AIST 2002 NRA Released
NASA’s Office of Earth Science Awards 21 Contracts for Advanced Information Systems Technology
09/16/2002 – The National Aeronautics and Space Administration (NASA) has awarded funding
for 21 new investigations for information systems technology development,
under the Advanced Information Systems Technology (AIST) Program, which
supports NASA’s mission to understand and protect our home planet. The
proposals, selected from a field of 200 submitted proposals, focus on
high-priority information technology areas: on-board processing, space-based
communications networks, mission automation, and high-end computing technologies
for modeling. The total funding for these investigations, over a period
of three years, is approximately $23 million; investigators hail from
14 states and the District of Columbia.
The main purpose of AIST is to invest in research and development
of new and innovative information technologies to support and enhance
the Earth Science Enterprise (ESE) science and applications objectives
in the 21st century. AIST focuses on creating mature technologies leading
to smaller, less resource-intensive and less expensive flight systems
that can be built quickly and efficiently, and on more-efficient ground-based
processing and modeling systems that improve the use of Earth science
data.
The technologies selected include the implementation of
science algorithms in hardware for onboard processing, as well as for
the direct distribution of change detection data products from space.
Reconfigurable on-board processors will also be matured to support onboard
data compression, short-range weather forecasting and cloud detection,
and sensor webs and formation flying.
Also selected were space-based network technologies to
facilitate seamless connection from instruments or sensors on-board the
spacecraft to the user community using existing commercial Ethernet standards.
An innovative technique for increasing current printed circuit board memory
capacity by an order of magnitude will enable advanced data processing
and the use of new high-resolution sensors.
Communication and mission automation technologies to enable
the implementation of cooperative observations using constellations and
sensor webs will also be developed. These include novel approaches to
network protocols and architectures, automated scheduling, and mission
automation using Earth phenomena observing systems.
Analysis of Solid Earth, Carbon and Weather data obtained
from Earth Observing satellites will be enhanced through the maturation
and integration of various high-performance computer modeling techniques.
These models will be complemented by the development of an intelligent
data analysis system that automates some ground system data processing
functions.
The investigations selected by NASA’s Office of Earth Science
are:
Abstracts for Panel Recommended Fund
Proposals
| Proposal Number | AIST-02-0028 (Panel 1A) |
| Title | An On-Board Processor for a Spaceborne Doppler Precipitation Radar |
| PI | Stephen L Durden |
| Abstract | |
| The objective of this work is to develop an on-board data processor for a spaceborne Doppler precipitation radar. The radar system and Doppler processor would be used to measure the reflectivity and vertical motion in precipitation, both of which are needed for direct calculations of latent heating. Precipitation and vertical motion measurements are directly related to NASA ESE strategic questions concerning atmospheric variability and predictability. Recent work has demonstrated the feasibility of measurement of vertical motion in the atmosphere using spaceborne Doppler radar. However, for small convective cells, the finite size of the radar footprint, in combination with the large platform velocity, creates large velocity errors when using conventional estimators. It has been shown that errors can be substantially reduced when the full Doppler spectrum, rather than just the mean Doppler, is used. Because of the large data volume associated with down-linking the raw radar data, on-board processing is highly desirable. The FPGA-based on-board processor proposed here performs Doppler spectral estimation, and averaging. The FPGA firmware design will be tested by simulation and by operation in an FPGA board. |
|
| Proposal Number | AIST-02-0045 (Panel 1A) |
| Title | On-Board Processor for Direct Distribution of Change Detection Data Products |
| PI | Yunling Lou |
| Abstract | |
| We will build an on-board imaging radar data processor for repeat-pass change detection and hazards management. This is the enabling technology for NASA ESE to utilize imaging radars. This processor will enable the observation and use of surface deformation data over rapidly evolving natural hazards, both as an aid to scientific understanding and to provide timely data to agencies responsible for the management and mitigation of natural disasters. Many hazards occur over periods of hours to days, and need to be sampled quickly. The new technology has the potential to save many lives and millions of dollars by putting critical information in the hands of disaster management agencies in time to be of use. The processor architecture integrates two key technologies by combining a Field Programmable Gate Array (FPGA) front-end with a reconfigurable computing back-end. Through this approach we are able to capitalize on the strengths of both technologies for the optimization of performance while maintaining flexibility where needed within the algorithmic implementation. A searchable on-board data archival will store the reference data sets needed for change detection processing. The benefit of this technology development is not limited to future spaceborne imaging radar missions, but will aid any NASA and commercial spaceborne mission that requires high-speed FPGAs, digital signal processors, and searchable data archival systems. This technology development is critical to mitigate the technical risk, cost, and schedule of on-board processing for future spaceborne missions. |
