 |
|
|
|
|
|
| Computational Science Projects |
|
| Viewpoints: High Performance Visualization for Large, High-dimensional Space Science, Earth Science, and Engineering Data. (PI: Levit, Creon,
NASA Ames Research Center
) | |
| Analysis and visualization of extremely large and complex data sets is one of the most significant challenges facing NASA
investigators. Current NASA missions, instruments, and simulations produce so much data of such high dimensionality that they
outstrip the capabilities of traditional visualization and analysis software. Hyperspectral images from satellites, multivariate data of
high dimensionality from sky surveys, time-varying three-dimensional flows from supercomputer simulations, and complex
inter-related time series from vehicle telemetry are but a few examples of these data. We have developed a prototype high performance,
cross platform, data visualization application called viewpoints for graphical exploratory data analysis of large, high-dimensional
NASA science and engineering data sets. The prototype is already in use by approximately 20 NASA scientists and engineers in support
of a broad range of research, development and analysis activities. We propose to leverage this operational experience and the attendant
input from our existing user community to add essential capabilities, transform the prototype into a production system, and deploy it
widely for use by NASA investigators and the general scientific community.
| |
| Presence, Personalization and Persistence: A New Model for Doing Science in a Collaborative Archive Environment (PI: McGlynn, Thomas,
NASA Goddard Space Flight Center
) | |
| We propose to explore a fundamentally new way of working with data archives using techniques and technologies that have been
popularized in on-line virtual realities and massively parallel on-line games. Current NASA data archives use the same fundamental
approach. This can limit the role of the archive in supporting science and limit the science that users contemplate. Using the approach
we suggest below, archives can directly support collaborative research, provide summary views of the data holdings and bridge the gap
between archival access and data analysis. In this research project we shall investigate whether the promise of these approaches can be
realized by building a prototype environment that provides new capabilities within the data archive of NASA's High Energy
Astrophysics Science Archive Research Center. However what we learn should be widely applicable to NASA data centers in all science
domains.
| |
| Subdueing RHSEG: The Marriage of Graph Based Knowledge Discovery (Subdue) with Image Segmentation Hierarchies (from RHSEG) for Data Analysis, Data Mining and Knowledge Discovery (PI: Tilton, James,
NASA Goddard Space Flight Center
) | |
| It is proposed to design and implement the integration of the Subdue graph based knowledge discovery system, developed at the
University of Texas Arlington and Washington State University, with image segmentation hierarchies produced by the RHSEG
software, developed at NASA GSFC, and perform pilot demonstration studies of data analysis, mining and knowledge discovery on
NASA data. Subdue represents a method for discovering substructures in structural databases. Subdue is devised for general-purpose
automated discovery, concept learning, and hierarchical clustering, with or without domain knowledge. Hence, the method can be
applied to many structural domains. As an unsupervised algorithm, Subdue searches for a substructure, or subgraph of the input
graph, that best compresses the input graph. As a supervised learning algorithm, Subdue can be provided with known instances of
target concepts and can use this information to learn a graph concept. The graph concept can be used on new data to determine its
class value. Substructure discovery using Subdue has yielded expert-evaluated significant results in domains including terrorist
activity, predictive toxicology, network intrusion detection, earthquake analysis, web structure mining, and protein data analysis.
Subdue was developed by Co-I, Diane J. Cook, and her colleague, Lawrence B. Holder. For Subdue to be effective in finding patterns
in imagery data, the data must be abstracted up from the pixel domain. An appropriate abstraction of imagery data is a segmentation
hierarchy: a set of several segmentations of the same image at different levels of detail in which the segmentations at coarser levels of
detail can be produced from simple merges of regions at finer levels of detail. The RHSEG program, a recursive approximation to a
Hierarchical Segmentation approach (HSEG), can produce segmentation hierarchies quickly and effectively for a wide variety of
images. Both RHSEG and HSEG were developed at NASA GSFC by the PI, James C. Tilton. The HSEG/ RHSEG approaches are
based on region growing, and include a provision for merging spatially disjoint regions into aggregate regions. This region aggregation
leads to a compact abstraction of even very complex images with very faithful representation of spatial detail. This proposed work is an
innovative combination of two established information technologies which will significantly improve our ability to extract and/or
discover information from imagery data. The proposed work also includes a number of demonstration projects designed to illustrate
the significant potential this combination of technologies has to increase the productivity of NASA's Science Mission Directorate
research endeavors, fostering collaborations across a wide range of space, Earth and computer science disciplines.
| |
| Directed Exploration of Complex Systems (PI: Burl, Michael,
Jet Propulsion Laboratory
) | |
| We will mature and broaden the applicability of a directed exploration capability designed to maximize scientific return from
large-scale numerical simulators. Numerical simulators provide scientists with a valuable tool for examining massively complex
systems that would be impossible to study otherwise. However, in many simulations, long run-times make a detailed, exhaustive study
of the ``input parameter landscape'' infeasible. The key idea we are advancing uses support vector machine (SVM) classification and
regression techniques along with active learning to cleverly explore input parameter space, as a means of improving the speed and
efficiency with which a set of simulations can provide scientifically useful results. In addition, Markov Chain Monte Carlo (MCMC)
sampling techniques will be introduced to aid in the exploration and visualization of these potentially high-dimensional input
parameter spaces. In proof-of-concept studies to date (including an AISR-funded seed project), we have successfully applied our
approach to two large-scale scientific simulations: (1) a smooth particle hydrodynamic (SPH)/N-body gravitational simulation of
asteroid collisions to narrow plausible initial conditions (impactor size, velocity, etc.) for production of asteroid satellites and for
generation of specific asteroid families (Emma, Karin, Baptistina) and (2) magnetospheric inversions driven by in-situ observational
data from the IMAGE spacecraft. Having demonstrated the basic feasibility and utility of the concept (speedups from 2-fold to 100-fold
while achieving comparable fidelity to exhaustive grid sampling), we seek to mature and broaden the approach so that it can be applied
to a variety of scientific investigations. At the recent Second NASA Data Mining Workshop, numerous scientists voiced concerns that
data mining outputs are rarely linked back to the underlying physical system and processes. A major strength of this proposal is that it
directly links data mining with models of physical systems, which are currently being used in projects at the cutting edge of scientific
research, and aims at increasing knowledge of how the underlying parameter spaces affect observable quantities. Another significant
strength is that we team domain experts (scientists who are expert in particular simulators) with computer scientists who can bring to
bear cutting edge research in computing techniques and technology. This work is intended to be applicable to a broad variety of
physical systems modeled with simulators. Our approach requires no modifications to the internals of a particular simulator; the
method can be deployed simply by writing glue code that enables the simulator to be called by the directed exploration capability and
providing one or more grading scripts that process the raw output of the simulator into the scientific quantities of interest for the
investigation. The primary asteroid collision applications directly address NASA strategic subgoal 3.C.1: ``Progress in learning how the
Sun's family of planets and minor bodies originated and evolved.'' Directed exploration fits well with several of the Applied
Information Systems Program goals, including developing and deploying tools to amplify the productivity of scientific users of
high-end computing resources and to increase science return, including enabling qualitatively new science through information science
and technology. Our collaboration has been successful in the past in bringing low-TRL concepts through AISR (and similar programs)
into deployment in NASA science research and analysis programs.
| |
| NASA's Cosmos (PI: Lang, Kenneth,
Tufts University
) | |
| New discoveries from NASA's investigations of the Cosmos will be shared with
the general public, space scientists, students and teachers through the
writing and publication of two new books, entitled "The Sun From Space, Second
Edition"; and "Astrophysical Formulae, Fourth Edition: An On Line Reference",
building upon a proven track record with NASA and AISRP. This work will
bring scientific credibility, human interest, and visual excitement to NASA
accomplishments. In a logical extension of proven abilities, the Principal
Investigator will use the past ten years of NASA Space Science results to
update, extend and stengthen two books, including one of them in on line,
electronic form. "The Sun From Space, Second Edition" will demonstrate how
recent Solar Spacecraft have helped solve fundamental solar problems such as
the heating of solar corona, the origin of the Sun's winds, the nature of
solar flares
and coronal mass ejections, and the Space Weather interactions of the Sun
with either the Earth or with unprotected astronauts and spacecraft in
outer space or on the Moon or Mars.
