DescriptionGeographic Information Science (GIScience), crosscutting many fields (e.g., geography, social sciences, computer science, geodesy, and information sciences), plays essential roles for transforming geographic data into geospatial information and knowledge, breaking through scientific problems, and improving decision-making practices of broad and significant societal impact. However, fulfilling such roles is increasingly dependent on the ability to handle very large spatial datasets and complex analysis and modeling methods based on synthesizing computational and spatial thinking enabled by cyberinfrastructure (CI), which conventional GIS software approaches do not provide. CI-based integration of geographic information systems (GIS) and spatial analysis and modeling, as a holistic solution, is leading to unprecedented capabilities for transforming geospatial sciences.The purpose of this project is to extend and sustain GISolve, a TeraGrid Science Gateway toolkit for GIScience, for establishing a high performance, distributed, and collaborative CyberGIS framework that couples CI, GIS, and geospatial analysis and modeling capabilities. Through the continuous TeraGrid resource allocation support from previous three years, a set of spatial middleware components has been built into the GISolve Toolkit to glue generic cyberinfrastructure capabilities and geospatial analysis methods. This toolkit has been used to build the TeraGrid GIScience Gateway as a collaborative geospatial problem-solving environment for multi-disciplinary researchers to perform large-scale geospatial analysis and modeling, and help non-technical users directly benefit from accessing TeraGrid capabilities. With the support of TeraGrid high-end computing resources, we have developed a set of high-performance parallel and distributed geospatial computational methods for our research projects. Scalability and efficient use of high-end computing resources are the foci in developing these methods. For example, the parallel agent-based modeling and parallel land use optimization code are scalable to thousands of processors on Abe and Ranger with impressive computational performance. The methods so developed have been applied in solving large- and multi-scale geospatial science problems that could not be solved before, such as the study of geospatial pattern of the impact of global climate change on crop yields. With GISolve being widely used in the GIScience community, new methods continue to be identified, proposed, and integrated in the GISolve Toolkit. To support community-contributed applications, we have developed a streamlined application integration process to facilitate cyberinfrastructure-enabled computation and efficient integration into the science gateway for sharing. This project has been growing dramatically with consistent and extended research collaboration and education efforts such as the collaboration with the U.S. Geological Survey (USGS) in the National Map project and outreach activities with the University Consortium for Geographic Information Science (UCGIS).
OrganizationUniversity of Illinois at Urbana-Champaign
DepartmentGeography and Geographic Information Science
Sponsor Campus GridOSG-XSEDE
Principal Investigator
Shaowen Wang
Field Of ScienceGeographic Information Science