“Knowledge Strength”: maximising the reliability of evidence derived from environmental modelling in the face of uncertainty – the case of the salmon louse (<i>Lepeophtheirus salmonis</i>)

Policy developments for a sustainable Blue Economy require scientific advice. An important component of the Blue Economy is salmon (Salmo salar) aquaculture, the sustainability of which is limited by salmon louse (Lepeophtheirus salmonis) infection. This parasite impacts both farmed and wild salmonid fish. Modelling is a valuable source for advice, but inevitable uncertainties exist. Here we develop an approach we call “Knowledge Strength” to maximise our confidence in model results; this is aimed at reducing uncertainties in model outputs and understanding the remaining uncertainty in these outputs, to maximise our confidence in model results, so that we can give policy makers the best advice to support informed decision making. The approach consists of addressing five questions: (1) What is the objective addressed by the model? (2) What are the causes of uncertainty in model outputs? We describe uncertainties due to (a) limitations of computing, (b) model building and (c) parameters, and (d) forcing data. (3) What is the statistical nature of uncertainty? Noise and bias are qualitatively different. (4) How can knowledge strength be maximised given those uncertainties? Approaches of resourcing (“power”) and analysis (“wisdom”) are considered. (5) How can information, including uncertainties, be communicated to different audiences? Policy makers/managers define and resource the objective for question 1, modellers address questions 2 to 4 - but their solutions are made transparent, and the communication question 5 is a two-way process with outputs transparent to immediate decision makers and external stakeholders. Examples of the policy environment behind salmon lice management are detailed in the Supplementary Material covering Scotland, Norway and the Faroe Islands.