"Astrophysical Formulae, Fourth Edition: An On Line Reference" wiil be
published in an updated electronic version, with its own search engine.
It will demonstrate how NASA spacecraft have contributed to studies of
black holes, cosmic evolution, the expanding Universe, dark energy, dark
matter, gamma ray bursts, planets around other stars, pulsars, quasars, and
the three-degree cosmic microwave background. Springer Verlag has agreed to
publish both books in late 2008, or early 2009, providing sufficient time to
assimilate new findings from recently launched spacecraft.
| |
| A Distributed Knowledge Extraction Framework Based on Semantic Web Services (PI: Graves, Sara,
UA Huntsville
) | |
| Project Summary
Key information technology advances in recent years include the emergence of distributed computing architectures based on web services; knowledge engineering efforts as evidenced by the development of science domain ontologies in the Semantic Web; and growing interest in scientific data mining as a means for automated knowledge extraction from the ever-increasing volumes of science observations and model data available. To facilitate exploitation of these promising techniques by the NASA science community, the University of Alabama in Huntsville (UAH) leads a multi-disciplinary, collaborative research team in proposing to prototype an extensible framework to integrate distributed resources including science data, related services, and the ontologies that describe them.
This project brings together a strong research team with a unique combination of skills and experience. Dr. Sara Graves' research center at UAH brings to the project ongoing data mining and semantics research to provide knowledge extraction, manipulation and analysis web services. The knowledge engineering aspects of this project wll capitalize on the long involvement in formal ontology design and development by Dr. Deborah McGuinness at Stanford University and McGuinness Associates. Dr. Peter Fox at the National Center for Atmospheric Research will provide science expertise to develop science domain ontologies and guide design of the prototype.
The result of this research will be a prototype semantic knowledge integration framework, SKIF, comprising a toolkit of data mining and knowledge extraction web services designed specifically for NASA data, and a series of linked ontologies describing both the data mining, manipulation and analysis services as well as the science problem domain. A web-based user interface will use the ontologies to allow users to discover and explore available data and services, compose workflows of data access, data mining and related services appropriate for their tasks, and invoke them to perform the desired analysis.
This proof of concept will serve two important purposes. First, SKIF will be a useful tool to assist researchers in creating data analysis workflows to address targeted Earth-Sun System science problems. Perhaps more importantly, associating semantic information with these knowledge extraction, manipulation and analysis services will not only address immediate usability concerns, it will also position them for integration with many other science data services in the emerging Semantic Web Services context.
| |
| EPISODE - Software for Trajectory Generation and Nonlinear Continuous Control in the Presence of Uncertainty (PI: Jewell, Jeffrey,
NASA JPL
) | |
| We will develop the software named EPISODE (Evaluation of the Posterior for the Inference of Solutions of Ordinary Differential Equations) for an implementation of a probabilistic (Bayesian) approach for reasoning about dynamical systems in the presence of uncertainty with application to intelligent mission simulation, trajectory generation, and nonlinear continuous trajectory control, in order to increase life cycle effectiveness and efficiency of the Science Mission Directorate research endeavors, in particular: 1) to reduce mission development time, risk, and cost through advanced simulation and design capabilities, and 2) to increase mission duration and reliability through autonomous operations and control.
The work proposed here provides algorithmic advances for the use of dynamical systems theory and optimal control, an approach to trajectory design that has already achieved a dramatic reduction in mission planning time for the Genesis mission, in solving for the trajectory design in less than ONE day as opposed to 8-12 weeks!! The work proposed here will further the automation and accuracy with which these trajectories can be discovered and optimal control laws synthesized for more complex missions for longer duration, with multiple objectives, over the full spectrum of fuel expense penalty (from longer time low thrust to large control inputs). Although this proposal specifically targets the mission design problem, we note that the EPISODE framework is completely general and may be applied to any engineering problem described by differential equations.
The proposed work directly addresses several of NASA�s strategic objectives as outlined in Table 1 of the Summary of Solicitation of this NRA, including 1) Undertake robotic and human lunar exploration to further science and to develop and test new approaches, technologies, and systems to enable and support sustained human and robotic exploration of Mars and more distant destinations, 2) Conduct robotic exploration across the Solar System for scientific purposes and to support human exploration in particular, explore Jupiter�s moons, asteroids, and other bodies to search for evidence of life, to understand the history of the Solar System, and to search for resources, 3) Develop and demonstrate power generation, propulsion, life support, and other key capabilities required to support more distant, more capable, and/or longer duration human and robotic exploration of Mars and other destinations.
| |
| Enhancing the Pyramid Parallel Unstructured Adaptive Mesh Refinement Library (PI: Norton, Charles,
NASA JPL
) | |
| The objective of this proposal is to introduce significant and relevant enhancements to the PYRAMID parallel unstructured adaptive mesh refinement library. Developed under the ESTO Computational Technologies Program, this freely available tool is currently used for large-scale mesh generation and solution adaptive mesh refinement for Earth and space science applications. More than 100 downloads of the current version of the tool exist, but there is growing interest in the research, commercial and private sector for additional functionallity to be incorporated into the tool. This proposal will add such features including improved memory usage, support for commercial mesh generator formats such as NASTRAN, IDEAS, Solidworks, and Unigraphics, full implementation of the C interface, a web tutorial for on-line training, additional improvements to mesh coarsening routines, retrofitting of 3D features to the 2D version, and parallel I/O support. AISR has supported similar activities for other AMR too development in the past so this proposal is relevant to the goals of the program.
The work will be accomplished by redesigning aspects of the central data structure to more effectively handle memory usage while providing support for unstructured adaptive mesh coarsening (based on techniques of R. Lohner) with quality control. (Quality control prevents creation of elements with poor geometry as the mesh structure changes.) Translators will be generated to support the commercial mesh generation formats, routines will be added to the C-interface library since PYRAMID is written in Fortran 95, and large-scale I/O will be managed using either HDF-5 and/or aspects of MPI-IO. No new numerical techniques need to be added to the library.
