Modelled and observed sea-spray icing in Arctic-Norwegian waters

• Published in: Cold Regions Science and Technology
• Main Authors: Samuelsen, Eirik Mikal, Edvardsen, Kåre, Graversen, Rune
• Format: Text
• Language: English
• Published: Elsevier 2016
• Subjects: VDP::Matematikk og naturvitenskap: 400::Fysikk: 430; VDP::Mathematics and natural scienses: 400::Physics: 430; VDP::Teknologi: 500; VDP::Technology: 500; Arktisk teknologi / Arctic Technology; Meteorologi / Meteorology; Polarforskning / Polar research; Arctic; Arktis*; Nordkapp
• Online Access: https://hdl.handle.net/10037/10503; https://doi.org/10.1016/j.coldregions.2016.11.002

Link to publishers version: http://dx.doi.org/10.1016/j.coldregions.2016.11.002 Hazardous marine icing is a major concern for ships operating in Arctic waters during freezing conditions. Sea spray generated by the interaction between a ship and ocean waves is the most important water source in these dangerous icing events. Although there exist several data sets with observations of ice accretion in conjunction with meteorological and oceanographic parameters, these data sets often have shortcomings and only a few are obtained in Arctic-Norwegian waters. In this study, icing rates from a large coast-guard vessel type, the KV Nordkapp class, are used for verification of a newly proposed Marine-Icing Model for the Norwegian COast Guard (MINCOG). Ship observations, NOrwegian ReAnalysis 10km data (NORA10), and wave data based on empirical statistical relationships between wind and waves are all applied in MINCOG and the results are compared. The model includes two different empirically-derived formulations of spray flux. It is found that in general the best results for different verification scores are obtained by using a combination of observed atmosphere and ocean-wave parameters from the ships, and wave period and direction from NORA10, regardless of the spray-flux formulation applied. Furthermore, the results illuminate that wave parameters derived from formulas based on empirical relationships between the local wind speed and significant wave height and wave period, compared to those obtained from observations or NORA10, considerably worsen icing-rate predictions in Arctic-Norwegian waters when applied in MINCOG.