Geospatial grid

Abstract

Geospatial science is the science and art of acquiring, archiving, manipulating, analyzing, communicating, and utilizing spatially explicit data for understanding both physical, biological, and social systems on the Earth's surface or near the surface. One of the very significant aspects of geospatial science is that it is not unique to any one discipline; instead it is an interdisciplinary science, built on the knowledge of many science disciplines. Because geospatial science can be utilized in virtually all situations and human activities that are spatially related, it is becoming more and more integrated into many aspects of contemporary life. Three significant features distinguish geospatial research from scientific endeavors in other scientific domains: 1) the research is multi-disciplinary; 2) the research needs a large amount of data and information and may be computationally intensive; and 3) the research can be on regions that span multiple spatial scales and may be micro to macro in nature (e.g., a leaf, a field, or an area of local, regional, continental, or global extent). Normally, the geospatial knowledge discovery process involves three consecutive data and information flow steps: 1) Geoquery, 2) Geodata and information assembly, and 3) Geocomputation. Geoquery is the location (discovery) and acquisition of data from data repositories. Geodata and information assembly is assembly of the data and information from potentially distributed and heterogeneous data repositories in a way that satisfies the needs of geocomputation. Geocomputation is analysis and simulation of the complex Earth system using the data and information from the geoquery process. Because of their importance in social and economic activities, large amounts of geospatial data have been collected by various public and private sector organizations mainly using remote sensing methods. Those data must be converted to information and knowledge before they become useful. Because of the multidisciplinary nature of geospatial science, a typical geospatial research and applications project requires access to and manipulation of data and information from multiple sources provided by 122 multiple data and information systems. Currently, datasets from data centers are diverse; the data products may differ in spatial/ temporal extent and resolution, origin, format, name conventions, and map projection. Scientists spend considerable time assembling the data and information into a form ready for analysis in the geocomputation step, even when the analysis is very simple. An estimated 50% ~ 80% of the research time spent by ESS scientists is for data and information discovery and assembly (Di and McDonald 1999). The fundamental problem cause is that system of, and the data, and information available from the different data repositories are not interoperable. This paper discusses the geospatial Grid technology that will significantly reduce the problems associated with archiving, manipulating, analyzing, and utilizing large volumes of geospatial data at distributed locations. © 2006 Springer-Verlag Berlin Heidelberg.