This work is significant to the objectives of the solicitation as it is cross-cutting and bridges NASA SMD interests in Earth and Space science. One of the major challenges facing these communities involves modeling and interpretation of data from observational measurements. As missions are planned, and as results are obtained, achieving analysis goals will depend on coupling physics-based solvers with advanced mesh/grid infrastructure tools for parallel computers, such as Project Columbia and others. This ranges from understand Earthquake science to modeling of space physics phenomena. We have, for example, used this tool as part of the Quakesim project that models surface deformation for active tectonics and other Solid Earth processes. Such a capability is directly relevant to the development of an InSAR analysis system as large-scale adaptive grids, when coupled with geophysics solvers, can assimilate InSAR measurements to model deformation processes. It has also been used for magnetosphere flow modeling as part of Earth-Sun system simulation models as well as problems in structural mechanics engineering.
This proposal responds to the following NASA strategic objective: Conduct a program of research and technology development to advance Earth observation from space, improve scientific understanding, and demonstrate new technologies with the potential to improve future operational systems.
| |
| Viewpoints: Hyperwall for the masses (PI: Levit, Creon,
NASA Ames
) | |
| Current NASA missions, instruments, and simulations produce so much data of such high dimensionality that they outstrip the capabilities of traditional visualization and analysis software. Hyperspectral images from satellites, multivariate data of high dimensionality from sky surveys, time-varying three-dimensional flows from supercomputer simulations, and complex inter-related time series from vehicle telemetry are but a few examples of these data.
The hyperwall, developed by the Computational Sciences Division at the NASA Ames Research Center, is a high-performance visualization cluster that was specifically designed to address this problem: Exploring, visualizing, and analyzing large, complex, multidimensional data. The hyperwall's capability represents a qualitative improvement over traditional approaches. But until now, this capability has required specialized hardware.
We propose to re-implement the technology of the hyperwall for general use: to create the necessary software, user interfaces and application programmer interfaces (APIs) so that hyperwall-like functionality is made available on an off-the-shelf high-end desktop workstation. We will apply this to a set of astrophysics and space science data analysis problems, and we will make the system available to the NASA and general scientific communities as open source.
| |
| Robust Grid Computing using Peer-to-Peer Services (PI: Sussman, Alan,
U of Maryland
) | |
| We propose to build a scalable infrastructure for executing grid
applications on a distributed set of resources. Such infrastructure must
be decentralized, robust, highly available, and secure, while efficiently
mapping applications to available resources throughout the system.
Fortunately, these are precisely the characteristics promised by new
techniques and approaches in peer-to-peer (P2P) systems research.
Our system is composed of a relatively loosely-coupled set of distributed,
cooperating users (peers). All peers contribute resources to an ad-hoc
resource pool, and all peers can submit jobs that are executed using
available resources. The overall system, from the point of view of a user,
can be thought of as a combination of a centralized, Condor-like Grid
system for submitting and running arbitrary jobs, and a system such as
SETI@home for farming out jobs from a server to be run on a (potentially
very large) collection of machines in a completely distributed environment.
Such a P2P grid system will have clear advantages over the current state of
the art platforms for so-called ``desktop grid computing'', which are based
on a client-server architecture. In these systems, a trusted server
supplies jobs to a set of client machines distributed across the Internet.
The server controls job placement, and also
collects all results. The server must, therefore, maintain state
for all jobs in the system, and the entire resource pool is rendered
useless if the server fails. The server can also become a performance bottleneck.
The proposal addresses the crucial problem of job placement in a
completely decentralized manner. We describe a
distributed algorithm for submitting jobs and efficiently matching them to
available resources. We also address issues related to security and the
forming of separate ad-hoc networks for different communities of users. We
use P2P techniques for both load-balance and for resilience; from our
preliminary analysis and simulations, we expect our scheme to scale with system size
and to be robust against peer failures and departures.
The proposed work is a collaborative effort between computer scientists and
astronomers. We will design, implement, and validate the
system using a set of problems in computational astronomy, including ones
that are directly relevant to the Deep Impact mission. Measuring
performance and observing the behavior of the algorithms in a
realistic environment will validate the usefulness of basic peer-to-peer
services in the grid context.
Applications that are suited for the system have both large
computational requirements and low I/O requirements. We have identified
several problems with these characteristics from the computational
astronomy domain, including several relevant to data analysis and
theoretical modeling for the Deep Impact Mission. For example, the detailed chemical network modeling
required to match the observations involves a large parameter space search
that is ideally suited for the proposed P2P scheme. Currently there are
insufficient computational resources available to the Deep Impact team to
carry out this search in a practical fashion using a dedicated brute-force
approach. The proposed system will enable employing unused cycles on any
machines made available by collaborators and other willing participants,
without significant administrative burden or security concerns.
| |
| Mission Extension Using Sensitive Trajectories and Autonomous Control (contd) (PI: Belbruno, Edward,
Princeton U
) | |
| This proposal represents a continuation of a current funded proposal from AISRP. The goal is to show how dynamically sensitive trajectories about a planetary body, eg the Moon, can substantially reduce the fuel required to stay in orbit for extended periods of time. The trajectories move in a region about the Moon which supports chaotic motion where the spacecraft is between capture and escape about the Moon. It has been demonstrated by this proposer that a tiny delta V will stabilize the motion and where the spacecraft can orbit the Moon for several months with negligible delta V. Two spacecraft used this proposers trajectory design to demonstrate this motion about the Moon - Japan's Hiten in 1991 and ESA's SMART-1 in 2004. The current work by this proposer has made significant progress in his current research supported by AISRP showing how small the delta V can be to maintain the motion about the Moon with a simple trajectory control. An interesting discovery was recently made to show that the inclination of a spacecraft moving in this way can be changed from 0 to 180 degrees for nearly zero delta V. This 'free' inclination change has enormous potential to be further studied since it takes a lot of delta V and hence fuel to change inclination by conventional means. Also, it is being proposed to fully design an optimal orbit control algorithm that can keep the spacecraft in motion about the Moon for very long periods of time for almost no delta V. The current effort can only touch on this. After an algorithm is developed, it can be programmed into a program to give the spacecraft an autonomous capability to maintain its motion about the Moon.
This capability can give extended mission duration which will enable the spacecraft to mine substantially more data than previously for the same amount of fuel. In fact, less fuel could be stored on the spacecraft thereby reducing its size and mass.
There are two approaches used to see how to substantially reduce the delta V for control: The first is to perform a tiny delta V when the spacecraft is captured at the Moon in the chaotic region to stabilize it with minimal fuel. Then the trajectory is followed to see how it becomes unstable, and where it is best to perform another delta V. This process is continued. The second approach is to perform an analysis of the dynamic nature of the instability of the motion from a mathematics perspective to better understand it. This approach uncovers surfaces where the spacecraft will go to and follow when it becomes unstable (called invariant hyperbolic manifolds). These two approaches taken together enable algorithm definition.
