NSFGEO-NERC Advancing Predictability of Sea Ice: Phase 2 of the Sea Ice Prediction Network (SIPN2)

The shrinking Arctic sea-ice cover has captured the attention of the world. The downward September
trend has accelerated over the last decade, with the 10 lowest September sea-ice extents occurring in the
last 10 years. An essentially ice-free Arctic during summer is expected by mid-century. Loss of the seaice
cover has profound consequences for ecosystems and human activities in the Arctic, so there is an
urgent need to advance sea-ice predictions in all seasons at both the pan-Arctic and regional scales. A key
finding that emerged from the Sea Ice Prediction Network (SIPN) effort is that predictions of September
sea-ice extent tend to have less skill in extreme years that strongly depart from the trend line. The
objective of proposed research under Phase 2 of SIPN (SIPN2) is to improve forecast skill through
adopting a multi-disciplinary approach that includes modeling, new products, data analysis, scientific
networks, and stakeholder engagement. Specifically, the team will: (1) Investigate the sensitivity of
subseasonal-to-seasonal sea-ice predictability in the Alaska Arctic to variations in oceanic heat and largescale
atmospheric forcing using a dynamical model (NCAR CESM) and statistical forecasting tools,
focusing on spatial fields in addition to total extent summaries; (2) Assess the accuracy of Sea Ice
Outlook (SIO) submissions based on methodology and initialization; (3) Develop new observation-based
products for improving sea-ice predictions, including sea-ice thickness, surface roughness, melt ponds,
and snow depth; (4) Evaluate the socio-economic value of sea-ice forecasts to stakeholders who manage
ship traffic and coastal village resupply in the Alaska Sector, and engage the public in Arctic climate and
sea-ice prediction through blog exchanges, accessible SIO reports, bi-monthly webinars, and by making
public data sources useful to non-scientists and scientists alike; and (5) Continue and evolve network
activities to generate SIO forecasts and reporting for September minima as in SIPN and expand SIPN2
forecasts to include full spatial resolution and emerging ice-anomaly-months (October/November).

A better quantification of the role of oceanic heat and climate variations in the Pacific sector, new observational-based sea-ice products, and network activities will advance understanding of seasonal predictability of Arctic sea ice, the limits of this predictability, and the economic value of forecasts for stakeholders. The network will examine origins and impacts of extreme ocean surface warming in preconditioning the ice cover in the Pacific Arctic for continued major reductions in sea-ice extent and duration.

This work will directly engage stakeholders that create and use sea-ice forecasts in Alaska and lead to improved safety around sea ice. Work under SIPN2 will also track public awareness and perceptions regarding sea ice, helping to raise understanding through accessible reports, discussions, and public data sources useful to non-scientists and scientists alike. Stakeholder engagement during the research process will potentially facilitate rapid research-to-operations implementation of the products of this work. Ongoing efforts by the SEARCH sea-ice action team and Sea Ice for Walrus Outlook (SIWO) will be leveraged to provide additional understanding of stakeholder needs.

Increasing uncertainty about future sea ice conditions presents a distinct challenge to industry, policymakers, and planners responsible for economic, safety, and risk mitigation decisions. The ability to accurately forecast the extent and duration of Arctic sea ice on different timescales provides significant implications for the operation of wide ranging Arctic maritime activities.
The Arctic sea ice prediction community has advanced rapidly in the past decade with many new sea ice forecast products and services that are targeted for different user groups. However, it is still unclear how well end users are able to utilize these products and services into their planning. There is a need for better engagement with a broad range of Arctic stakeholders and a need to tailor new products and services to end user-specific requirements. Towards this end a workshop is being hosted at Arctic Frontiers in Tromso (January 2018) by PI Stroeve, in collaboration with CliC, EU Polar Net and University of Bergen to bring together sea ice stakeholders and forecasters to:
1) Assess the value of forecasts by the user community.
2) Determine if and how ice forecasts are currently being used in decision making.
2) Communicate the relevant metrics needed by various stakeholders.
3) Identify where improvements in sea ice forecasts would help stakeholders make decisions.
4) Communicate the limits and opportunities of current forecasting systems.