Detectability of Subpolar North Atlantic Climate Change in the Past Century

In the past century, climate change signals have been reported in a range of variables in the North Atlantic and are significant to observe and prepare the world for the unpredictable future. Complex climate relationships complicate detection and attribution’s uncertainties such as historical forcing, internal variability, and model physics further complicate climate science research. The impacts on detection and attribution from these uncertainties remain unknown. Two research objectives are 1) to identify North Atlantic climate change signals in sea surface temperature (SST), ice areal extent, and precipitation, and 2) to quantify and decipher the uncertainties in model simulations of historical climate change signals in the Subpolar North Atlantic. To address these challenges, CESM-LE (Community Earth System Model – Large Ensemble) with CMIP5 forcing (Coupled Model Intercomparison Project 5), CESM2-LE (Community Earth System Model 2 – Large Ensemble) with CMIP6 forcing, and CESM2-CMIP5 under the Historical simulation were utilized. The linear trend of SST, ice areal extent and precipitation, area-weighted over the subpolar North Atlantic, and the maximum and minimum trend as simulated by ensemble members were calculated to quantify uncertainties for each climate variable. Results showed significant North Atlantic climate change signals in SST, ice areal extent, and precipitation. An increase was present in signals in SST and precipitation, and shrinkage was present in signals in ice areal extent over time. Large uncertainties in North Atlantic SST and ice areal extent were caused by internal variability, differences in model physics, and uncertainties in historical forcing, while uncertainties in precipitation mainly comes from internal variability. An effect on the attribution and detection of past climates and the prediction of future climates seems to exist due to the large uncertainties.