The objectives of this proposal are significant since they
1. Enable autonomous operations of a spaceraft in motion about a planet
2. Data mining is substantially enhanced due to extended motion about the planet for much longer periods of time than using standard orbital mtion
3. The discovery of the ability to perform inclination changes for negligible delta V has enormous implications since by standard methods, inclination changes require a lot of fuel
4. Autonomous motion about the Moon for little delta V and to perform inclination changes for little delta V has important applications for returning to the Moon by substantially reducing the cost of orbital operations.
| |
| An Integrated Software Environment to Design Polymorphic Fault Tolerant Processors for Command and Control Functions on RadHard FPGAs (PI: Dasu, Aravind,
Utah State U
) | |
| Commercial FPGA (field programmable gate arrays) have started making a powerful impact on space science missions, owing to the success of their deployment on the Mars rovers. They offer highly adaptable/reconfigurable electronic fabrics which if efficiently utilized can outperform the traditional monolithic microprocessors for on-board processing. The thrust in deploying autonomous command and control (C&C) software on-board, driven by sophisticated scheduling and navigation algorithms, has created a need for highly adaptable and computationally powerful on-board processing systems. It is expected that for tasks such as orbital rendezvous and atmospheric entry, the processing rates and states of underlying kalman filters rapidly increase as compared to more mellow times of the missions. This requires the on-board computing systems to quickly change their capabilities at run time. FPGAs offer the ability to host such adaptable processing systems, but the onus is on the user to design such systems. In order to create systems that can be adapted quickly to (i) a new mission, or (ii) a new processing speed for a different task on the same mission, or (iii) a variant of a computation algorithm due to change in health of the spacecraft (in a fault tolerant manner), the circuits need to be (a) polymorphic and (b) tailored well to the data/control flow nature of the application. In this project, we propose to capture the data/control flow nature of C&C applications by first compiling their C/C++ programs into machine independent Intermediate Representation (IR) forms. Then, through a set of feature extraction compiler passes, we will extract their core data/control flow patterns. These patterns will act as templates to soft macros that will form the core of the data-paths and state machines on an FPGA. We will grow polymorphic architectures in the Viva design language, on top of these soft-macros (which act as seed elements). These architecture modules will be capable of order tensor, information rate and data type polymorphism. To provide strong fault tolerance capabilities in these modules, we will embed triple modular redundancy and built in self test features at the seed element level. To provide a higher degree of freedom in programmability of the overlaying architecture, we will partition the application�s architecture into a group of soft microprocessors on the FPGA. This will permit us to partition and compile the application�s C/C++ programs onto the custom instruction sets of the soft microprocessors. In this project we will integrate state of the art technologies and tools for code generation of mission/planning & state estimator algorithms, FPGA synthesis tools and compilers, into a single design environment. Through this system we will be able to provide a cost effective and time saving design environment for NASA engineers and scientists for future space missions.
| |
| Development of Active Learning Capability for Numerical Simulations (PI: Merline, William J.,
Southwest Research Institute
) | |
| We propose to develop efficient machine learning techniques that will enable vast improvements in large-scale numerical simulations in the space sciences. Simulations play a fundamental role in studies by scientists and engineers across NASA, DoE, NOAA, FAA, industry, and academia. In many cases, simulations provide the means to examine processes that could otherwise be infeasible or impossible to study. Here, we team domain experts (scientists expert in particular simulations) with computer scientists who bring to bear cutting-edge research in computing techniques and technology. We will employ novel active-learning techniques, for which we have already shown proof-of-concept. In a just-completed NASA Intelligent Systems grant, we demonstrated that active learning can make running suites of numerical simulations dramatically more efficient. Here, we seek to mature that technology to the extent that it can be presented for funding to a NASA OSS research program. To do this, we choose as our primary application the simulation of asteroid collisions, resulting in creation of asteroid satellites. This application will demonstrate the depth of our methodology, and when complete, will allow a substantial increase in the speed of learning from simulators. We will also demonstrate the wide-ranging potential applicability of our active learning research by development under two additional simulation areas --- understanding of data on Earth's magnetosphere, and understanding of the structure of the atmosphere of Titan, a moon of Saturn. Our three applications are very different from one another, and thus will demonstrate the power of the technique, in terms of flexibility and enhancement of science understanding.
| |
| HYDRA: A New Paradigm for Astrophysical Modeling, Simulation, and Analysis (PI: Wise, Michael,
Massachusetts Institute of Technology
) | |
| Regardless of the source or mission, the analysis of astrophysical data invariably relies on the interplay between predictive physical models, a detailed understanding of the observing instruments, and the discriminating comparison of predictions and actual observations. With HYDRA, we are developing a new platform which provides scientists a more flexible and extensible way of constructing models for astrophysical sources including geometric information, physical emission and absorption mechanisms, transport processes, and projection effects. We will also provide an interface through which users may link existing astrophysical models to HYDRA. Similarly, modules describing the performance of the instrumentation, be it ground-based telescope or orbiting satellite, can be defined in a mission-independent way to allow realizations of these models in the form of simulated observations. By combining source and instrumentation models, HYDRA can serve as a tool for both observational planning and instrument design and calibration. Simultaneously, HYDRA provides an analysis environment allowing source models to be compared to existing observations and iteratively adjusted. As part of its design, the HYDRA system will also include advanced visualization capabilities providing users with additional diagnostic ability as well as a potentially powerful educational tool.
| |
| Mission Extension Using Sensitive Trajectories and Autonomous Control (PI: Belbruno, Edward,
Innovative Orbital Design, Inc.
) | |
| The main objective of this proposal is to describe how to design new types of mission trajectory designs based upon the use of dynamically sensitive(chaotic) trajectories together with autonomous spacecraft control. This combination would produce new designs which would substantially increase mission duration and require little fuel in the process. Once recently discovered type of motion called ''quasi-stationary motion'' is a promising new development where a spacecraft remains nearly fixed relative to the earth in a large region of space, distinct from the Lagrange points-offering a number of new scientific applications. Such a motion is very sensitive and unstable, and autonomous control would likely be necessary in practice. This combination would increase mission duration substantially, and increase both data collection and scientific return. Another application is to a special class of lunar transfers already demonstrated in 1991 which require little fuel and are also dynamically sensitive. Autonomous control of such transfers would enhance the capabilities of the transfers with regards to mission duration, reliability, and scientific return. They would also increase the mass that could be brought to the Moon by a factor of two at a low cost. The methods that will be used in this proposal are to numerically simulate the given motions and the algorithms required for autonomous control. The algorithms will be designed to keep the spacecraft on their sensitive motions by using little fuel. As a result these algorithms will increase mission durations substantially. The proposed work would enhance space science productivity supporting the goals of OSS. The development of special algorithms to maintain autonomous control of dynamically sensitive trajectories would provide a novel way to increase the usefulness of missions and open new possibilities for mission designs. This supports the goals of NASA for the exploration of the solar system, and it also supports the new directive for NASA to return to the Moon.
| |
| Autonomic Computing via Dynamic Self-Repair of Hardware Faults (PI: Sorin, Daniel,
Duke University
) | |
| The goal of this research project is to develop computer systems that use dynamic self-repair to enable autonomic operation in the presence of permanent hardware faults. Autonomic operation is crucial for reliability when mission-critical computer systems are deployed for a long time with little or no opportunity for human repair. Existing solutions for tolerating general classes of hard faults, such as triple modular redundancy, are very expensive-in terms of power and hardware cost. In this research project, we seek to develop lightweight hardware techniques for self-repair that can achieve high reliability and robust performance without consuming vast amounts of hardware or power. Our first step in this direction, called Self-Repairing Array Structures (SRAS), masks hard faults in microprocessor array structures. Our experimental evaluation of SRAS shows that it achieves high reliability with far less power consumption than processor-level redundancy. Our current research is exploring Hierarchical Modular Redundancy (HMR), a self-repair scheme for tolerating faults in more general structures than just arrays. Ongoing and future research will further develop SRAS and especially HMR, including efficient hardware implementations and thorough experimental evaluations. We will also explore a range of computing systems other than just microprocessors, including embedded processors and network processors. As part of this project, we will develop significant simulation infrastructure for evaluating these research ideas and for modeling the effects of emerging fault models.
