New validation method for hydrodynamic fjord models applied in the Hardangerfjord, Norway. (5th November 2020)

• Record Type: Journal Article
• Authors: Dalsøren, Stig B.; Albretsen, Jon; Asplin, Lars
• Is Part Of: Estuarine, coastal and shelf science. Volume 246(2020)
• Publication Date: 2020-11-05
• Subjects: Hydrodynamic model -- Current measurements -- Validation method -- Fjords -- Hardangerfjord -- Norway; Estuarine oceanography -- Periodicals; Coasts -- Periodicals; Estuarine biology -- Periodicals; Seashore biology -- Periodicals; Coasts; Estuarine biology; Estuarine oceanography; Seashore biology; Periodicals; 551.461805
• DOI: 10.1016/j.ecss.2020.107028
• Languages: English
• ISSNs: 0272-7714

Abstract

Abstract: We introduce a new intuitive evaluation method for comparison of fjord model results and current measurements. The approach is tested using high resolution model simulations and measurements in the Hardangerfjord, a large fjord system in Norway with huge aquaculture production. The method is easy to interpret, clearly distinguishes periods with good and poor model performance, and relate them to physical driving forces. This makes it possible to identify potential shortcomings in the models' representation of physical processes. The applied model mostly performs well in the Hardangerfjord. Good performance often coincides with strong local fjord forcing (i.e. strong winds in the fjord). In periods with poor model performance, internal waves induced by pressure perturbations on the coastal shelf tend to propagate erroneously into the fjord. Stratification biases in coastal waters, connected to the applied model boundary conditions, seems to be an important cause. Demonstrated flexibility of time frame and performance criteria suggests applicability of the validation method for a wide set of geophysical variables in various physical environments. Graphical abstract: Image 1 Highlights: New evaluation method for comparison of fjord models and current measurements. Applied for high resolution simulations and measurements in the Hardangerfjord. Model performs well >60% of the time for the seasons and depths studied.