Statistical Dimensioning of Nutrient Loading Reduction: LLR Assessment Tool for Lake Managers

• Published in: Environmental management (New York) Vol. 56; no. 2; pp. 480 - 491
• Main Authors: Kotamäki, Niina, Pätynen, Anita, Taskinen, Antti, Huttula, Timo, Malve, Olli
• Format: Journal Article
• Language: English
• Published: New York Springer US 01.08.2015; Springer Nature B.V
• Subjects: Aquatic Pollution; Assessments; Atmospheric Protection/Air Quality Control/Air Pollution; Bayes Theorem; Bayesian analysis; Biogeochemistry; Chlorophyll; Chlorophyll - analysis; Chlorophyll A; Chlorophylls; Coastal waters; Datasets; Decision making; Earth and Environmental Science; Ecology; Environment; Environmental Management; Environmental Monitoring - methods; Environmental Monitoring - statistics & numerical data; Eutrophication; Finland; Forestry Management; Freshwater; Lakes; Lakes - chemistry; Management planning; Mathematical models; Models, Theoretical; Nature Conservation; Nutrient concentrations; Nutrient loading; Nutrient retention; Nutrients; Phosphorus; Reduction; Retention; River basin management; River basins; Statistical models; Uncertainty; Waste Water Technology; Water Management; Water Pollution Control; Water Quality; Water quality assessments; ANE, Finland; Uncertainty; Water Framework Directive; Lake water quality modeling; Bayesian inference; Hierarchical modeling; River basin management
• ISSN: 0364-152X; 1432-1009
• DOI: 10.1007/s00267-015-0514-0

Implementation of the EU Water Framework Directive (WFD) has set a great challenge on river basin management planning. Assessing the water quality of lakes and coastal waters as well as setting the accepted nutrient loading levels requires appropriate decision supporting tools and models. Uncertainty that is inevitably related to the assessment results and rises from several sources calls for more precise quantification and consideration. In this study, we present a modeling tool, called lake load response (LLR), which can be used for statistical dimensioning of the nutrient loading reduction. LLR calculates the reduction that is needed to achieve good ecological status in a lake in terms of total nutrients and chlorophyll a (chl- a ) concentration. We show that by combining an empirical nutrient retention model with a hierarchical chl- a model, the national lake monitoring data can be used more efficiently for predictions to a single lake. To estimate the uncertainties, we separate the residual variability and the parameter uncertainty of the modeling results with the probabilistic Bayesian modeling framework. LLR has been developed to answer the urgent need for fast and simple assessment methods, especially when implementing WFD at such an extensive scale as in Finland. With a case study for an eutrophic Finnish lake, we demonstrate how the model can be utilized to set the target loadings and to see how the uncertainties are quantified and how they are accumulating within the modeling chain.