implications for ship navigation). Like Metcalf, he also noted the critical importance of wind and ocean currents in driving ice conditions.
During a discussion on key gaps and challenges in observations for future sea ice prediction, Hajo Eicken of the University of Alaska Fairbanks noted a number of gaps in our current understanding of sea ice. These observational challenges include predicting the seasonal decay of ice and heat fluxes over decaying ice. He indicated that when interacting with the modeling community, the decadal timescale is very important, and snow on sea ice is a particularly relevant factor on this scale. He also mentioned that it is important to accurately define the questions that stakeholders would like answered (while also acknowledging that different stakeholders have different needs). He indicated that a path forward may be to define the sea ice services that stakeholders require, translate those needs into specific prognostic variables, and determine the predictive success that is acceptable to the stakeholders.
Walt Meier of the National Snow and Ice Data Center also emphasized the importance of filling gaps in observations, although he pointed out a number of products that are already available for ice extent and concentration, ice thickness, and ice motion. Some current limitations include quantitative error estimates, the harmonization of spatial and temporal scales, melt state and albedo, snow depth, information on ice deformation, in situ and airborne measurements, integrated products, and data access. He also mentioned the importance of continuity and contingency plans in satellite missions (to avoid large gaps in data availability). Ron Lindsay of the University of Washington discussed the importance of in-depth conversations between stakeholders and researchers. This is essential, not only to be sure that we are getting the observations that we actually need, but also to determine where some possible improvements in skill would most help the stakeholders. He noted that, using this information, we can begin to focus on those problems that we can more readily solve and would also be most helpful (instead of using this time on questions that have a limited likelihood of being solved, or on issues that are of limited importance to stakeholders).
Following the panel discussion, members of the breakout groups convened to address additional challenges and strategies associated with sea ice observations. Breakout group rapporteurs mentioned that there is a wide range of needs on many spatial and temporal scales, and that key parameters should be clearly defined depending on stakeholder needs. One group suggested that there are linkages between the need for specific ice parameters and broader scale questions such as: Is there ice? What is it like? Where is it going? This helps to drive observational needs for defining ice extent, ice character, and ice motion. Other issues that were discussed in the breakout groups include quantifying uncertainty, assimilating observations into models, improving bathymetry data, the