|
| Proposal Number | AIST-02-0157 (Panel 1A) |
| Title | Reconfigurable Hardware in Orbit |
| PI | Brian Schott |
| Abstract | |
| This proposal is focused on reconfigurable computing (RC) in space, using field-programmable-gate-array (FPGA) technology. The techniques to be developed address the increased demand of computations in space, resulting from the use of more complex algorithms necessary to acquire, process, and transmit data. This research addresses not only the computational / transmission requirements, but also the desirability to adapt the on-board processing algorithms in real-time, through software up-loads. Thus, this technology seeks to enable the use of Reconfigurable-Hardware-in-Orbit (RhinO), via an integrated design tool-suite aiming to reduce risk, cost, and design time of multi-mission reconfigurable space processors using SRAM-based FPGAs. The research team has considerable expertise in these areas as applied to reconfigurable/adaptive computing systems, radiation effects, power optimization (FPGA-specific), and development of design tools. The effort leverages technologies developed at the University of Southern California s Information Sciences Institute (ISI) under DARPA-funding since 1997 on Adaptive Computing Systems (ACS). This effort is carried-out in collaboration with the NASA centers at Goddard Space Flight Center (GSFC) and Ames Research Center (ARC). GSFC and ARC will provide NASA ESE-driven applications and algorithms, and discuss, guide, and focus appropriately the research findings; furthermore, ARC will participate actively on the design of robust FPGAs for space operations. Other team members are Los Alamos National Laboratory (LANL), and Brigham-Young University (BYU); both of which were actively involved with ISI on the DARPA-sponsored ACS program that developed the baseline technology for RC used today. LANL and BYU will participate in the development of the RHinO toolkit and testing FPGAs in space environment. |
|
| Proposal Number | AIST-02-0076 (Panel 1B) |
| Title | Radiation Tolerant Intelligent Memory Stack (RTIMS) |
| PI | Jeffery Alan Herath |
| Abstract | |
| The Earth Science Enterprise (ESE) has identified many systems that require on-board satellite data processing for future science needs. Continuing themes in addressing these are the ever increasing resolution, quality and quantity of collected data. The ability to efficiently handle these large amounts of data necessitates the use of larger on-board memories. This project will develop a radiation tolerant memory, suitable for both geo-stationary and low earth orbit missions, which provides 2 gigabits of error corrected digital memory to address this mounting need. The Radiation Tolerant Intelligent Memory Stack (RTIMS) will be designed, built, tested and integrated onto a 6U Compact PCI printed circuit board (PCB) by the completion of this project; thereby increasing the potential memory capacity of such a PCB, currently used in space missions, by over 10 times. RTIMS will take advantage of the circuit stacking technology of 3D-Plus USA and new package level radiation shielding. By using FPGA technology for the memory controller, RTIMS can also be used as a key element in adaptive / reconfigurable computing applications. RTIMS will significantly enhance a broad range of high data rate missions. This project will develop and demonstrate the RTIMS technology, which will enable new Earth observation measurements and information products, increase the accessibility and utility of Earth science data, and reduce the risk, cost, size, and development time of ESE space-based systems. |
|
| Proposal Number | AIST-02-0099 (Panel 1B) |
| Title | 10/100 mb/sec flight ready Ethernet Hardware |
| PI | Mike Lin |
| Abstract | |
| The objective of this proposal is to develop Ethernet Local Area Network (LAN) hardware technology in radiation hardened components and build flight capable prototype hardware to validate the designs and reduce risks for future insertion into NASA flight programs. The benefits of Ethernet technology would enable high performance networks & processors, lower end-to-end spacecraft development costs, and the ability to leverage off of the extensive existing commercial infrastructure. Ethernet immediately and directly targets several of the ESE Strategic mission, goals, and objectives for applications and technology outlined in the road map for 2003 to 2025, including missions such as GPM. The approach is as follows. First we leverage off of the previous ESTO funded work by NASA GSFC in the development of a breadboard Ethernet Network Interface Card (NIC) and a breadboard Ethernet switch. The Ethernet NIC was built and tested successfully at 10 and 100 Mb/sec, and the 8-port Ethernet switch is currently being developed. Using this work as a starting point, we would transition both designs into radiation tolerant Field Programmable Gate Arrays (FPGA). The new flight ready designs, NIC and switch, would be built and tested. NASA GSFC and Orbital Sciences would perform the work. The project will take a little over one year to complete. The cost of this development over two fiscal years would be $557K, providing a low cost networking solution for ESE missions, based on the commercial Ethernet standard. This Activity elevates the TRL for flight Ethernet from 4 to 5. |
