b'Risk-based modelling of phosphorus pollution in IrishAgricultural CatchmentsGlendell M1 and Mellander P-E2,31The James Hutton Institute, Environmental and Biochemical Sciences Group, Aberdeen, Scotland, UK2Agricultural Catchments Programme, Teagasc, Johnstown Castle Environment Research Centre, Wexford, Co. Wexford, Ireland3Crops, Environment and Land Use Programme, Teagasc, Johnstown Castle Environment Research Centre, Wexford, Co. Wexford, IrelandPhosphorus (P) pollution remains a major cause of surface water quality failures. Abating P pollution in agricultural catchments requires informed decisions about the likely effectiveness of land management mitigation measures and their spatial targeting. Furthermore, its important to balance the environmental benefits of mitigation measures with their potential impact on farm productivity and profitability. Therefore, user-friendly transparent decision supports tools are required that allow the integration of uncertain information on potential effects and outcomes, whilst accounting for the uncertainty in the understanding of both the model structure and the data.Bayesian Belief Networks (BBNs) are probabilistic graphical models that allow the integration of both quantitative and qualitative information from a range of sources (including data, other model outputs and non-scientific knowledge, such as expert opinion) in one model. BBNs allow system-level thinking, revealing possible causal relationships between controlling factors that may not be apparent otherwise and in situations where controlled experiments are not possible, such as diverse river catchments.Here we present a spatial BBN to facilitate the understanding of the effects of land use on Ppollution risk in Irish ACP catchments. The PhosphoRisk decision support tool facilitates the co-construction of the modelling outcomes by the academic and the stakeholder communities. The modelled scenarios will help to inform targeting of water quality mitigation measures in high risk areas, while the quantified model uncertainties will inform further research and datacollection. The tool will provide a user-friendly interface with clear visual outputs that can beeasily updated as new data and understanding become available.32'