parallel codes has been perceived
to be difficult. I describe a new fast parallel-processing image-analysis program called crblaster which does
cosmic ray rejection using van Dokkum's L.A.Cosmic algorithm. crblaster is written in C using the industry
standard Message Passing Interface (MPI) library. Processing a single 800×800 HST WFPC2 image takes 1.87
seconds using 4 processes on an Apple Xserve with two dual-core 3.0-GHz Intel Xeons; the efficiency of the
program running with the 4 processors is 82%. The code can be used as a software framework for easy development
of parallel-processing image-anlaysis programs using embarrassing parallel algorithms; the biggest required
modification is the replacement of the core image processing function with an alternative image-analysis function
based on a single-processor algorithm. I describe the design, implementation and performance of the program.
Planning is underway for a possible post-cryogenic mission with the Spitzer Space Telescope. Only Channels 1
and 2 (3.6 and 4.5 μm) of the Infrared Array Camera (IRAC) will be operational; they will have unmatched
sensitivity from 3 to 5 microns until the James Webb Space Telescope is launched. At SPIE Orlando, Mighell
described his NASA-funded MATPHOT algorithm for precision stellar photometry and astrometry and presented
MATPHOT-based simulations that suggested Channel 1 stellar photometry may be significantly improved by
modeling the nonuniform RQE within each pixel, which, when not taken into account in aperture photometry,
causes the derived flux to vary according to where the centroid falls within a single pixel (the pixel-phase
effect). We analyze archival observations of calibration stars and compare the precision of stellar aperture
photometry, with the recommended 1-dimensional and a new 2-dimensional pixel-phase aperture-flux correction,
and MATPHOT-based PSF-fitting photometry which accounts for the observed loss of stellar flux due to the
nonuniform intrapixel quantum efficiency. We show how the precision of aperture photometry of bright isolated
stars corrected with the new 2-dimensional aperture-flux correction function can yield photometry that is almost
as precise as that produced by PSF-fitting procedures. This timely research effort is intended to enhance the
science return not only of observations already in Spitzer data archive but also those that would be made during
the Spitzer Warm Mission.
The performance of an adaptive optics system is typically given in terms of the Strehl ratio of a point spread function (PSF) measured in the focal plane of the system. The Strehl ratio measures the normalized peak intensity of the PSF compared to that of an ideal PSF, i.e. aberration-free, through the system. One advantage of this metric is that it has been shown to be proportional to the rms wavefront error via the Marechel approximation. Thus, Strehl ratio measurements are used to determine the performance of the system. Measurement of the Strehl ratio is frequently problematic in the presence of noise as can be the peak determination for critically sampled data. We have looked at alternative metrics, in particular the S1 sharpness metric. This metric measures the compactness of the PSF by the normalized sum of the squared image intensity and therefore relates to the intensity variance of the image. Using simulated AO PSFs, we show that there is a unique relationship between S1 and the Strehl ratio and we can therefore relate it back to the rms wavefront error.
We present astronomical results from K-band adaptive optics (AO) observations of the wide binary system σ Corona Borealis with the Lick Observatory natural guide star adaptive optics system on 2004 August 27-29. Seeing conditions were excellent and the AO compensation was very good, with Strehl ratios reaching 50% at times. The stellar images were reduced using three different analysis techniques: (1) Parametric Blind Deconvolution, (2) Multi-Frame Blind Deconvolution, and (3) the MATPHOT stellar photometry code. The relative photometric and astrometric precision achievable with these three analysis methods are compared. Future directions that this research can go towards achieving the goal of routinely obtaining precise and accurate photometry and astrometry based on near-infrared AO observations are described.
Innovative image analysis software has the potential to act as a technology driver for advancing the state-of-the-
art in the design of space telescopes and space-based instrumentation. Total mission costs can sometimes
be significantly reduced by using innovative compact optical designs that create ugly Point Spread Functions.
Most traditional astronomical image analysis techniques, like precision stellar photometry and astrometry, were
developed for the analysis of ground-based image data and many photometric reduction codes cleverly take full
advantage of the blurring caused by the Earth's atmosphere. Image data from space-based cameras, however,
is typically characterized by having significant amounts of power at high spatial frequencies. Mission designers
have a penchant to approve of optical designs that are undersampled. Although excellent justifications can often
be made for using complex optical designs that have ugly Point Spread Functions (e.g., reduced total mission
cost) or for using detectors that are too big at a given wavelength (e.g., giving a wider field-of-view), the analysis
of resultant image data from these designs is frequently problematical. Reliance upon traditional ground-based
image analysis codes may preclude the use of innovative space-based optical designs if such designs are rejected
during the design review process for the very practical reason that there is no proven way to accurately analyze
the resultant image data. I discuss ongoing research efforts to develop new image analysis algorithms specifically
for space-based cameras that may help NASA and ESA to enhance the scientific returns from future astrophysical
missions while possibly lowering total mission costs.
The MATPHOT algorithm for digital Point Spread Function (PSF) CCD stellar photometry is described. A theoretical photometric and astrometric performance model is presented for PSF-fitting stellar photometry. MATPHOT uses a digital representation of the sampled PSF consisting of a numerical table (e.g., a matrix or a FITS image) instead of an analytical function. MATPHOT achieves accurate stellar photometry with under-sampled CCD observations with super-sampled PSFs. MATPHOT currently locates a PSF within the observational model using a 21-pixel-wide damped sinc interpolation function. Position partial derivatives of the observational model are determined using numerical differentiation techniques. Results of MATPHOT-based design studies of the optical performance of the Next Generation Space Telescope are presented; observations of bright stars analyzed with the MATPHOT algorithm can yield millimag photometric errors with millipixel relative astrometric errors -- or better -- if observed with a perfect detector. Plans for the future development of a parallel-processing version of the MATPHOT algorithm using Beowulf clusters are described. All of the C source code and documentation for MATPHOT is freely available as part of the MXTOOLS package for IRAF (http://www.noao.edu/staff/mighell/mxtools). This work is supported by a grant from NASA's Office of Space Science.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.