Jess Kozman (CLTech) with co-authors from Schlumberger and the Arizona Geological Survey showed how MetaCarta’s geographic place-name-based search engine can leverage the content of the newly created GeoScience Information Network (GIN). GIN is a unified front end to over 3,000 data stores maintained by US Geological Surveys. GIN leverages open source standards from the Open Geospatial Consortium, a data exchange standard based on the GeoSciML geoscience mark-up language and a ‘prototype’ catalog web service for data discovery. Kozman argues that the use of natural language processing and geographic search applications like MetaCarta improve on keyword-search on more structured data sets. The technology can be considered to provide the US with a ‘virtual’ National Data Repository.
Anwar Dahish (Malaysia Technological University) showed how remote sensing with space borne radar has been used to delineate oil spills in coastal areas. Despeckling, image segmentation and digital classification, and geo-referencing of remote sensing data were used to determine the relative impact of spills of different origin. Digital filtering enabled fingerprinting of light and heavier petroleum products on the water surface.
Upendra Dadi (Oak Ridge National Laboratory) described the query and visualization of extremely large network datasets over the web using ‘quadtree’ based KML. Quadtrees, a spatial indexing technique, allow for scale-dependent selection of image or vector features.
© Oil IT Journal - all rights reserved.