Development of a GIS based contaminated land decision support system (CLDSS)
PublisherUniversity of Liverpool
MetadataShow full item record
This thesis is mainly focused on developing a contaminated land risk assessment model in GIS. For this purpose Contaminated Land Exposure Assessment (CLEA) and Human Exposure Assessment - Soil, (HERA-Soil) are selected. CLEA is a popular model in construction and remediation industry in the UK is selected. This model is published by the Environment Agency in England and Wales and estimates the generic assessment criteria (GAC) for soil which provides a threshold to assess brownfield sites. CLEA, includes four land use scenario, Residential with Garden, without Garden, Allotment and Commercial, and estimates the exposure through ten exposure pathway for the critical receptor of each scenario. HERA-Soil is a developed version of CLEA that includes the same land uses, exposure pathways; however HERA-Soil applies the Alternative Integration Procedure (AlP) to estimate the GAC, which reduces the simulation time. CLEA and HEAR-Soil models are currently available via Excel spread sheets and are lacking visual representation of the results on a map and GIS functionalities. Another shortcoming of these models is the lack of statistical tests for soil samples of a site. This is important since, in site investigation, the GAC is compared with the soil sample of the site to decide whether the site is contaminated or not. A detailed study of CLEA, HERA-Soil and related statistical issues is undertaken in this study and two plug-ins are created. Plug-in, is a set of software components which adds specific abilities to a larger application. The first plug-in which is HERA-Soil-GIS, which implements HERA-Soil in an open source GIS application, Quantum GIS, thus makes this model available in a GIS environment and empowers it with GIS tools. The second plug-in is Soil-Statistics-Gl S, consists of a variety of statistical outlier tests on the maximum value in a sample and includes six various outlier tests. These tests are consistent with the size of the sample and the level of significance that is required. These two plug-ins complement each other and lead to a detailed comprehensive contaminated land risk assessment. To validate the HERA-Soil-GIS, results of this plug-in are compared with those from CLEA and HERA-Soil, for 54 compounds and almost resemble each other. In order to demonstrate how HERA-Soil and CLEA benefit from GIS a case study is undertaken. Google map and Satellite map can be easily loaded for any location on the globe. This case study involves using HERA-Soil and GIS tools such as voronoi polygon and other GIS features. The site can be divided into blocks (zones) that clarify how the concentration for each contaminant varies across the site. For the purpose of testing the Soil-Statistics-GIS and describing its abilities three samples of varying size and distribution are used and all the tests used in this plug-in are applied. The importance of statistical outlier tests in decision making is discussed by making a decision in one case with and without identifying potential outliers using these tests. To validate this plug-in, all the tests are applied twice, by running Soil-Statistics-GIS and applying the tests on soil samples in Excel and comparing the outcome. Contributions made by this research is connecting HERA-Soil, a developed version of CLEA, to GIS and making it available in a GIS environment, also creating statistical outlier tests available in the same GIS environment for this model.