|
| Proposal Number | AIST-02-0113 (Panel 1B) |
| Title | TCP/IP Router Board (TRB) with Ethernet Interfaces |
| PI | James Edward Joseph |
| Abstract | |
| Increasing spacecraft and instrument capability coupled with faster downlink data rates and tightened funding demand seamlessly accessible instruments and sensors. A TCP/IP Router Board (TRB) with Ethernet Interfaces provides that solution with significant benefits over existing interfaces: 1) Reduced EGSE cost by 50%, 2) Scalable and reconfigureable for future needs, 3) Seamless TCP/IP Access, and 4) Based on existing commercial standards. Ethernet is the most common technology used for networks in a terrestrial environment to support IP. Routers are now being used in business environments to divide company networks into smaller interconnected networks using shared resources. While a spacecraft has only a few network nodes, a router can isolate those nodes into networks by function and provide increased overall bandwidth on the network. Spectrum Astro proposes combining its Ethernet Multipurpose Board (EMB) (TRL4+) with an embedded RISC processor (TRL3) and TCP/IP router software (TRL3) to develop a Space-Qualified TCP/IP Router (TRL7). After design trades identify the preferred configuration, the EMB is modified to support 4-port router development, router software is implemented, and EGSE is developed to test the router. Then a cPSB backplane and a flight-qualifiable TCP/IP router are designed and fabricated. The TCP/IP router is tested thermally and mechanically demonstrating that it meets the requirements for a flight-qualified component. The proposed effort integrates our NASA GRC funded TCP/IP Ethernet hardware with existing router software taking the evaluation model to a flight design. The period of performance is eighteen months. |
|
| Proposal Number | AIST-02-0128 (Panel 1B) |
| Title | A Reconfigurable Computing Environment for On-Board Data Reduction and Cloud Detection |
| PI | Jacqueline Le Moigne |
| Abstract | |
| The objective of this proposal is to investigate the use of reconfigurable computing for on-board automatic processing of remote sensing data. The Reconfigurable Data Path Processor (RDPP), a radiation tolerant alternative to Field Programmable Gate Arrays, developed at NASA/Goddard under ESTO funding, will be the computation engine of our study. For this feasibility study, two basic methodologies will be implemented on the RDPP: data reduction and cloud detection. These two types of processing represent the most important first steps to be performed on-board, since they enable to reduce communication bandwidth, and make subsequent computations simpler, faster and more accurate. An adapted software environment developed for the RDPP will be utilized for the design of these new algorithms. For validation purposes, three applications are being targeted: (1) dimension reduction of hyperspectral data for land use classification purposes , (2) cloud detection of Landsat-7 data, and (3) data reduction and cloud detection of AIRS data for data assimilation purposes. To insure the success of this endeavor, our team represents a blend of researchers with broad expertise in hardware and software design, as well as Information and Earth science, with impressive records of research and numerous publications. The significance of this research goes beyond the development of well-chosen technologies as the results of our work will be the basis to build on-board autonomous systems for future Earth Science and formation flying missions. |
|
| Proposal Number | AIST-02-0065 (Panel 2) |
| Title | Hybrid Ground Phased Array Prototype for Low Earth Orbiting Satellite Communications |
| PI | Dan Mandl |
| Abstract | |
| The purpose for this proposal is to validate new antenna and array signal processing technologies that either used separately or in combination significantly lower the cost to purchase, maintain and operate ground antenna systems in the X and Ka-band for low earth orbiting satellites. This proposal is in the category of ”Space-Based Communications Networks”. These technologies will increase reliability by minimizing or eliminating moving parts. They will increase operational flexibility and capability. If successful, these antenna technologies can enable progressive mission autonomy by providing a cost-effective demand access messaging service to low-earth orbiting satellites via a cell-tower network of antennas. The proposal outlines a set of experiments with some back-end (adaptive beam forming algorithms) and some front-end technology (space fed lens and reflectarray) that in combination can optimize cost and capabilities. The technologies begin at Technology Readiness Level (TRL) 3 for the space fed lens and the reflectarray and begin at a TRL 4 for the adaptive beamforming. For the NASA Ground Network, the typical 11 m dish costs NASA $2 4 million. The sensitivities for the 11 meter antennas are higher than needed and so this new technology will be built to meet the required threshold for typical satellites such as EO-1. This effort explores technologies to lower acquisition cost to a target of under $100K. |
