Sensitivity of boundary data in a shallow prairie lake model

Abstract

A good water quality model needs sufficient data to characterise the waterbody, yet monitoring resources are often limited. Inadequate boundary data often contribute to model uncertainty and error. In these situations, the same water quality model can also be used to determine where sampling efforts are best concentrated for improving model reliability. A sensitivity analysis using a one-at-a-time approach on a shallow, eutrophic, Prairie reservoir model investigates which boundary conditions are contributing most to variability in the model. The model results show the lake model has greater sensitivity to its catchment processes than to its in-lake processes. Flows are shown to have the greatest influence on model predictions for all water quality variables tested, followed by air temperature. The lake is facing pressure from climate change, and water management decisions. Results indicate defining the water balance accurately will be a crucial factor in future monitoring programs and modelling efforts.