| |
| The Virtual Cosmos: A Public Portal to the National Virtual Observatory (PI: Craig, Nahide,
University of California Berkeley
) | |
| The National Virtual Observatory (NVO) holds tremendous potential for Education and Public Outreach (EPO) opportunities. The possibilities for EPO with the NVO, which promises to make widely available the great majority of the worldís astronomical data, are too numerous for any one EPO effort to ever hope to develop. Therefore, it is critical that the NVO EPO program develop an infrastructure and tools flexible enough that any EPO program can make use of it. We propose to assemble the infrastructure and tools for NVO EPO and to construct the first portal that makes use of NVO resources for non-scientists. This portal, called The Virtual Cosmos, will be designed to provide the user with an integrated model of the Universe and the ability to navigate to regions of interest linked to live databases. The portal will stand as an example of many possible EPO resources that can make use of the NVO EPO infrastructure. Such a robust end-to-end prototype of an EPO interface is in the true spirit of NVO and goes beyond the traditional online portals to the various datasets. The Virtual Cosmos will inform, excite, and educate the public about space science and astronomy, and serve as a catalyst for scientific and technological literacy.
| |
| Low Power Architectures for Real-Time Hyperspectral Image Processing (PI: Bose, Tamal,
Utah State University
) | |
| The goal of this project is to design and implement re-configurable low power hardware tools for real-time processing of hyperspectral images. The main idea is to use only quantized power-of-2 algorithms in which all multiplications are reduced to simple shifting operations. This will guarantee that the hardware will use the minimum power, provided other design constraints are met. In a previous AISR project, we have designed and developed several multiplier-free algorithms. One important and large component which was not part of that project is ''Spectral Unmixing.'' In hyperspectral images, data are often modeled as linear combinations of radiance spectra. Therefore, the spectral data needs to be ''unmixed'' before usage. In this project, we will develop multiplier-free algorithms for spectral unmixing. We will then design hardware modules for all our DSP algorithms. We will design low-power VLSI architectural modules for several different DSP algorithms including: (a) Fixed digital filters, (b) Adaptive filtering algorithms, (c) Image restoration algorithms, (d) Spectral unmixing algorithms, (e) Adaptive Pulse Code Modulation (ADPCM), (f) Transform coders, and (g) Entropy coders. All of these modules will be designed to be multiplier-free and low-power. These features also translate to low cost and smaller chips. We will combine these modules to build an entire low power system for hyperspectral image restoration and compression. Utah State University/Space Dynamics Laboratory (USU/SDL) has developed a low power computer system for command and control, attitude determination, and telemetry for small spacecrafts. The system has been developed for the 15-kilogram class Ionospheric Observation Nanosatellite Formation (ION-F) satellites. The low-power hyperspectral image processor will be integrated with this system to develop a complete spacecraft system. Since all hardware design is typically done in VHDL or Verilog, it will not be difficult to perform this integration. In fact, we will also integrate the electronics for science instruments with our system. This includes instrumentation for image detection, electron density, electric field, magnetic field and neutral particle motion and composition. The ultimate goal is to build a complete ìspacecraft system on a chip.î
| |
| Estimating Missing Data in Sensor Network Databases Using Data Mining to Support Space Data Analysis (PI: Gruenwald, Le,
University of Oklahoma
) | |
| Recent advances in Micro Electro Mechanical Systems (MEMS) based sensor technology, low-power analog and digital electronics, and low-power Radio Frequency (RF) design have made possible the development of relatively inexpensive and low-power wireless micro sensors that can be integrated in a network. The purpose of such a network is to monitor, combine, analyze and probably respond to the data collected by hundreds (or even thousands) sensors distributed in the physical world in a timely manner. This network can be used to support space data collection and analysis. For example, to facilitate solar system exploration missions, mobile sensors mounted on robots as well as hundreds of static micro sensors can be placed on MARS to collect its data and to send the collected data to a base station residing on MARS for real-time data analysis. The base station can then use the analysis results in real-time to determine actions that the robots should take next. However, in a wireless sensor network, a significant amount of sensor readings sent from the sensors to the data processing point(s) (servers) may be lost or corrupted. In this research we propose a power-aware technique that uses association rules mining to handle such a problem. In this technique, to save battery power on sensors and to meet real-time requirements for data analysis, instead of requesting the sensor nodes (MS), the readings of which are missing, to resend their last readings, an estimation of the missing value(s) is performed by using the values available at the sensors relating to the MS through association rule mining. Temporal data mining using data clustering is also employed to improve data estimation. This research derives solutions for both centralized and distributed wireless sensor networks where transmissions can be single hops or multiple hops, and sensors/servers can be static or mobile. It then conducts performance evaluations using NASA sensor data to compare its proposed technique with existing statistical approaches.
| |
| PARAMESH: A parallel, adaptive, grid tool for the Space Sciences. (PI: Olson, Kevin,
Drexel University
) | |
| PARAMESH is a portable high performance software toolkit that enables parallel, adaptive-mesh, computer simulations. It was originally conceived to support hydrodynamic and MHD applications of the Solar atmosphere. It has been so successful that it is now being used in a much broader range of space science applications. These applications include General Relativistic models of colliding compact objects in preparation for the LISA mission, space weather models of the sun, inner heliosphere and magnetosphere, and models of radiation hydrodynamics in novae, supernovae, and gamma-ray bursts. These different applications stress the PARAMESH package in different ways and will all benefit from a PARAMESH redesign which absorbs their specific needs. We propose to extend the functionality of PARAMESH and rationalize its internal structure to best achieve this. Extensions will include support for radiation transfer, particle techniques, and improvements specific to General Relativity. Among these extensions are support for multigrid algorithms, elliptic solvers, curvilinear coordinates and unstructured grids, divergence of B control, and higher order interpolation proceedures. An additional major goal in this effort will be to configure the development process to include outside developers from these application groups and beyond, by adopting an effective open source development model. This will enable PARAMESH to more easily grow in future years in ways defined by its user community.
| |
| Automated Data Analysis with Knowledge Ontologies (PI: Shaya, Edward J.,
University of Maryland College Park
) | |
| A major advance in the direction of automated scientific data analysis becomes clear by recognizing that the graphical representation of procedures in Visual Programming (VP) is a directed graph that can be represented by standard ontology such as the W3Cís OWL. A merger of these two technologies would brings together set theory, first order logic reasoning, planning and scheduling, and XML technologies of ontologies, with the advanced modeling, on-the-fly programming, and interchangeable software components of VP. Additionally, both ontology and VP interface well with the distributed execution and storage technologies of WebServices and the Grid. Therefore, such an integrated system is not limited by the local hardware capabilities. For this AISR we propose to make use of mostly existing open source software to create an extremely powerful space science data analysis tool that implements expert knowledge about space science and learns and grows by experiences shared by multiple users across the web. We will replace the graphical front end typically used in a visual programming (VP) software (SCIRun from http://software.sci.utah.edu/) with an ontological front-end (ProtÈgÈ http://protege.stanford.edu) that provide knowledge of the relationships between transformations/functions and data classes. We will add to the existing set of functions available to SCIRun a selected set of functions needed in space physics and astronomy. We will provide read and write for standard data formats (CDF, FITS and VOTable). We will provide query tools for extracting data from the Virtual Observatories and individual data centers. We will add the capability of transparently shifting computations to the Grid when the load is too heavy for local resources. Finally, we will provide a web based repository where working procedures and relationships are stored as OWL/RDF files. This is a novel computational method that will increase productivity of researchers by making it irrelevant to create detailed programs/scripts or even flow diagrams for each step in a scientific investigation. Scientific return and public outreach will be enhanced by simplifying the path to creating visualization of complex data. We will be providing a key step in enabling recent IT developments in semantic Web and advance problem solving for scientific output of OSS missions. In addition, this package will make possible increased data return through onboard science processing and reduction which also leads to huge compression factors.
