Continuum‐based teleconnection indices of United States wintertime temperature variability.

It is well‐documented that United States (US) wintertime temperature variability is related to teleconnection patterns that are commonly extracted using Empirical Orthogonal Function (EOF) analysis. However, recent work suggests that EOF patterns are not necessarily physical or strongly related to temperature variability at a specified location. In this paper, US wintertime temperature variability is understood using an alternative continuum approach to teleconnection patterns, contrasting with many previous studies that adopt an EOF‐based approach. Using the continuum approach, two patterns resembling the East Pacific‐North Pacific pattern are identified that are more strongly related to US wintertime temperature than common EOF‐based patterns. Using a simple correlation method and a teleconnection impact index, 13 teleconnection indices were ranked based on how much US wintertime temperature variability they can explain. The ranking methodologies revealed that three continuum‐based dipole patterns are the leading modes of US wintertime temperature variability and that common EOF‐based patterns like the Arctic Oscillation are relatively weak temperature influencers. It is also shown that the dipole patterns are related to a tropical sea surface temperature pattern termed the pan‐Niño (PN) pattern whose associated index was created using the continuum approach. The PN pattern was found to have moderate to strong relationships with the dipole patterns and US wintertime temperature, suggesting that tropical processes have contributed to recent wintertime temperature variability. These results have important implications for research focusing on seasonal forecasting, recent cooling trends across the eastern US, and the impact of Arctic warming on US winter weather. [ABSTRACT FROM AUTHOR]