IN33C-01 INVITED
Pain and Gaps in IT Infusion
The process of adopting a new information technology 'X' within geoscience research projects is hindered by
two strong barriers: The pain associated with learning about, adopting and adapting to X, and corresponding
gaps in the 'ease-of-adoption' process left by the builders of X. As builders and providers of two such X's we
discuss several lessons learned from two distinct points along the data pipeline (data acquisition, storage,
retrieval, archival, cleaning, provenance, browsing and analysis). We begin with work at Microsoft Research
to generalize the CUAHSI Observations Data Model to a "next generation" Environmental Data Model (EDM)
with the idea of supporting trans-disciplinary information across remote sensing, in situ, sample analysis,
archival, and model data spaces. We then turn to an in situ sensor network microserver developed through
NASA support for harsh environment data acquisition. The primary 'IT infusion' candidate research project
here is SEAMONSTER, the Southeast Alaska Monitoring Network for Science, Telecommunications,
Education and Research. We trace the adoption pathway, including gaps and pain, from deployment through
to data registration on an EDM data catalog server. We discuss architecture, documentation and technical
support in terms of an end-result success metric: How easily can this project's open data results be
discovered and used?
IN33C-02
Earth Information Exchange --- An Interoperable Spatial Web Portal for sharing Earth Science Information and a Testbed for infusing new technologies
The bridging of accumulated Earth Observation and Model Simulation Earth science information and urgent
needs of the information in different communities and applications poses a global challenge in the 21st
century. The Earth Science Information Partnership (ESIP) proposes to develop an Earth Information
Exchange (EIE) in testing relevant technology mechanisms to provide a solution/platform for addressing the
challenge. Through the development in the past years of infusing different new technologies to support the
platform, a near operational EIE is ready.
The EIE provides widely needed functions through new technologies in 1) accessing multiple catalogs, such
as GCMD and ECHO, from the same entry point; 2) supporting interoperable interfaces, such as OGC
WMS/WFS/WCS and CS-W, and community standards, such as OpenDAP and HDF-EOS; 3) providing multi-
dimensional data selection and visualization with NASA World Wind and Google Earth; 4) providing
knowledge based information search with Earth Science ontologies; and 5) searching and discovering
catalogs progressively.
Technologies gradually integrated into the EIE includes 1) generic IT technologies, such as JSR168/268 and
Ajax; 2) interoperable interfaces, such as OGC specifications and ISO/TC211 and FGDC standards; 3)
multiple catalogs, such as GOS and NCDC; 4) Service Oriented Architecture (SOA), such as geoscience
interoperability; 5) semantic and ontology for knowledge reasoning, such as SWEET and NOESIS; 6) 3D and
4D visualization techniques, such as that provided by World Wind and Google Earth; 7) and many others.
This paper will introduce the strategies we took in integrating those technologies to support the functionalities
of information and knowledge sharing for different communities and applications.
http://eie.esipfed.org
IN33C-03
Land Cover Change Community-based Processing and Analysis System (LC-ComPS): Lessons Learned from Technology Infusion
The Land Cover Change Community-based Processing and Analysis System (LC-ComPS) combines grid
technology, existing science modules, and dynamic workflows to enable users to complete advanced land
data processing on data available from local and distributed archives. Changes in land cover represent a
direct link between human activities and the global environment, and in turn affect Earth's climate. Thus
characterizing land cover change has become a major goal for Earth observation science. Many science
algorithms exist to generate new products (e.g., surface reflectance, change detection) used to study land
cover change. The overall objective of the LC-ComPS is to release a set of tools and services to the land
science community that can be implemented as a flexible LC-ComPS to produce surface reflectance and
land-cover change information with ground resolution on the order of Landsat-class instruments. This
package includes software modules for pre-processing Landsat-type satellite imagery (calibration,
atmospheric correction, orthorectification, precision registration, BRDF correction) for performing land-cover
change analysis and includes pre-built workflow chains to automatically generate surface reflectance and
land-cover change products based on user input.
