Was the 2015 subpolar North Atlantic cold blob predictable?

Abstract:

Sea surface temperature (SST) in the subpolar North Atlantic reached near-record cold conditions during 2015 before rapidly warming to climatology in 2016. 2015 was also remarkable for the persistently positive conditions of North Atlantic Oscillation (NAO), which was necessary for this cold blob’s evolution. Here, we examine how predictable this extreme event was using the Community Earth System Model version 1 Decadal Prediction Large Ensemble (CESM1-DPLE), its uninitialized counterpart the CESM1 Large Ensemble (CESM1-LE), and additional sensitivity experiments. The 1-year prediction from the CESM1-DPLE reproduces the multidecadal swings in observed SST and its 40 ensemble members encompass observed SST for every year – except for the 2015 cold blob. This predictive failure occurs within 6 months of initialization and despite already cold initial conditions. Results from the CESM1-LE indicate that the unlikeliness of the observed winter 2015 NAO is at least part of the explanation for why this event was not encompassed in the CESM1-DPLE's spread. To test another possibility – that deficiencies in the atmosphere, ocean, or sea ice initial conditions degraded the prediction – additional hindcasts using the CESM1-DPLE protocol but with different initial conditions are performed. Altering the initial conditions did not improve the simulation of this event, and in some cases, degraded it. Given the difficulty of predicting this event, this case could be a useful testbed for future prediction system development.