Anthropogenic aerosols mask increases in US rainfall by greenhouse gases

Abstract

A comprehensive understanding of human-induced changes to rainfall is essential for water resource management and infrastructure design. However, at regional scales, existing detection and attribution studies are rarely able to conclusively identify human influence on precipitation. Here we show that anthropogenic aerosol and greenhouse gas (GHG) emissions are the primary drivers of precipitation change over the United States. GHG emissions increase mean and extreme precipitation from rain gauge measurements across all seasons, while the decadal-scale effect of global aerosol emissions decreases precipitation. Local aerosol emissions further offset GHG increases in the winter and spring but enhance rainfall during the summer and fall. Our results show that the conflicting literature on historical precipitation trends can be explained by offsetting aerosol and greenhouse gas signals. At the scale of the United States, individual climate models reproduce observed changes but cannot confidently determine whether a given anthropogenic agent has increased or decreased rainfall.

Fig. 1. Fraction of the contiguous United States (CONUS) with a significant attribution conclusion for the slow and fast precipitation response in each season across spatial scales.

a, b show results for precipitation rate and 20-year return values, respectively. Conclusions are based on null hypothesis tests of no effect for the fast and slow response, and we show results for successively subdividing the CONUS into one, two, four, 13, or 75 regions as well as 0.25° × 0.25° grid boxes. Testing individual subregions or grid boxes accounts for the effect of internal variability, and we include a multiple testing adjustment to yield statistical significance with both moderate and strong significance (see “Methods”).

Fig. 2. Grid-box attribution for the fast aerosol (Fast-AER), slow aerosol (Slow-AER), and greenhouse gas (Slow-GHG) precipitation responses.

a, b show results for seasonal precipitation rate and 20-year return values, respectively. Hatching indicates where we can attribute a statistically significant human influence, with either moderate (− hatching) or strong (+ hatching) significance. Each subpanel shows the fraction of grid boxes with conclusive attribution at either moderate or strong significance (strong significance only). 20-year return values are calculated as the $1-\frac{1}{20}$ quantile of the fitted generalized extreme value distribution .

Fig. 3. In situ-based, United States-average trajectories of seasonal mean (top) and extreme (bottom) precipitation anomalies from a pre-industrial climate for isolated forcing agents and the combined anthropogenic (ANT) response.

The combined ANT response is the sum of three anthropogenic agents: the slow response from greenhouse gases (GHG; red), the slow response from aerosols (AER-glob; blue), and the fast response from aerosols (AER-local; green). Each trajectory includes a 90% bootstrap confidence band. Dashed vertical lines denote the year of emergence for the isolated GHG signal (red) and combined ANT response (black), defined as the first year in which the 90% confidence band departs from zero and does not return to zero by 2020.

Fig. 4. Differences in emergence times for the combined anthropogenic response (ANT) versus the isolated greenhouse gas (GHG) signal across spatial scales.

a, b show results for seasonal precipitation rate and 20-year return values, respectively. The plotted color represents the difference in the year of ANT emergence minus the year of GHG emergence; green colors indicate masking from local aerosols while pink and purple colors indicate anti-masking from local aerosols. For cases where only the ANT signal emerges, we show the year in which the ANT signal emerges in black and white text.

Fig. 5. Comparison of the United States-average effect of each forcing agent on precipitation for historical Coupled Model Intercomparison Project Phase 6 (CMIP6) simulations and rain gauge measurements from the Global Historical Climate Network.

a, b show results for seasonal precipitation rate and 20-year return values, respectively, and we show both individual CMIP6-historical ensemble members and the ensemble mean. Estimates compare maximum versus minimum levels of each forcing agent (see "Methods"), and the y-axis shows the fraction of the multi-model ensemble that is consistent with each value on the x-axis. CMIP6 estimates involve weights based on the internal consistency of each global climate model (see "Methods"), and error bars overlaid on each density summarize the central 90% and median area under each curve.