In order to meet the project objectives, the team created the infrastructure (i.e., client-server system with
graphical and machine interfaces) to expand the use of these existing science algorithm capabilities in a
community with distributed, large data archives and processing centers. Because of the distributed nature of
the user community, grid technology was chosen to unite the dispersed community resources. At that time,
grid computing was not used consistently and operationally within the Earth science research community.
Therefore, there was a learning curve to configure and implement the underlying public key infrastructure
(PKI) interfaces, required for the user authentication, secure file transfer and remote job execution on the
grid network of machines. In addition, science support was needed to vet that the grid technology did not
have any adverse affects of the science module outputs. Other open source, unproven technologies, such
as a workflow package to manage jobs submitted by the user, were infused into the overall system with
successful results. This presentation will discuss the basic capabilities of LC-ComPS, explain how the
technology was infused, and provide lessons learned for using and integrating the various technologies while
developing and operating the system, and finally outline plans moving forward (maintenance and operations
decisions) based on the experience to date.
http://esipapp03.umiacs.umd.edu:8080/LC-ComPS/
IN33C-04
Satellite Sensornet Gateway Technology Infusion Through Rapid Deployments for Environmental Sensing
The Satellite Sensornet Gateway (SSG) is an ongoing ESTO Advanced Information Systems Technology
project, at the University of Southern California. The major goal of SSG is to develop a turnkey solution for
building environmental observation systems based on sensor networks. Our system has been developed
through an iterative series of deployment-driven design, build, test, and revise which maximizes technology
infusion to the earth scientist.
We have designed a robust and flexible sensor network called Sensor Processing and Acquisition Network
(SPAN). Our SPAN architecture emphasizes a modular and extensible design, such that core building blocks
can be reused to develop different scientific observation systems.
To support rapid deployment at remote locations, we employ satellite communications as the backhaul to
relay in-situ sensor data to a central database. To easily support various science applications, we have
developed a unified sensor integration framework that allows streamlined integration of different sensors to
the system. Our system supports heterogeneous sets of sensors, from industry-grade products to research-
specific prototypes. To ensure robust operation in harsh environments, we have developed mechanisms to
monitor system status and recover from potential failures along with additional remote configuration and
QA/QC functions.
Here we briefly describe the deployments, the key science missions of the deployments and the role that the
SSG technology played in each mission.
We first deployed our SSG technology at the James Reserve in February 2007. In a joint deployment with the
NEON project, SDSC, and UC Riverside, we set up a meteorological station, using a diverse set of sensors,
with the objective of validating our basic technology components in the field. This system is still operational
and streaming live sensor data.
At Stunt Ranch, a UC Reserve near Malibu, CA, we partnered with UCLA biologist Phillip Rundel in order to
study the drought impact on deep and shallow rooted plants. Our system was deployed in December 2007
and monitors sap flow on various plant species, while using a satellite link for real-time data access.
In April 2008, in a joint deployment with UCLA, UC Merced, and GLEON, our SSG technology was used to
study the impact of agricultural run off in a series of salt lakes near Bahia Blanca, Argentina. Our system
collected meteorological data that were combined with water quality measurements taken from boats and
buoys.
Our SSG technology was used at the PASI workshop in June 2008 at the La Selva Biological Research
Station in Costa Rica. As part of a two-week curriculum, students from throughout the americas used our
system to collect measurements in the rain forest and later analyzed the data. La Selva plans to install
several SSG nodes throughout the reserve and make mobile nodes available for visiting researchers to use
in their research.
We are currently planning a deployment with environmental engineer Tom Harmon from UC Merced to build
an autonomous water quality flow path and reactive transport observation system near Merced, CA. SSG
technology will be deployed to monitor soil, groundwater, and surface water parameters.