|
| Proposal Number | AIST-02-0086 (Panel 2) |
| Title | RF Agile Low Power Transceiver (LPT) Technology for Future Space-Based Communications Networks |
| PI | Dan Weigand |
| Abstract | |
| Low Power Transceiver (LPT) technology, the result of cooperative government and industry efforts, is an emerging enabler for future, sophisticated space-based communications networks. LPT applies commercially available components and open standards, and serves as a host platform capable of: accommodating mission specific requirements; providing communications across satellite constellations encompassing diverse orbits and communications characteristics; and ensuring connectivity from scientist to instrument. The proposed effort focuses on the addition of programmable RF tuning based on application of emerging high speed converter and programmable filter technologies that, when complete and integrated with LPT, will yield the first true Software Programmable Radio platform for space, encompassing the entire RF to baseband signal processing path. Termed fLPT, this unit will enable future space missions to reconfigure on-orbit operating frequencies in real-time to contend with evolving mission challenges, such as: ad hoc constellation formation; time-varying RF interference/congestion; and geographic spectrum allocations. ITT proposes to develop and test key receive/transmit modules to demonstrate the application and benefits of direct RF sampling techniques for RF agile space transceivers. This effort builds upon a current AIST program – – emphasizing fault tolerant digital electronics in LPT — and produces results that can apply to a broad range of NASA, AFRL and other government/commercial space programs. This proposed AIST effort will start at TRL 3 and progress through TRL 4 and 5 over the 3-year program duration, with emphasis on: developing prototypes of new electronics modules; integration with current LPT capabilities; and validating applicability to space via environmental testing. |
|
| Proposal Number | AIST-02-0125 (Panel 2) |
| Title | Seamless Handover in Space Networks |
| PI | Mohammed Atiquzzaman |
| Abstract | |
| There is significant interest in deploying the Internet protocol in space. A number of NASA-funded projects are studying the possible use of Internet technologies and protocols to support all aspects of data communication with spacecraft. These include prototyping, testing and evaluating various IP-based approaches and solutions for space communications. A spacecraft, moving between ground stations, acts like a mobile node in a communication scenario. Consequently, NASA has been experimenting with Mobile IP for managing handover of spacecraft between ground stations. The handover mechanism in Mobile IP may be too time and bandwidth-consuming in the limited resources environment of spacecraft communication. Moreover, Mobile IP suffers from a number of limitations such as packets losses, increased overhead due to IP-in-IP encapsulation, wastage of network bandwidth, and overloading of routers. The objective of this project is to investigate an alternative handover scheme with a view to solving some of the issues related to Mobile IP. The scheme could thus become a potential candidate for managing handovers and maintaining seamless connectivity in an IP-based space network. In this project, we propose to leverage the dynamic connection setup and multihoming features of SCTP, a new transport layer protocol, to develop seamless satellite handovers. The study will be carried out using simulation and laboratory testbed. The outcome of this project will result in a space-friendly scheme for managing handovers in space networks. |
|
| Proposal Number | AIST-02-0142 (Panel 2) |
| Title | Reconfigurable Protocol Chip for Satellite Networks |
| PI | Andrew A. Gray |
| Abstract | |
| This task proposes the development of methods, architecture, and a hardware prototype for realizing a space-based dynamic reconfigurable protocol chip for Earth Science Enterprise satellite constellations. This architecture, its associated object oriented design methods, and partial reconfiguration techniques enable rapid autonomous reconfiguration of space communications network functions. This reconfiguration provides long-life space communications infrastructure, enables dynamic operation within space networks with heterogeneous nodes, and compatibility between heterogeneous space networks (i.e. distributed spacecraft missions using different protocols). An architecture for implementing a software reconfigurable network processor for satellite communication applications is proposed; a laboratory prototype is developed as part of the proposed effort. This work builds upon numerous advances in commercial industry as well as NASA and military software radio developments to develop reconfigurable space network processing and processors. The development of such radios and the network protocol chip presented here require defining the correct combination of processing methods (!?objects!?) and developing appropriate dynamic reconfiguration techniques as a function of system goals and operating parameters.?n Dynamic reconfiguration techniques to be developed as part of this effort include autonomous network/protocol identification and autonomous network node reconfiguration. Technologies developed will target future satellite constellations and applicable to the CALIPSO mission, flying in formation with Aqua (2004), COACH (2006), EOS-9 Global Precipitation Mission (2007), and Grace follow-on: formation flying spacecraft (2010). They also conform fully to technology thrust areas described in the Earth Science Enterprise Strategic Plan. |
