Validation metrics for ice edge position forecasts

摘要: Abstract. The ice edge is a simple quantity in the form of line that can be derived from spatially varying sea concentration field. Due to its long history and relevance for operations in the Arctic, position should an essential element any system that is designed monitor or provide forecasts physical state Arctic Ocean and adjacent ocean regions. Users monitoring forecast products must provided with complementary information on expected accuracy data or model results. Such information traditionally available as set metrics provide an assessment quality. In this study we survey that are presently included product quality Copernicus Marine Environment Monitoring Service (CMEMS) Arctic Marine Forecasting Center forecast. We show when results from different are compared, mismatching polynya local freezing at the coasts continents archipelagos have large impact assessment. Such situations, which occur regularly examine, not been properly acknowledged sets constructed. examine using total 15 edge position. These analysed synthetic examples, well selected cases of actual forecasts, finally full year weekly bulletins. Using necessity and simplicity guideline, recommend four that sheds light various aspects consider. Moreover, user interested limited part geographical domain, so domain-wide integrated quantities may value. Consequently, also made appropriate of sub-domains. Furthermore, find decorrelation timescales are much longer than present range. Hence, our final recommendation include depictions gridded positions maps ice edge error.