Arsenic contamination of drinking water in Ireland: A spatial analysis of occurrence and potential risk.

The presence of arsenic in groundwater has become a global concern due to the health risks from drinking water with elevated concentrations. The Water Framework Directive (WFD) of the European Union calls for drinking water risk assessment for member states. The present study amalgamates readily available national and sub-national scale datasets on arsenic in groundwater in the Republic of Ireland. However, due to the presence of high levels of left censoring (i.e. arsenic values below an analytical detection limit) and changes in detection limits over time, the application of conventional statistical methods would inhibit the generation of meaningful results. In order to handle these issues several arsenic databases were integrated and the data modelled using statistical methods appropriate for non-detect data. In addition, geostatistical methods were used to assess principal risk components of elevated arsenic related to lithology, aquifer type and groundwater vulnerability. Geographic statistical methods were used to overcome some of the geographical limitations of the Irish Environmental Protection Agency (EPA) sample database. Nearest-neighbour inverse distance weighting (IDW) and local indicator of spatial association (LISA) methods were used to estimate risk in non-sampled areas. Significant differences were also noted between different aquifer lithologies, indicating that Rhyolite, Sandstone and Shale (Greywackes), and Impure Limestone potentially presented a greater risk of elevated arsenic in groundwaters. Significant differences also occurred among aquifer types with poorly productive aquifers, locally important fractured bedrock aquifers and regionally important fissured bedrock aquifers presenting the highest potential risk of elevated arsenic. No significant differences were detected among different groundwater vulnerability groups as defined by the Geological Survey of Ireland. This research will assist management and future policy directions of groundwater resources at EU level and guide future research focused on understanding arsenic mobilisation processes to facilitate in guiding future development, testing and treatment requirements of groundwater resources.

Pesticide occurrence in groundwater and the physical characteristics in association with these detections in Ireland.

This study explores the associations of pesticide occurrence in groundwater to geological characteristics of the monitoring points (MPs) contributing area. Pesticide analyses were undertaken during a 2-year groundwater monitoring campaign which generated 845 samples. MCPA and mecoprop were the most frequently detected pesticides in groundwater. Each MP (n = 158) had a specifically delineated zone of contribution (ZOC) and the dominant physical characteristics present from nine national datasets were recorded for each ZOC. Associations between detections in groundwater and the dominant physical characteristic in each MPs ZOC tested were then statistically analyzed using Fisher's exact test, logistic regression, and multiple logistic regression. The original physical characteristic datasets used that were associated with detections in groundwater were the type of MP, aquifer type, and Quaternary deposit type. Logistic regression revealed that springs, regionally important aquifer types, aquifers with a karstic flow regime, and alkaline Quaternary deposits in existence above karst aquifers in a MP's ZOC were more likely to have a pesticide detection in groundwater. Multiple regression from this exploratory work showed some mutual dependency between soil association, aquifer type, and the Geological Survey of Ireland groundwater vulnerability map. The combination of national monitoring data and physical attribute datasets can be used to explore key areas where groundwater is more vulnerable to pesticide contamination.