|
| Proposal Number | AIST-02-0202 (Panel 2) |
| Title | Multi-Satellite Virtual Private Network for Space-Based Applications (SpaceVPN) |
| PI | Marcos A Bergamo |
| Abstract | |
| This research addresses the unique problems of multi-satellite networking for creating a secure IP-based communications infrastructure for SensorWebs in space. BBN’s approach focuses on two critical aspects of the SensorWeb problem: the architectural framework for extending standard IPsec-based Virtual Private Networks to space and the Multi Satellite Network (MSN) technologies required to enable on-demand real-time access to on-board instruments, and integration with the terrestrial Internet. BBN’s architectural framework integrates the use of low-cost ground stations, networking between satellites operating in diverse orbits and ground stations, the opportunistic use of the terrestrial Internet to fill in sparse areas in the satellite constellation, and standard secure IP (IPsec) tunnels terminating at the satellites or ground stations. The space network leverages the TRL-3 technologies BBN is developing for NASA’s High Throughput Distributed Spacecraft Network (HiDSN – Contract NAS 3 01101) that integrate the use of dynamic null-steered multi-beams (for maximum spatial reuse of the spectrum) and BBN’s Tdma with Cdma-encoding Multiple Access TCeMA (for variable rates and maximum connectivity), with network layer protocols optimized for space applications. The key results of the proposed work will be the definition of a common ground/space networking capability that could be used to complement GEO relays in future NASA satellites and ground stations. This will be coupled with demonstrations of the developed technologies in a laboratory prototype capable of realistically emulating the satellite mobility in LEO orbits and communication links including null steered beams, and signal derived from beams of widely varying received power. |
|
| Proposal Number | AIST-02-0135 (Panel 3) |
| Title | Intelligent Dataset Identification, Assimilation, Collection and Transformation System |
| PI | Kara Nance |
| Abstract | |
| Earth observation platforms require the identification, collection and synthesis of diverse, geographically distinct, heterogeneous datasets in an attempt to understand or model a phenomenon. At present, in many cases, significant effort is required by the scientists and modelers to gather, organize, and transform the necessary data prior to scientific analysis or model execution. Each of these steps takes time away from the scientific analysis process. This project will address the problems raised as increasingly complex scientific models attempt to utilize multiple, possibly geographically distributed, data sources as their inputs. The solution will facilitate a single, standardized query to a middle-layer system (IDACT), which will automatically handle the tasks of data retrieval and conversion from multiple data sources. IDACT will also utilize intelligent hybrid Petri-net technology for real-time scheduling of analysis and model runs, based on the availability of data and the desired completion deadlines, ensuring that the hardware is utilized as efficiently as possible while providing model inputs in a timely manner. In addition, it will contribute to a middle-ware data archive thus minimizing recalculation and retransformation of identified datasets. The benefits of this approach to the scientist and modeler include seamless autonomous data collection, data system operation, and management of heterogeneous entities in support of scientific analysis and modeling, thus reducing the associated operational time and costs. This project will be conducted using an open-source approach, and all developments will be distributed under the GNU Public License (GPL). |
|
| Proposal Number | AIST-02-0183 (Panel 3) |
| Title | Mission Automation for A Train” Correlative Measurements Using the Earth Phenomena Observing Systems (EPOS)” |
| PI | Stephan E. Kolitz |
| Abstract | |
| Draper will utilize its optimization-based, hierarchical, real-time mission planning and control technology to increase the value of science data gathered by NASA s EOS (Earth Observing System) through optimized allocation of sensor, data, communication, and ground processing resources. Our technology will enable the rapid development of optimized tasking plans for taskable sensors, e.g., TES on Aura. The results of the proposed effort will be useful in the development of innovative concepts of operations for current confederations of satellites, such as the A Train and the Morning Constellation, as well as in the development of revolutionary concepts of operations for future confederations. We will enhance and extend EPOS, developed under the current AIST contract, from its current TRL 3 to TRL 4 in the proposed two-year effort, and to TRL 5 in the option year. Optimization-based plan generation, closed-loop planning and control, and hierarchical decomposition characterize the technology. We will use data from MODIS on Terra to cue the pointing of TES on Aura. The Cloud Mask produced from Terra s MODIS data will be utilized to generate the value function used in the optimization-based planning for tasking TES to maximize science value, i.e., TES will be tasked to point at high-value locations of interest that are not covered with clouds. In addition, we will combine the MODIS Cloud Mask with a cloud mask generated from GEOS-8 data for further improvement in the forecast of cloud cover, which will subsequently improve the value of the science data that results from observations by TES. |
