Using the Summed Rank Cusum for monitoring environmental data from industrial processes

Dave Stewardson and Shirley Coleman

Journal of Applied Statistics, 2001, vol. 28, issue 3-4, 469-484

Abstract: Environmental issues have become a hot topic recently, especially those surrounding industrial outputs. Effluents, emissions, outflows, by-products, waste materials, product de-commissioning, land reclamation and energy consumption are all the subject of monitoring, either under new legislation or through economic necessity. Many types of environmental data are often difficult to understand or measure because of their unusual distribution of values however. Standard methods of monitoring these data types often fail or are unwieldy. The scarcity of events, small volume measurements and the unusual time scales sometimes involved add to the complexity of the task. One recently developed monitoring technique is the Summed Rank Cusum (SRC) that applies non-parametric methods to a standard chart. The SRC can be used diagnostically and this paper describes the application of this new tool to three data sets, each derived from a different problem area. These are measuring industrial effluent, assessing the levels of potentially harmful proteins produced by an industrial process and industrial land reclamation in the face of harmful waste materials. The use of the SRC to spot change points in time retrospectively is described. The paper also shows the use of SRC in the significant-difference testing mode, which is applied via the use of spreadsheets. Links to other similar methods described in the literature are given and formulae describing the statistical nature of the transformation are shown. These practical demonstrations illustrate that the graphical interpretation of the method appears to help considerably in practice when trying to find time-series change points. The charts are an effective graphical retrospective monitoring technique when dealing with non-normal data. The method is easy to apply and may help considerably in dealing with environmental data in the industrial setting when standard methods are not appropriate. Further work is continuing on the more theoretical aspects of the method.

Date: 2001
References: View complete reference list from CitEc
Citations:

Downloads: (external link)
http://www.tandfonline.com/doi/abs/10.1080/02664760120034180 (text/html)
Access to full text is restricted to subscribers.

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:taf:japsta:v:28:y:2001:i:3-4:p:469-484

Ordering information: This journal article can be ordered from
http://www.tandfonline.com/pricing/journal/CJAS20

DOI: 10.1080/02664760120034180

Access Statistics for this article

Journal of Applied Statistics is currently edited by Robert Aykroyd

More articles in Journal of Applied Statistics from Taylor & Francis Journals
Bibliographic data for series maintained by Chris Longhurst ().