| |
| Advanced Techniques for High-Performance Computer Simulations of Rarefied Neutral Gas and Plasma Flo (PI: Combi, Michael,
University of Michigan Ann Arbor
) | |
| As computers increase in speed and memory, particle kinetic simulations will continue to expand in their capability to address physical problems in which non-equilibrium microphysical processes are of central importance and for which fluid-based approaches are inadequate. With previous support from the Applied Information Systems Research Program we have successfully developed a highly parallelized kinetic model based on Direct Simulation Monte Carlo (DSMC) for application to a variety of space science rarefied gas applications. The features and capabilities of the model are as follows: ï Collisional kinetic neutral and ionized gas physics with rotational/vibrational degrees of freedom and photodissociation on 1D, 2D, 3D unstructured meshes, ï An integrated Euler hydrodynamics solver incorporating a fully functional hybrid scheme with adaptive boundary feedback between kinetic and fluid domains, and ï Dust particles for modeling dusty-gas flows such as dusty gas cometary atmospheres and volcanic plumes on outer planet satellites. A five-year research program is proposed in which an important set of developments will be studied and implemented using our kinetic ion/neutral particle simulation model for space science as a test bed. Just as our past and proposed advances are built on the published work of others, the dissemination of innovative techniques to be developed are intended to contribute to the general advancement of particle kinetic simulation work. Our specific goals are as follows: ï Extension of our already-developed finite-element based Poisson solver for self-consistent electric fields to self-consistent solution of electric and magnetic fields in plasma particle simulations. A finite element approach has the potential to permit using unstructured meshes in fully electromagnetic particle plasma simulations. ï A new computational mesh method based on mixed unstructured (tetrahedrons) and structured (hexahedrons) cells. This could lead to a computational speed-up of up to a factor 6. ï A novel noise reduction/optimization scheme--for the first time in DSMC--which is similar to the complex particle kinetic (''blob'') methods now in forefront plasma kinetic particle-in-cell (PIC) research. This could lead to a dramatic decrease in the number of simulation particles required for many applications. ï New physics attributes for the model including fluid electrons and gas-phase chemistry. The addition of more complex physics will greatly extend the capability of the application of particle kinetic methods to even more realistic and practical problems.
| |
| The NASA'S Cosmos Project (PI: Lang, Kenneth,
Tufts University
) | |
| Central Objectives: The excitement and scientific content of NASAís Space Science Missions will be shared with the public. Notable books will be written, and public accessibility and exposure enhanced in an enjoyable NASAíS COSMOS Web Site, located at http://ase.tufts.edu/cosmos/. Previous Accomplishments: Previous Applied Information Systems Research (AISR) Program funding has resulted in the publication of two books, entitled The Cambridge Encyclopedia of the Sun (2001) and The Cambridge Guide to the Solar System (2003), and the completion of the Solar System phase of the NASAís COSMOS Web Site, which now has more than a million hits per year. Other accomplishments include a Discovery Channel show, The Savage Sun, presentations to the National Academy of Sciences and United Nations Workshops, and contributions to the Elsevier, Phillips and Microsoft Encarta Encyclopedias. Proposed Methods/Techniques: The proposed work will update and extend previous work by incorporating new NASA discoveries in the Solar System, and including NASA results for the entire Cosmos outside the Solar System. A new book, entitled Parting the Cosmic Veil, will be written, and new Navigation Tabs for the NASAíS COSMOS Web Site will be prepared, within categories entitled Discovering the Unknown, Brave New Worlds, Motion, Content and Form, Pervasive Violence, Cosmic Evolution, Illusory Emptiness, and Origins and Destinies. Each thematic concept will include an Overview of key concepts, a Tutorial with text and images, Images that present hundreds of visually appealing space shots and line drawings, and Resources with access to all of the relevant Space Science Mission Home Pages. This new material will include topics of immense public interest, such as black holes, cosmic evolution, the expanding Universe, dark energy, dark matter, gamma-ray bursts, planets around other stars, pulsars, quasars, and the three-degree cosmic microwave background. Significance of Proposed Work to ROSS-2003 and NASA interests and programs: This comprehensive and innovative program of education and public outreach describes accomplished and anticipated scientific results of the Office of Space Science (OSS) in all its 2003 science themes and strategic goals, and many of its 2003 research focus areas. The proposed work will bring scientific credibility, historical authority, human interest, and visual excitement to these elements of the OSS program, sharing all this information with a broad cross section of the American public that funded them.
| |
| Block-Adaptive Parallel Implicit Methods for Semirelativistic Multifluid Hall-MHD (PI: Gombosi, Tamas I.,
University of Michigan Ann Arbor
) | |
| Space missions have transformed our understanding of space and astrophysical environments into a more global picture. This wealth of data, however, will be of value only if we derive from these observations, and the subsequent interpretations, a clear understanding of the underlying and governing physical processes. It is here that a unifying multiscale, 3D model plays an essential role. The model must be sufficiently versatile to capture the complexity of the actual physical system. Then, by creating models that are consistent with the observations, the underlying and governing physical processes are revealed, and new phenomena, which are not currently available to observation, are suggested. This proposal requests support for the development of the next generation of such models. We propose a five year project to develop, implement and test a solution-adaptive, parallel code for multifluid semi-relativistic Hall MHD. The method used will be valid for magnetic fields ranging from very weak to strong enough that the Alfven speeds become relativistic. The analytical work will extend our earlier work in ideal MHD, semi-relativstic MHD, and high-moment models for rarefied flows. The code development will build on the implicit, parallel, block-adaptive framework developed here with previous support from the AISR program. The increased physical sophistication of the multifluid Hall model will allow us to calculate with improved fidelity space-weather events, Earth-ionosphere response, and outer-planet magnetosphere structure, as well as other applications in which resistivity and/or multiple species play a crucial role. Dealing with the additional complexity due to multiple species and the added Hall physics, as well as that due to the collisional processes, will require new development of numerical methods. In particular, work will need to be done on understanding the underlying Riemann problem of the extended equations, and on the effects that the collisional terms have on the evolution of the conservation-law system. However, this development can build on the existing framework for solution-adaptive, parallel ideal and semi-relativsic MHD, leading to a sophisticated physical model in a powerful, efficient computational framework. This proposal meets all three goals of this solicitation. It will help to reduce mission development time, risk, and cost through advanced simulation and design capabilities. These capabilities will be relevant for missions in solar and heliospheric physics, magnetospheric physics, planetary exploration and astrophysics. Finally, it increases interdisciplinary collaboration between space scientists (Gombosi), computational scientists (De Zeeuw, Powell), (De Zeeuw, Powell), plasma physicists (Sokolov, Toth) and computer scientists (Stout). This collaboration started over a decade ago and has resulted in a tightly integrated interdisciplinary team with a series of successful projects.
