A Risk Index Methodology for Potentially Polluting Marine Sites (PPMS)
Title | A Risk Index Methodology for Potentially Polluting Marine Sites (PPMS) |
Publication Type | Journal Article |
Year | 2014 |
Authors | Masetti, G, Calder, BR |
Journal | Environment Systems and Decisions |
Volume | 34(3) |
Pages | 391-405 |
Date Published | 2014/05/25 |
Publisher | Springer |
Keywords | level of confidence, PPMS, Quality Symbol, risk assessment, Risk Index |
Attempting to assess the risk of a release from a Potentially Polluting Marine Site (PPMS) can be a very subjective process. The Marine Site Risk Index (MaSiRI) is designed to provide a more objective approach to this process by adopting a table-based evaluation scheme, while still allowing for the inevitable unknown conditions by including a subjective “expert correction” in a suitably controlled manner. Building on a Geographic Database (GeoDB) of PPMS records, the MaSiRI algorithm applies data filters to remove PPMS records for which it is not applicable, and then estimates a basic risk index based on core data that almost all sites would contain. It can then refine the results for those sites that have auxiliary data, varying the assessed risk as appropriate, according to standard rule-sets. A Risk Level of Confidence (RLC) is computed and adjusted to express dynamic confidence in the risk value (e.g., due to reliance on estimates rather than measured values), and where appropriate an upper and lower bound of risk can be used to assess the range of values associated with an estimated parameter. This information can be visualized by a composite quality symbol proposed here. MaSiRI is demonstrated on three illustrative shipwrecks, and then compared against the DEEPP project database from the Pelagos Sanctuary in the western Mediterranean. The aggregate results of the comparison are broadly similar to DEEPP, within the limits of the comparison, but provide a more detailed analysis in the case of estimated pollutant volume, and ubiquitous assessment of levels of confidence. | |
Publication Link | http://link.springer.com/article/10.1007/s10669-014-9504-7# |