|
| Proposal Number | AIST-02-0191 (Panel 3) |
| Title | Realtime-Reconfigurable Distributed-Computing for Adaptive Science Operations in Satellite Formations using Heterogeneous CPUs & Heterogeneous Connectivity |
| PI | Eric Alan Byler |
| Abstract | |
| Formation Computing Environment (FCE) is an algorithmically-reconfigurable, distributed processing environment enabling automated operation of distributed sensors. FCE enables mission automation of clusters, formations, and distributed sensing systems providing simple, low-cost distributed aperture formations. Satellite formations are key elements of Earth Science Enterprises’ Strategic Plan in support of Space and Earth Sciences Vision 2010. Implementation of FCE will enable a new class of instruments supporting future ESE science applications including Bi-directional Reflectance Distribution Measurements, CME/Magnetospheric Imaging Radio Arrays, Synthetic Aperture Radars, and related distributed measurement constellations. A satellite-formation middleware tool suite containing adaptive network agents and support services for heterogeneous computing environments will be developed and tested on three different pre-existing hardware configurations. The key objective is algorithmically re-configurable distributed processing for use on heterogeneous clusters of autonomous multi-sensor spacecraft. The agents respond to requests for instrument reconfiguration, fault management, calibration updates, and real-time events to coordinate formation re-pointing, allocation of computing resources, and data storage rates. We will develop a single, integrated solution for all formation operations, allowing scientists to operate a distributed aperture sensor or cluster as a single instrument. The solution includes validated agents for a range of local communications systems, providing formation designers with flexibility to design for cost and performance without affecting integration schedules or software development. FCE development will occur in an integrated, six-spacecraft formation mission test-bed for realistic formation interactions. |
|
| Proposal Number | AIST-02-0240 (Panel 3) |
| Title | Planning and scheduling of coordinated science observations |
| PI | Robert Morris |
| Abstract | |
| The scientific need for multiple sources of earth science data in order to explain complex phenomena will require coordination in instrument scheduling and operations. NASA is developing mission and instrument operations concepts to meet the challenges of managing numerous EOS spacecraft. Nonetheless, there is currently little planning infrastructure to enable the coordination of observations for Earth observing satellites. At the present time, coordination is done between distinct missions with their own mission operations centers. Each of these missions builds their own schedules, and coordination is done manually. To meet the imminent challenges in coordinated science planning for future EOS missions, a team comprised of experts in planning technology, systems integration and engineering, earth science, and mission operations is required. Such a team has been assembled for the proposed project. We propose to develop both a set of technology concepts and a prototype system based on these concepts that will address problems in coordinated scheduling while acknowledging the realities of present day mission design concepts. More specifically, two major tasks will be undertaken. The first task involves defining and solving problems of constellation coordination using advances in automated planning technology. We propose to design algorithms that can be used to integrate new observation requests with an EOS constellation’s existing operations plan. The second task is an integration of science observation scheduling technology developed at NASA Ames with the Automated Mission Planning and Scheduling (AMPS) system, developed at NASA Goddard Space Flight Center for use with EOS missions involving either single or multiple satellites. The final products of this two year effort will be, first, a technology plan for coordinated science planning for multiple missions, and second, an enhanced automated tool for coordinated science planning that can form the basis for infusion into future EOS missions. |
|
| Proposal Number | AIST-02-0036 (Panel 4) |
| Title | Implementing an efficient supercomputer-based Grid Compute Engine for end-to-end operation of a high-resolution, high data-volume terrestrial carbon cycle model. |
| PI | Peter E Thornton |
| Abstract | |
