Seasonal Predictability of Global and North American Coastal Sea Surface Temperature and Height Anomalies

Abstract

A Linear Inverse Model (LIM) is constructed to evaluate predictability of seasonal sea surface temperature (SST) and sea surface height (SSH) anomalies over the ice-free global ocean. Its ensemble-mean hindcast skill is also compared to that of the North American Multi-Model Ensemble (NMME) for 1982–2010. Both have similar skill for dominant modes of SST variability, but regional NMME SST skill is somewhat higher in many locations. However, the LIM has considerably more Atlantic and Southern Ocean SSH skill. Skill is generally comparable along the North American coastline, but LIM skill is greater for several highly productive coastal zones and East Coast tide gauge stations. Diverse, often predictable ENSO events drive teleconnections providing predictability in the North Pacific and along the US West Coast. Predictability in the Atlantic and along the US East Coast is associated with Gulf Stream strength modulation. Overall, the LIM shows potential for seasonal prediction of coastal ocean conditions.