LAUSR

The predictive skill of summer sea ice concentration (SIC) in the Arctic presents a steep decline when initialized before June, which is the so‐called spring predictability barrier for Arctic sea ice. This study explores the potential influence of surface heat flux, cloud and water vapor anomalies on monthly to seasonal predictions of Arctic SIC anomalies. The results show an enhancement in skill predicting Arctic September SIC in the models that use surface fluxes, clouds, or water vapor in combination with SIC and surface sea temperature as predictors when initialized in boreal spring. This result shows the potential to reduce the spring barrier for Arctic SIC predictions by including the surface heat budget. The enhanced predictive skill can be very likely linked to the improved representation of the thermodynamics associated with water vapor and cloudiness anomalies in spring.