Circulation patterns associated with trends in summer temperature variability patterns in North America

Abstract

This study improves the understanding of circulation patterns associated with regional temperature trends by characterizing boreal summer temperature variability patterns in North America using rotated S-mode principal component analysis. We analyzed gridded observational 2-m temperature datasets and the ERA5 reanalysis temperature dataset to examine the climate patterns associated with long-term trends and inter-annual variability of temperature variability patterns in North America. Our analysis revealed significant trends among some classified temperature variability patterns from 1979 to 2022 summers, with inter-annual amplitudes (i.e., a departure from the mean state) signaling toward the warm regime. The anticyclonic circulation anomaly over the temperature coherent regions associated with Greenland/northeastern Canada, and Alaska, respectively, is linked to an increase in warm air advection and above-average temperatures, while cyclonic circulation over the northeast Pacific coast enhanced warm air advection and temperature increases in the coherent region comprising the northwestern portion of North America. The increase in global mean land and ocean temperatures is strongly associated with the long-term increase in the amplitude of atmospheric circulations associated with warm regimes in parts of North America. At the interannual time scale, temperature increase over Greenland/northeastern Canada is strongly associated with the negative phase of the Arctic Oscillation. These findings highlight the modulating effects of global temperature increase and warming of the western tropical Pacific Ocean on the increasing amplitude of circulations associated with warm regimes in North America. Our results further indicate that the enhancement of anticyclonic circulations over the Arctic contributes to nearly 68% of the observed reduction in sea ice extent.

Figure 1

Temperature variability patterns in North America during JJA, from 1979 to 2022. The first row is the first five PCs from CPC 2-m temperature; the second row is the first five PCs from GHCN 2-m temperature; and the third row is the first five PCs from ERA5 2-m temperature.

Figure 2

Composite anomaly maps of SLP (green contours), wind vector (black vectors), and atmospheric layer thickness between 1000 and 850 hPa (color) for the positive (a) and negative (b) phases of the temperature variability patterns in Fig. 1. Only values exceeding the 95-percentile confidence limit based on the permutation test are plotted Thick (dashed) contours indicate positive (negative) SLP anomalies.

Figure 3

Composite anomaly maps of 2-m temperature over land and oceans, for the positive and negative phases of the temperature variability patterns in Fig. 1. Only values exceeding the 95-percentile confidence limit based on the permutation test are plotted.

Figure 4

Time series of the JJA temperature variability patterns in Fig. 1 from 1979 to 2022. Time series is computed as the annual mean of the PC scores. A change point detection test that detects significant shifts in the mean of the time series was conducted with several methods in the trend package (R studio) [62]. Only statistically significant results at a 95% confidence level are reported (dashed vertical lines).

Figure 5

Five-year moving average of PC1 time series and time series of climate indices correlated with PC1, with Kendall correlation coefficient at least ~ 0.5 (top panel); and the detrended component of the PC1 and the AO index (bottom panel). PC1 was multiplied by minus 1 to keep it in phase with the AO index.

Figure 6

Seasonal climatology of temperature based on the 1992–2022 climatology.