| We propose to implement an existing and much-used model of the terrestrial carbon cycle as a high-resolution, high data-volume Grid Compute enabled system, demonstrating end-to-end functionality in a parallel supercomputing environment. The purpose of this system is to automate the location, retrieval, and translation of input model driver data, and provide a flexible and convenient high-throughput control and execution platform for large model simulations, analyses, and visualizations. The system will consist of five functional components integrated by a single software framework allowing remote user control and inteaction. The five components are: 1)An interpolation engine that acquires surface weather observations and produces high-resolution gridded output; 2)A state-of-the-art model of terrestrial biogeochemical cycles that acquires gridded surface weather fields, performs a user-defined sequence of simulations and produces a multi-dimensional output dataset; 3)A post-processing analysis and evaluation engine that integrates model output with data from existing operational remote sensing, flux network, and global coupled model data streams, producing a user-defined set of model summaries and evaluations; 4)A visualization engine that acquires products from all other components and produces a user-defined array of static and dynamic visualizations to assist in model development, interpretation of results, and transfer of knowledge to broad audiences; 5)A mass storage system with high speed bidirectional connection to the computational engines and their input and output data. Our goal is to provide a powerful parallel implementation of a state-of-the-art simulation package as an end-to-end solution for carbon cycle research and model development, without requiring on-site access to all (or any) of the necessary hardware and data components. |
|
| Proposal Number | AIST-02-0053 (Panel 4) |
| Title | Complexity Computational Environments: Data Assimilation SERVO Grid |
| PI | Andrea Donnellan |
| Abstract | |
| We will use Web (Grid) service technology to demonstrate the assimilation of multiple distributed data sources (a typical data grid problem) into a major parallel high-performance computing earthquake forecasting code. Such a linkage of Geoinformatics with Geocomplexity would demonstrate the value of the SERVO Grid concept, and advance Grid technology by building the first real-time large-scale data assimilation grid. It demonstrates the data assimilation component of the envisioned problem solving environment where it will integrate best practice including that of Goddard’s DAO group with Grid technology and adapt it for solid earth applications. It exploits the current CT effort in Active Tectonics by building on interoperability framework and parallel codes being developed in this project. Note that our project differs from current datagrid activities such as those in particle physics and the astronomical virtual observatory where the simulations are not large scale tightly coupled parallel codes. In the recent ESTO computational technology workshop, the solid earth community identified some key-unifying infrastructure that would benefit the field. Here we develop the next steps for both the Solid Earth Research Virtual Observatory (SERVO) concept and the identified need for a Solid Earth problem-solving environment. We use a challenging motivating problem of importance to NASA namely integrating NASA space geodetic observations with numerical simulations of a changing earth. |
|
| Proposal Number | AIST-02-0071 (Panel 4) |
| Title | Merging the NetCDF and HDF5 Libraries to Achieve Gains in Performance and Interoperability |
| PI | Russell Keith Rew |
| Abstract | |
| The proposal will merge Unidata’s netCDF and NCSA’s HDF5, two widely-used scientific data access libraries. Users of netCDF in numerical models will benefit from support for packed data, larger datasets, and parallel I/O, all of which are available with HDF5. Users of HDF5 will benefit from the availability of a simpler high-level interface suitable for array-oriented scientific data, wider use of the HDF5 data format, and the wealth of netCDF software for data management, analysis and visualization that has evolved among the large netCDF user community. The overall goal of this collaborative development project is to create and deploy software that will preserve the desirable common characteristics of netCDF and HDF5 while taking advantage of their separate strengths: the widespread use and simplicity of netCDF and the generality and performance of HDF5. To achieve this goal, Unidata and NCSA will collaborate to create netCDF-4, using HDF5 as its storage layer. Using netCDF-4 in advanced Earth science modeling efforts will demonstrate its effectiveness. The success of this project will facilitate open and free technologies that support scientific data storage, exchange, access, analysis, discovery, and visualization. The technology resulting from the netCDF-4/HDF5 merger will benefit users of Earth science data and promote cross-disciplinary research through the provision of better facilities for combining, synthesizing, aggregating, and analyzing datasets from disparate sources to make them more accessible. |
|
| Proposal Number | AIST-02-0133 (Panel 4) |
| Title | Coupling High Resolution Earth System Models Using Advanced Computational Technologies |
| PI | Christa D Peters-Lidard |
| Abstract | |
