Reassessing the relative role of anthropogenic aerosols and natural decadal variability in driving the mid-twentieth century global "cooling": a focus on the latitudinal gradient of tropospheric temperature.

The global mean surface temperature cooled slightly in the mid-twentieth century despite a continuous increase in greenhouse gas concentrations. The cooling was strongest in the Northern Hemisphere mid-latitudes, while the Southern Hemisphere mid-latitudes experienced moderate warming. This apparent contradiction is often attributed to internal multi-decadal variability originating from Pacific and Atlantic ocean-atmosphere interactions. Given the rapid increase of industrial activities in North America and Europe during that period, it is also plausible that anthropogenic aerosol (AA) emissions (as an external forcing) contributed to the stronger Northern Hemisphere cooling. This paper aims to quantify the contributions of AA and decadal variability to the 1948–1978 cooling. We analyzed the latitudinal temperature trend asymmetry in 60° S–60° N throughout the troposphere, using multiple reanalysis datasets and the Coupled Model Intercomparison Project phase 5 (CMIP5) multi-model ensemble that bears significant similarity with the observed patterns. We show that both AA increase and the North Atlantic Variability Index (NAVI) transition into its negative phase are the major contributors to the latitudinal asymmetry of cooling. At the surface level, based on the horizontal pattern correlation method, AA and NAVI have similar contribution fractions (20 vs. 16%), but the contribution fraction of AA is much larger at 500 hPa (55 vs. 8%). Attributions based on vertical pattern correlation and latitudinal gradient show consistent results. Natural forcings (NAT) also contribute to the cooling asymmetry during mid-20C, but with a much smaller impact compared to AA and NAVI. Therefore, we argue that previous studies that mostly focused on surface variables may have underestimated the role of AA in the mid-twentieth-century climate change. The study suggests that the three-dimensional thermal structure and atmospheric circulation change should be closely examined in future climate attribution analysis. [ABSTRACT FROM AUTHOR]