Title: Use Of Pattern Recognition To Identify The Source Of An Oil Spill On An Inland Water
Author: W Whalley
Author: P Robotham
Author: M O'Connor
Author: Environment Agency
Document Type: Monograph
Annotation: Environment Agency Project ID:EAPRJOUT_355, Representation ID: 89, Object ID: 1699
Abstract:
A comprehensive literature search has shown that statistical and neural network techniques of pattern recognition have been successfully used to identify standard refined oils and the source of crude oil spills affecting coastal areas. However, no evidence has been found of the use of pattern recognition techniques to identify the source of an oil spill affecting inland waters. It has been shown that oil spills on inland waters present a different kind of problem to that addressed by earlier studies, owing to the inevitable sparcity of data on the potential sources. It is concluded that in this case a two-stage approach to the problem is required, the first to identify the type of refined oil responsible for the spll; and the second to identify the true source from all potential sources. Potential solutions to the first task are well known, since it presents a similar problem to that addressed by many earlier studies. A potential solution to the second task has been found via an exploratory investigation based upon a visual pattern recognition exercise involving a specific case study. It is concluded that the use of computer based pattern recognition techniques would improve the reliability and defensibility of evidence given in court, and recommendations are made for a phased programme of research leading to the development of a comprehensive source identification system.
Publisher: Environment Agency
Subject Keywords: Rivers; Surface water; Brackishwater environment; Freshwater ecology; Marine environment
Extent: 97
Permalink: http://www.environmentdata.org/archive/ealit:4346
Total file downloads: 276
Download PDF Display PDF in separate tab