| The primary objective of the work is to apply advanced computational technologies to the problem of coupling high-resolution (e.g., 1 km, or cloud-scale) Earth system models. Specifically, we propose to combine the emerging technologies of the Earth System Modeling Framework (ESMF) and the Land Information System (LIS) to couple complex Earth system model components, including the Weather Research and Forecasting Model (WRF) and the Goddard Cumulus Ensemble (GCE) model to state-of-the-art land surface models including the Community Land Model (CLM) and the NOAH Land Surface Model, which are already part of the Land Information System. The technologies to be advanced and applied in this work include 1) support for parallel computing architectures (via MPI on the LIS Beowulf cluster); and 2) support for distributed data access, transport, translation, mining and conversion (via the Grid Analysis and Display System/Distributed Oceanographic Data System (GrADS/DODS) server and the ESMF). Both technologies are currently being demonstrated in the LIS for the problem of uncoupled land surface modeling, and this proposal seeks to further their technological readiness levels by applying them to the problem of high-resolution coupled modeling. These technologies are critical to advance ESE science and prediction goals, in which geographically distributed databases may hold petabytes of Earth system observations needed for high-resolution modeling and data assimilation to understand and predict water, energy, and carbon cycles. Accomplishments of prior work. The PI has significant experience in the application of advanced computational technologies to land surface modeling. Since 1996, Dr. Peters-Lidard has been collaborating with colleagues at the North Carolina Supercomputing Center to develop a high spatial resolution land surface model for numerical weather prediction and air quality applications (http://www.emc.mcnc.org/projects/dashmm/). The PI has experience coupling a land surface model to a meteorological model using an Input/Output Applications Programming Interface (I/O API) designed to facilitate multiscale, distributed, parallel environmental computing. As part of this work, a hundred-fold reduction in CPU, RAM, and disk space requirements has been achieved. The PI and Co-I Houser are currently leading the ESTO/CT LIS project, whose TRL3 technologies will be advanced and applied in the currently proposed project. In addition to the LIS project, Co-I Houser is a pioneer in the development of Land Data Assimilation Systems (LDAS) and the CLM model. Co-I Tao has over 25 years of experience developing and improving (microphysics, cloud-radiation, advection, surface processes) non-hydrostatic atmospheric models, including over 110 publications. Dr. Tao is a member of the WRF science steering team and the lead developer of the Goddard Cumulus Ensemble (GCE) model. Outline of proposed work and methodology. The proposed work will proceed according to well-defined milestones in order to apply and advance the LIS TRL3 technologies to TRL5. The first phase of the project will focus on coupling the LIS to the WRF and GCE models using parallel techniques with the ESMF framework for timing, gridding and control and the GrADS/DODS server for distributed data access and mining. The second phase of the project will focus on the evaluation of these systems for the case study of the International H2O Project (IHOP) field campaign, including performance tuning. In the Earth system modeling arena, this will represent an advancement to TRL4, since this field campaign is a ”natural laboratory”, which will allow thorough evaluation of the technologies. The third phase of the work will focus on expanding the application to regional and continental scales in order to advance the technology to TRL5, which is representative of a typical application environment. |
|
| Proposal Number | AIST-02-0160 (Panel 4) |
| Title | Integration of OGC and Grid Technologies for Earth Science Modeling and Applications |
| PI | Liping Di |
| Abstract | |
| We propose to integrate OGC and Grid technologies for making NASA EOSDIS data easily accessible to Earth science modeling and applications communities. OGC web service technology is developed for providing interoperable access and services of geospatial data. The built-in OGC geospatial services include subsetting, resampling, georectification, reprojection, reformatting, and visualization. GRID technology is developed for sharing data, storage, and computational powers of high-end computing within a virtual organization. The integration of the two technologies will make GRID technology geospatial enable and OGC standard compliant and make OGC technology GRID enable. The integration will allow researchers to focus on science and not issues with data receipt, format, and data set manipulation. The power of the integration will be demonstrated in NASA EOSDIS Data Pools, which host peta-bytes of data on-line, to show the huge volumes of EOS data can be easily accessible by Earth sciences models and applications in ready-to-analyze forms through OGC interfaces supported by the Grid technology at the backend. This project is built upon results of our previous and current projects, including OGC compliant NASA HDF_EOS Web GIS Software Suite (NWGISS) and CEOS Grid testbed (Di), Globus and NASA Information Power Grid (IPG) (Johnston), and DOE’s Earth System Grid (ESG) (Williams). |
|