Climate change alters teleconnections

TL;DR

The paper addresses whether anthropogenic climate change reshapes teleconnections among internal climate variability modes. It uses a covariance fingerprinting framework to project observed changes in monthly temperature covariance onto model-forced covariance changes, producing a scalar $S_{obs}$ and comparing it to a null distribution $S_{null}$. It also applies a multilinear regression approach to isolate regional contributions of GMST and five modes (ENSO, NAO, SAM, IOD, PDO), with bootstrapping to assess significance. The findings show that observed teleconnection changes between 1960–1990 and 1990–2020 cannot be explained by natural variability alone and are amplified in several regions; projections under RCP8.5 indicate further evolution, implying meaningful impacts on regional temperatures.

Abstract

Internal modes of climate variability, such as El Niño and the North Atlantic Oscillation, can have strong influences upon distant weather patterns, effects that are referred to as "teleconnections". The extent to which anthropogenic climate change has and will continue to affect these teleconnections, however, remains uncertain. Here, we employ a covariance fingerprinting approach to demonstrate that shifts in teleconnection patterns affecting monthly temperatures between the periods 1960-1990 and 1990-2020 are attributable to anthropogenic forcing. We further apply multilinear regression to assess the regional contributions and statistical significance of changes in five key climate modes: the El Niño-Southern Oscillation, North Atlantic Oscillation, Southern Annular Mode, Indian Ocean Dipole, and the Pacific Decadal Oscillation. In many regions, observed changes exceed what would be expected from natural variability alone, further implicating an anthropogenic influence. Finally, we provide projections of how these teleconnections will alter in response to further changes in climate.

Figures (4)

• Figure 1: Covariance changes across thirty-year epochs. The left column shows covariance averaged over all grid cells (top row), the Niño3.4 region (second row), the IOD region (third row), and the PDO region (bottom row) for 1960--1990. The middle column is the same as the left column, but for the 1990--2020 period, and the right column shows the difference between the two periods. Note, the PDO covariance values in the bottom row are divided by 4 to facilitate the use of the same color range as the rest of the panels.
• Figure 2: Covariance fingerprint.$S_\text{obs}$ (red line) is larger than any of 1000 realizations of $S_\text{null}$ (histogram). $S_\text{obs}$ is the observed covariance difference between 1960--1990 and 1990--2020 projected onto that produced by CESM2-LE historical runs. Each $S_\text{null}$ realization is the same as $S_\text{obs}$ but is computed using pairs of 30-year periods that are randomly drawn from the CESM2 long control run.
• Figure 3: Point-wise attribution. Differences in regression coefficients between 1960--1990 and 1990--2020 for five internal variability modes and GMST. Positive values indicate increased co-variation of monthly temperatures with the corresponding mode. Locations where the differences between the periods can be attributed to anthropogenic climate change are indicated by pink crosses.
• Figure 4: Climate change projections. Similar to Fig. 3, but for the difference in regression coefficients between 1960-1990 and 2070-2100 using simulations. Temperatures and mode indices are from CESM2-LE simulations. In this case, locations where results are statistically insignificant are indicated by gray hatching.