Motivated by public health concerns, several mid-latitude nations have regulated emissions of aerosols and their precursors over the last several decades, with continued decreases projected worldwide under most emission scenarios for the coming decades.  While increasing rates of global aerosol emissions during the mid-to-late 20^th century have garnered attention for masking warming that otherwise would have occurred in response to rising greenhouse gases, open questions surround the extent to which regional temperature and precipitation responses depend on local versus remote aerosol emissions.  Our work under this project has begun to close this knowledge gap by documenting climate responses in three different Earth System models, and identifying significant responses that are robust across all models.  In particular, we are developing a set of aerosol emission-climate response relationships that can be used to emulate the climatic responses to regional emission scenarios without requiring the computationally expensive fully coupled Earth System model simulations that we used to develop these emulators.

Our approach utilizes three U.S. Earth System (chemistry-climate) models (GFDL-CM3, GISS-E2, and NCAR-CESM1) that fully couple the atmosphere to ocean, sea ice, and land models, and include full tropospheric and stratospheric chemistry, as well as aerosol-radiation and aerosol-cloud interactions. Relative to a long present-day control simulation of up to 400 years with perpetual year 2000 or 2005 emissions, we conducted a set of ~200 year idealized regional emission perturbation simulations in which aerosols (black carbon, organic carbon) and/or an aerosol precursor gas (sulfur dioxide) are reduced, for a total of 14 separate regional emission perturbations. For example, we set U.S. sulfur dioxide emissions to zero and imposed similar magnitude emission reductions in Europe, China, and India.  The long control simulation enabled us to assess statistical significance of the changes induced by the aerosol emission reductions relative to internal climate variability. We conducted a parallel set of control and emission perturbation simulations in atmosphere-only versions of each model, with fixed sea surface temperature and sea ice, to quantify effective radiative forcing (ERF) that includes the rapid adjustments of the atmosphere and land surface to each perturbation.   

We applied this novel set of simulations to develop multi-model estimates of regional temperature responses in different latitude bands to the regional aerosol emissions reductions and associated ERF.  Our new estimates provide a measure of model structural uncertainty that can now be incorporated into integrated assessment models which to date have relied on estimates from a single model.  A consistent correlation emerges across the models between global annual mean surface temperature change and the global ERF.  The most robust regional response across our simulations is a strong Arctic warming, underscoring the sensitivity of this region to remote aerosol emissions.  We explore the specific mechanisms at play in the remote Arctic response by decomposing the total Arctic warming into contributions from ERF, climate feedbacks, and changes in atmospheric heat transport.  Our results show that feedbacks associated with changes in tropospheric temperature and surface albedo explain most of the difference in Arctic warming across the simulations.  An examination of climate extremes reveals that local aerosol emission reductions increase the hottest day each year within the source region by up to one degree Celsius.  In two of the three models, we find that the climate sensitivity to regional aerosol emission perturbations can exceed that diagnosed from doubled carbon dioxide simulations, with an overall range across the aerosol perturbation simulations of 0.5-1.0 K (W m^-2)^-1.

In each of our simulations, we find that aerosol emission reductions increase precipitation near the source region.  Our analysis reveals that remote precipitation responses in the tropical Pacific, Amazonia and South Asia may be modulated by the El Ni?o-Southern Oscillation (ENSO), which can explain up to 20% of the total precipitation response to regional aerosol emission perturbations.  Earlier work has partly attributed the observed drying of the Sahel region of northern Africa from the 1950s-1980s to the concurrent rise in global aerosol emissions.  We find that Sahelian precipitation responds robustly to multiple regional aerosol emission perturbations, with the largest responses occurring when aerosol emissions are reduced within the higher latitude regions of Europe or the United States. We demonstrate that the ratio of the warming in the northern hemisphere to that in the southern hemisphere induced by the removal of regional aerosol emissions can explain at least half of the total variance in boreal summer precipitation response in the Sahel across the simulations. Our findings suggest that efforts in recent decades to clean the air in the U.S.A. and European nations may have had unexpected benefits to the remote Sahel region by increasing precipitation and thus ameliorating drought conditions.

Last Modified: 10/25/2020
Modified by: Arlene M Fiore