On using numerical sea-ice prediction and indigenous observations to improve operational sea-ice forecasts during spring in the Bering Sea

Abstract

Impacts of a rapidly changing climate are amplified in the Arctic. The most notorious change has come in the form of record-breaking summertime sea-ice retreat. Larger areas of open water and a prolonged ice-free season create opportunity for some industries, but bring new challenges to indigenous populations that rely on sea-ice cover for subsistence. Observed and projected increases in Arctic maritime activities require accurate sea-ice forecasts on the weather timescale, which are currently lacking. Motivated by emerging needs, this study explores how new modeling developments and local-scale observations can contribute to improving sea-ice forecasts. The Arctic Cap Nowcast/Forecast System, a research sea-ice forecast model developed by the U.S. Navy, is evaluated for forecast skill. Forecasts of ice concentration, thickness, and drift speed produced by the model from April through June 2011 in the Bering Sea have been investigated to determine how the model performs relative to persistence and climatology. Results show that model forecasts can outperform forecasts based on climatology or persistence. However, predictive skill is less consistent during powerful, synoptic-scale events and near the Bering Slope. Forecast case studies in Western Alaska are presented. Community-based observations from recognized indigenous sea-ice experts have been analyzed to gauge the prospect of using local observations in the operational sea-ice monitoring and prediction process. Local observations are discussed in the context of cross-validating model guidance, data sources used in operational ice monitoring, and public sea-ice information products issued by the U.S. National Weather Service. Instrumentation for observing sea-ice and weather at the local scale was supplied to key observers. The instrumentation shows utility in the field and may help translate the context of indigenous observations and provide ground-truth data for use by forecasters.