Analyzing the Pacific-North American Teleconnection and its Relationship to Present and Future Climate in the CMIP5 Models

We assess the presence of the Pacific-North American (PNA) teleconnection pattern in 27 atmosphere-ocean general circulation models (AOGCMs) and earth system models (ESMs) from the fifth phase of the Coupled Model Intercomparison Project (CMIP5) under the Historical run and the RCP8.5 future climate projection. We use the National Centers for Environmental Prediction and Atmospheric Research (NCEP/NCAR) Reanalysis as a quasi-observational baseline in the evaluations of the CMIP5 climate models. We assess the PNA teleconnection both temporally and spatially, using a rotated principle component analysis (RPCA) to extract the PNA index time series and identify the spatial loading patterns. Each model's performance is assessed visually and by employing a series of calculations to quantify the model's simulation of the PNA under the future climate projection. We compare these results to the historical simulation for each model and to NCEP in order to understand how the PNA evolves throughout the 21st century. We also analyze the relationship between the PNA and surface air temperature and precipitation. The CMIP5 models vary substantially in their ability to simulate different features of the PNA teleconnection. Most models are unable to resolve the temporal variability of NCEP for both future time periods; however, most of the models capture the PNA as a low-frequency oscillation, albeit with slightly varying periods. Spatial variability among the models is present as well, with few models agreeing on the nominal appearance of the PNA pattern in the future. It is also evident that many models lack the ability to simulate the barotropic instability that initiates wave energy propagation through the 500-hPa geopotential height field, leading to a phase-locking and thus indistinguishable positive and negative modes of the PNA. The behavior of the models relative to NCEP is more conclusive than the trends within the models. The CMIP5 models consistently underestimate the strength of the anomalies associated with the three "centers of action" that define the PNA pattern. Both the spatial and temporal discrepancies that arise among the models result in inconsistent relationships among the PNA and surface climate variables. Based on a comprehensive analysis, only a few models stand out in their ability to consistently represent all aspects of the PNA teleconnection.