In China's Guizhou Province, we are collaborating with researcher Sarah Rothenberg, who is studying
mercury cycling in rice paddies. Our SSG system will collect soil parameters such as pH and ORP, in addition
to environmental measurements such as PAR, and UV.
This presentation will describe the SSG project, the SPAN prototype and our experience with technology
infusion from the deployments.
class="ab'>
IN33C-05
Autonomous Mission Operations for Sensor Webs
We present interim results of a 2005 ROSES AIST project entitled, "Using Intelligent Agents to Form a Sensor Web for Autonomous Mission Operations", or SWAMO. The goal of the SWAMO project is to shift the control of spacecraft missions from a ground-based, centrally controlled architecture to a collaborative, distributed set of intelligent agents. The network of intelligent agents intends to reduce management requirements by utilizing model-based system prediction and autonomic model/agent collaboration. SWAMO agents are distributed throughout the Sensor Web environment, which may include multiple spacecraft, aircraft, ground systems, and ocean systems, as well as manned operations centers. The agents monitor and manage sensor platforms, Earth sensing systems, and Earth sensing models and processes. The SWAMO agents form a Sensor Web of agents via peer-to-peer coordination. Some of the intelligent agents are mobile and able to traverse between on-orbit and ground-based systems. Other agents in the network are responsible for encapsulating system models to perform prediction of future behavior of the modeled subsystems and components to which they are assigned. The software agents use semantic web technologies to enable improved information sharing among the operational entities of the Sensor Web. The semantics include ontological conceptualizations of the Sensor Web environment, plus conceptualizations of the SWAMO agents themselves. By conceptualizations of the agents, we mean knowledge of their state, operational capabilities, current operational capacities, Web Service search and discovery results, agent collaboration rules, etc. The need for ontological conceptualizations over the agents is to enable autonomous and autonomic operations of the Sensor Web. The SWAMO ontology enables automated decision making and responses to the dynamic Sensor Web environment and to end user science requests. The current ontology is compatible with Open Geospatial Consortium (OGC) Sensor Web Enablement (SWE) Sensor Model Language (SensorML) concepts and structures. The agents are currently deployed on the U.S. Naval Academy MidSTAR-1 satellite and are actively managing the power subsystem on-orbit without the need for human intervention.
IN33C-06
Sensor-web Operations Explorer (SOX)
The Sensor-web Operations Explorer (SOX) is a research task under the Advanced Information Systems Technology project of the National Aeronautics and Space Administration (NASA). The objective of SOX is to develop an integrated software infrastructure (combining air-quality observations with models and data- assimilation tools) that permits a focused analysis of the chemical state and that can adapt to meteorological and chemical "events" over daily time scales. Processes governing the distribution and evolution of trace gases and aerosols have a profound impact on air quality and climate. Trace gases and aerosols do not only affect air quality, but they may also impact regional and global climate through longer-lived greenhouse gases, e.g., O3, CO2, and CH4 Aerosols can have a net cooling or heating effect depending on their type and vertical distribution. The quantification of these processes requires an integrated approach that combines observations from satellites, aircraft, sondes, and surface measurements with chemistry and transport models acting on both regional and global scales. The integrated observation is approached in two modes, an exploratory observation mode and a targeted observation mode. Currently, the exploratory observation mode is fully supported by the SOX on-line service employing a concept-design and an observing system simulation experiments (OSSE) framework. The exploration process needs to be iterated for maturation of a complex sensor-web operation scenario design. For the targeted observation mode, a 4D-variational adjoint framework is being developed in collaboration with the Global Earth Observation System for Chemistry (GEOS-Chem) research teams at Jet Propulsion Laboratory and Havard University. In addition to remote sensing, advances in global chemistry and transport models (along with 4-D variational assimilation techniques) provide powerful tools for the development of sensor webs that could, in principle, be deployed at operational time scales to provide up-to-date information on air pollution useful for decision support as well as enhanced scientific return. The integrated campaign plan describes the assets used in a sensor web along with the assimilation and modeling technologies that combine these assets. Currently, the SOX system can support OSSEs for air-borne sensors and for space-borne sensors (in low- Earth orbit (LEO) and geosynchronous Earth orbit (GEO)). The SOX system supports integrated air-quality campaigns involving both space-borne and air-borne sensors by forecasting the influence of current observations over the entire globe at multiple atmospheric layers and by analyzing the maximum impact zone. In the future, the SOX system will complete the development extending the OSSE services to include in-situ sensors that track surface emissions. The SOX technologies are being infused to several mission concept studies including the Climate Absolute Radiance and Refractivity Observatory (CLARREO) and the Geostationary Coastal and Air Pollution Events (GEOCAPE), part of theTier-1 and Tier-2 missions recommended by the NRC decadal survey. This paper will present technical approaches and implementation details to achieve these successful technology infusions.
