Application of multi-scale assessment and modelling of landfill leachate migration : implications for risk-based contaminated land assessment, landfill remediation, and groundwater protection
There are a large number of unlined and historical landfill sites across Britain, contaminating groundwater and soil resources as well as posing a threat to human health and local communities. There is an essential requirement for robust methodology when carrying out risk-based site investigations prior to risk assessment and remediation of landfill sites. This research has focused upon the methods used during site investigations for two reasons. Firstly, the site investigation is often conducted using field instruments and methods that do not account for the heterogeneous conditions found at landfill sites. Interpreting geophysical conditions between sampled points is a common practise. Given the complex and heterogeneous conditions at landfill sites, such methodology introduces uncertainty into data sets. Secondly, risk estimation models that simulate groundwater flow and contaminant transport require extensive field information. The data used during model construction will significantly impact contaminant transport simulations. Modelling guidelines also need further development, ensuring that sound modelling practises are adhered to during model construction. To address these concerns, four research objectives were identified: (1) Two new multi-spatial field assessment methods (remote sensing and ground penetrating radar), previously applied in other fields of science, were tested on landfill sites; (2) Kriging was used as a tool to improve landfill-sampling strategies; (3 & 4) Groundwater flow and contaminant transport models were used to evaluate whether different scales of field data and modelling practises influenced modelling assumptions and simulation. The utility of novel field- and airborne-based remote sensing methodologies in identifying the location and intensity of vegetation stress caused by leachate migration and inferring pathways of near surface contamination using patterns of vegetation stress was proven. The results from the kriging investigations demonstrated that additional insight into field conditions could be resolved to identify locations of additional sampling points, and provide information about variability in hydrological data sets. The Ground Penetrating Radar investigations provided three types of valuable near-surface information that could assist in determining landfill risks: buried landfill features, leachate plume locations and local hydrogeological conditions. These combined methods provided detailed synoptic geophysical and contaminant information that would otherwise be difficult to determine. Their application and acceptance as site assessment methods (used under certain landfill conditions) could increase the accuracy of assessing risks posed by landfill leachate. These applications also demonstrated that the most effective site assessments are achieved when integrated with other field data such as soil, vegetation, and groundwater quantity measurements, contaminant concentrations and aerial photographs, providing comprehensive information needed for risk estimation modelling. The modelling analyses found that close attention must be paid to site-specific and model-specific characteristics, as well as modelling practises. These factors influenced model results. By using additional data to infer model parameters, it was evident that the amount of data available will influence the way in which risk will be perceived. The more data that was available during model construction, the higher the risk prediction. This was the case for some seventy- percent of the models. By improving the accuracy of site investigation methodology, and by adhering to robust assessment and modelling practices, a higher level of quality assurance can be achieved in the risk assessment and remediation of contaminating landfill sites. If the improvements and recommendations presented in this research are considered, uncertainties inherent in the site investigation could be reduced, therefore enhancing the accuracy of landfill risk assessment and remedial decisions.