| |
| Exploration of Novel Methods to Visualize Genome Evolution. (PI: Gogarten, Peter,
University of Connecticut
) | |
| Goal of the proposed research is to develop and test approaches that can be used to visualize and dissect the mosaic nature of genomes. This will allow tracing histories of individual genes, detection of co-evolving traits, and correlation of the molecular record with the fossil and geological records. In the past we have developed tools that map the support of individual gene families for the different possible phylogenetic trees. However, because of the huge number of possible trees topologies, this approach cannot readily be extended to the depiction of many genomes. To facilitate the processing of phylogenetic information, we will organize the data from analyses of multiple genomes into matrices. For each gene family we will utilize either the significantly supported bipartitions, or the support for the different quartets that can be formed by selecting four of the analyzed species. All of these data matrices will be of very high dimension ([number of orthologous genes] x [number of possible tree topologies / quartets / bipartitions]). We will explore different tools to analyze these matrices. First, we will consider the local linear embedding algorithm (LLE), which maps each point in the high dimensional space into a point in two-dimensional space via an encoding based on a covariance matrix of the distances to a selected set of nearest neighbors. This algorithm performs well in providing low dimensional projections of high dimensional data retaining the essential structure of the original data. Another algorithm that will be considered is the self-organizing map (SOM) algorithm. This neural network-based algorithm attempts to detect the essential structure of the input data based on the similarity between the points in the high dimensional space. We will investigate the usefulness of these and other approaches (principal component analysis (PCA), multidimensional scaling (MDS), support vector based kernel PCA, and ISOMAP) to produce informative two-dimensional maps depicting phylogenetic relationships among genomes, and we will develop tools to make gene types and tree topologies readily recognizable in each map. We aim for analyses and diagrams that depict gene families with similar histories as neighboring, whereas genes with different histories are classified into separate clusters. We will explore the utility of the different approaches using well-studied test cases: the eukaryotic genome where currently three large groups of genes are recognized based on their different evolutionary history (those form the host genome and those from the endosymbionts that evolved into mitochondria and plastid), and selected bacterial genomes where genes involved in the same function of metabolic pathway often have the same evolutionary history.
| |
| Idea to Observations: User Support Tools for the Next Decade (PI: Koratkar, Anuradha P.,
University of Maryland Baltimore County
) | |
| This project has ended.
| |
| Development of a Data Federation Facility for the National Virtual Observatory (PI: Brunner, Robert,
University of Illinois Urbana-Champaign
) | |
| Astronomy is entering a new, information-rich era as multiple, large area, digital sky surveys, including many NASA missions, are or will soon be in production. The resulting datasets are truly remarkable in their own right; however, a revolutionary step arises in the aggregation of complimentary multi-wavelength datasets (i.e., the cross-identification of billions of sources). In fact, before any advanced data exploration or mining tools can be employed, the data of interest must be federated. Indeed, this data federation service is one of the primary requirements for the National Virtual Observatory (NVO). Federating these different datasets, however, is a challenging task. In this proposal, we focus on identifying solutions to the problems inherent in the dynamic, multi-wavelength cross-identification of large numbers of Astronomical sources.
These problems arise from both a computational science side (data volumes algorithmic complexities, persistence mechanisms), as well as from the astronomical side (variations in physical phenomena between multiple wavelength observations). In this proposal, we present our plan for developing a robust data federation service for the forthcoming NVO, which includes three separate sub-projects whose own utility exceeds the combined goal. First, we propose to develop AstroForge, a distributed collaborative development facility for scientific computing software projects based on the extremely popular SourceForge.net open source software development website. Second, we will develop a cross-identification toolkit, hosted by AstroForge, to facilitate open discussion and development, that will utilize probabilistic associations to federate disparate astronomical data sets. Finally, we will integrate the data-federation service with a data-mining service on the separately funded NVO framework as a demonstration facility. Different integration mechanisms will be explored, including the Web Service model, the Peer-to-peer model (e.g., using JavaSpaces or JXTA), and direct database integration (e.g. using SQL and Stored Procedures). All of the results from this project, including technical evaluations, will be made publicly available to the community.
| |
| Software Technology to Enable Reliable High-Performance Distributed Disk Arrays (PI: Warren, Michael,
Los Alamos National Laboratory
) | |
The advent of commodity microprocessors with adequate floating-point performance and low-priced fast Ethernet switches contributed to the emergence of Beowulf clusters in the mid-90s. We are currently poised for a similar advance in distributed disk arrays (DDAs), due to the dramatic decline in the price of commodity disk drives. The implementation of reliable DDAs will revolutionize data storage and retrieval in practically all area of NASA information science. The cost per Gbyte for disk storage is currently less than$2.00. Several groups (including ours) have demonstrated fault-tolerant terabyte servers for a total cost of under $4000. Used in a parallel cluster environment, multi-terabyte disk arrays with achievable read/write bandwidths that greatly exceed available Gigabit local and wide-area networking technology are possible. Additionally, the greater CPU/storage ratio in DDA offers techniques which are not possible in traditional RAID arrays. While projects such as the parallel virtual file system have demonstrated clear untility, they lack the fault-tolerance that could be obtained via the efficient calculation and storage of parity or mirroring information between nodes (analogous to RAID techniques within a node). This additional functionality would add orders-of-magnitude to the reliability of mass storage on clustered systems. Also, while disk areal density has been improving at an annual rate of about 60% per year, disk latency has been improving 10%, so disks are becoming increasing unbalanced in terms of capacity and latency. By intelligently replicating and caching data in a DDA, it is possible reduce latencies to access terabytes or more of data by an order of magnitude. Our objectives are to: - Build upon software techniques such as the RAID-x architecture at USC and the xFS project at Berkeley to create high-performance fault-tolerant DDA.
- Demonstrate the ability to trade capacity for reduced latency in DDA using software-only disk head prediction mechanisms which work on a wide range of off-the-shelf hard drives.
- Demonstrate techniques for cheaply and efficiently duplicating and transporting multi-terabyte datasets.
- Address limitations in the transmission-control protocol which limit bulk data transfers.