IN33C-07
A Multi-Agent Framework Manages a Representative Sensor Web
NASA's vision of a Sensor Web (which includes a distributed global observation system) consists of a large number of elements, such as remote spacecraft hosting multiple instruments, in situ terrestrial and oceanic sensor networks, and airborne assets. Researchers and developers of a Sensor web face a number of challenges that arise from (1) the inherent heterogeneous and geographical distributed nature of the Sensor web; (2) the myriad mission goals and objectives that must be satisfied by the Sensor web, ranging from an improved understanding of earth science, weather forecasting, and disaster management to an alleviation of societal problems; and (3) the need to support myriad operational modes, such as long and short-term monitoring and targeted observations. Resolving these challenges requires some form of autonomy - typically embodied in software. Agent technology has emerged both as a salient purveyor of entities that exhibit autonomous behavior and also as a paradigm for constructing complex software systems with a large number of interacting heterogeneous components. This paper describes our experiences integrating the Multi-agent Architecture for Coordinated Responsive Observations (MACRO) into the SouthEast Alaska MOnitoring Network for Science, Telecommunications, Education, and Research (SEAMONSTER). MACRO provides agents at (1) the mission level, where agents interact with users to define science goals and then translate these goals into a set of prioritized tasks that have to be executed to achieve these goals, and (2) the resource level, where agents translate tasks into activities related to data collection, data analysis, and data communication. As a representative small-scale sensor web situated in multiple locations on the Juneau Icefield, SEAMONSTER affords an unparalleled opportunity to develop, mature, and showcase MACRO's multi-level agent capabilities. MACRO is developed by the Lockheed Martin Space System Company's Advanced Technology Center (ATC), and the Institute for Software Integrated Systems (ISIS), Vanderbilt University. SEAMONSTER is developed at the University of Alaska Southeast. Both projects are recipients of funding from the NASA Earth Science Technology Office's (ESTO) Advanced Information Systems Technology (AIST) program.
IN33C-08
A Forest Fire Sensor Web Concept with UAVSAR
We developed a forest fire sensor web concept with a UAVSAR-based smart sensor and onboard automated response capability that will allow us to monitor fire progression based on coarse initial information provided by an external source. This autonomous disturbance detection and monitoring system combines the unique capabilities of imaging radar with high throughput onboard processing technology and onboard automated response capability based on specific science algorithms. In this forest fire sensor web scenario, a fire is initially located by MODIS/RapidFire or a ground-based fire observer. This information is transmitted to the UAVSAR onboard automated response system (CASPER). CASPER generates a flight plan to cover the alerted fire area and executes the flight plan. The onboard processor generates the fuel load map from raw radar data, used with wind and elevation information, predicts the likely fire progression. CASPER then autonomously alters the flight plan to track the fire progression, providing this information to the fire fighting team on the ground. We can also relay the precise fire location to other remote sensing assets with autonomous response capability such as Earth Observation-1 (EO-1)'s hyper-spectral imager to acquire the fire data.