- Validate the reliability and performance of our tools by applying them to a number of on-going astrophysics and data science projects at LANL.
| |
| N-Chilada: A framework for parallel data analysis and visualization (PI: Quinn, Thomas,
University of Washington
) | |
| We will develop, test and distribute open-source, object-oriented simulation and analysis/visualization framework that will apply generically to a broad range of problems that can be modeled with particles, such as: evolution of large-scale structure in the Universe; galaxy formation; formation and long-term evolution of planetary systems; granular dynamics, protein folding; etc. The goal is to enable the rapid development of parallel tree and adaptive-mesh based algorithms operating on particles, particularly for visualization and analysis. For example, we will implement a tree-based scalable particle dataset visualization algorithm. This method would allow the efficient calculation of smoothed quantities such as density and pressure. The framework will also be adapted for visualization and analysis of large astronomical catalogues such as those that will become available in the National Virtual Observatory. The software will be scalable from single workstations through Beowulf clusters to massively parallel supercomputers, and will provide a Web-based interface for data retrieval, manipulation, and processing.
| |
| Windows to the Universe (PI: Johnson, Roberta,
UCAR
) | |
| We propose a three-year effort to continue work supporting Windows to the Universe project, as a component of the Information Technology Research program. This effort will include: expansion and updating of existing content for space science related sections of the web site; implementation of emerging technologies appropriate to space science education and public outreach; support of student learners utilizing information technology in their math, science, and technology education; correspondence with users; analysis of usage trends; and maintenance of the server and associated site software. In addition to developing this content, we will also disseminate information about the project at the National Science Teachers Association through presentations, participation in the National Earth Science Teachers Association Share-A-Thons, exhibit time, demonstrations and short course, and participate as requested in the NASA ODD Education ecosystem.
| |
| Support for NVO Science Definition Team (PI: Djorgovski, Stanislav George,
California Institute of Technology
) | |
|
| |
| An Interoperable Framework for Data Mining and Analysis of Space Science Data (PI: Graves, Sara,
University of Alabama, Hunstville
) | |
| The F-MASS research team, composed of Information Technology and Space Science researchers propose to create an interoperable Framework for Mining and Analysis of Space Science Data (F-MASS). This framework will extend an existing scientific data mining system by providing Space Science data filters, Data Mining algorithms and customized user interfaces appropriate for the Space Science research domain. Data mining techniques such as statistical analysis, pattern recognition, feature classification, and image processing will be selected for the framework using Space Science research problems as drivers for this development. The F-MASS team consists of Space Science researchers who will be working with the teams Information Technology researchers to adapt the existing ADaM data mining system to provide a comprehensive framework for Space Science data investigations. The framework will focus on problems such as the handling of large datasets inherent in the Space Science communities, the heterogeneity of the data, the incorporation of specialized data mining algorithms, and domain-specific user interfaces. The expected result of this project will be a framework that can be used across the Space Science research community for advanced data analysis and mining. The full proposal details specific Space Science research scenarios that will be used as use case drivers for F-MASS and some candidate data mining techniques that will be incorporated into the framework.
| |
| Astrostatistical Tools in Python (PI: Loredo, Thomas,
Cornell University
) | |
| Astronomers are developing new and powerful statistical methods in response to the recent explosion in astronomical data quantity and quality. These methods promise to significantly increase the amount and quality of science distilled from complex data. Though powerful and computationally complex, most new methods are conceptually straightforward and can be used without knowledge of implementation details. Our projects primary objective is to enable astronomers to use sophisticated methods without requiring them to master the art of statistical computing. To accomplish this, we will pursue four subsidiary objectives:
(1) Within a unifying framework, implement a broad and deep collection of statistical software tools built on algorithms from recent and current research in astrostatistics and computational statistics, wit extensive and accessible documentation. These tools will span may problem domains (e.g., time series, surveys, spectroscopy, imaging), and will offer choices of various approaches wherever possible (e.g., conventional frequentist methods and Bayesian counterparts) so users can easily compare competing methods. A parametric inference engine will supplement basic tools with a framework implementing functionality common to many parameter estimation methods (e.g., constrained optimization, multidimensional exploration and integration, Monte Carlo methods).
(2) Exploit the capabilities of a modern, object-oriented computer language Pythonto implement the tools efficiently and in a form that is seductively easy to use despite the sophistication of the underlying algorithms. Python allows simplified, high-level interfaces to methods with little compromise in computational efficiency. It is being used to rewrite the widely-used IRAF data analysis environment, so IRAF users will have particularly easy access to our tools.
(3) Undertake an outreach program to inform astronomers of the methods and tools by demonstration of their scientific utility. This will include high-visibility presentations at major astronomical meetings demonstrating use of our methods.
(4) Enable collaboration between astrostatisticians and computer scientists to ensure the combined scientific and computational quality of the tools. This project will allow astrostatisticians and a leading architect of Pythons numerical capability to devote significant effort to constructing a robust, well-documented tool kit.
| |
| Multiplierless Filters for Real-Time Processing of Hyperspectral Images (PI: Bose, Tamal,
Utah State University
) | |
| There is a recent thrust by NASA to develop high-resolution hyperspectral imaging systems for use in geostationary vehicles. The concept of hyperspectral imagery is that the imaging interferometer has a sensor module which consists of a NxN array of sensors. Each of these sensors provides a Mx1 spectrum. Therefore, for each spectral component, there is an image representing the amplitude over
a mapped space. This represents a 2-D increase in information over conventional spectrometers which have only one sensor. However, the down side is that there is an incredible amount of data to process, compress, deliver, and analyze. Real-time processing of such data sets is nearly impossible with current signal processing algorithms and hardware technology.
Conventional algorithms use a variety of different filters for processing image data. These filters are implemented using many multipliers and adders. The multipliers are a bottleneck to high speed computations. A solution is therefore to design filters with coefficients that are powers-of-two. The multiplications can now be implemented using simple shifting operations. In VLSI mplementations,
multiplierless filters are ideal because of their high speed and low power. The design of 2-D multiplierless filters is a long standing open problem.
The main goal of this project is to develop the tools and techniques for designing 2-D multiplierless digital filters. These filters will then be used to process hyperspectral image data, both in the spatial
and temporal directions. The final goal is to derive algorithms using these filters for the restoration and compression of hyperspectral images.
Although multiplierless filters and algorithms are ideally suited for processing hyperspectral images from FT spectrometers, they will also be useful in other industrial applications that use, process, store
and transmit images and video.
| |
| Development of a Real-Time Automated Image Enhancement Pipeline for NASA's IRTF and Other Users (PI: Young, Eliot,
Southwest Research Institute
) | |
| The goal of this project is to develop an automated pipeline to perform PIXON image reconstruction on images from NASA's Infrared Telescope Facility (IRTF) with minimal interaction from the
observer. Preliminary work with the Spacewatch data set has overcome many of the challenges facing such a pipeline, such as automated characterization of background and sky noise, automated
extraction of PSFs, and automated identification of non-stellar objects such as galaxies or cosmic ray hits. Several images from the IRTF's recently commissioned SpeX guider camera have been PIXON-processed with extremely promising results.
Based on the early results with SpeX images, we expect to deliver diffraction-limited spatial resolution across the entire 0.85 to .5 micron range of the instrument. The addition of real-time
PIXON processing to the IRTF will open new avenues of research. Eventually enhanced images produced by this pipeline will become an expected benefit of observing with the IRTF. We will concentrate a significant part of our effort on making the pipeline robust and easy to use to aid its adoption by amateurs and other large telescopes.
| |
| Science Education Gateway II (PI: Hawkins, Isabel,
University of California Berkeley
) | |
|
| |
| Testbed Web-based Tool Development to Involve Non-Professionals in Space Science Research (PI: Stern, S. Alan,
Southwest Research Institute
) | |
| Inspired by the SETI@home effort, which has received much attention in
the press, and which already involves over 10$^6$ ``subscribers'' who
have together contributed over 60,000 years of computer time, we
desire to construct a broader set of philosophically-similar
Publically-Accessible Research Application Web Tools (``PARAWTs'') for astronomy and the broader space sciences. Our goal is to multiply the available computing and data gathering capabilities of the astronomical and planetary research communities by demonstrating that {\it SETI@home} success was not an isolated occurance, but instead a new paradigm with broader implications for future public outreach and research resource leveraging.
| |
|
|
|
|
|
