View of industrial exhaust rising from smokestacks in the background, with electrical transmission towers and poles in the foreground
Human activities like heavy industry and power generation emit aerosol particles into the atmosphere, where they filter incoming sunlight and influence weather and climate patterns. Despite many years of scientific research into aerosol effects in the atmosphere, large uncertainties remain. Credit: Ralf Vetterle, Pixabay license

If you have ever seen a reddish sunset through a haze of smog, dust, or smoke from a forest fire, you have witnessed the large effects that tiny aerosols can produce in the atmosphere. Scientists have learned a great deal about these diminutive particles in recent decades, but much remains to be revealed. In July 2019, about 100 scientists from 18 countries met in Beijing, China, to discuss recent advances in observing and measuring atmospheric aerosols and in our understanding of how aerosol particles interact with Earth’s climate. The group also discussed how to improve computational models of atmospheric aerosols.

Atmospheric aerosols are liquid or solid particles between 1 nanometer and 10 micrometers in diameter that are suspended in the atmosphere. They may be produced directly in the form of mineral dust from deserts, sea salt from oceans, or black carbon and other particulates from fossil fuel and biomass burning. They may also be produced indirectly through chemical reactions between precursor gases emitted into the atmosphere. For example, power plants emit sulfur dioxide, which forms sulfate aerosols, and automobiles emit nitrogen oxides, which form nitrate aerosols.

Aerosols are not passive particles; they actively affect the climate system and how we experience it.

In the atmosphere, where winds transport them laterally and vertically, aerosols participate in a myriad of complicated processes. They can act as cloud condensation nuclei, for example, or ice-nucleating particles that then serve as catalysts for cloud formation. Aerosols are eventually removed from the atmosphere when they attach themselves to Earth’s surface (a process called dry deposition) or when rain or snow washes them out of the air (processes called in-cloud and below-cloud scavenging). However, aerosols are not passive particles; they actively affect the climate system and how we experience it. For example, they reduce visibility by scattering and absorbing sunlight, and they modify the ways in which clouds interact with light and other electromagnetic waves (radiative properties) by changing the number of droplets or crystals in a cloud, the amount of water clouds contain, cloud lifetime, and precipitation.

Persistent Uncertainties

In contrast to the warming effects of greenhouse gases (GHGs), aerosols are generally considered to cast a cooling effect on Earth’s climate. Thus, aerosols have canceled or masked a significant fraction of Earth’s warming (a positive radiative flux change) attributable to increased greenhouse gas concentrations in the atmosphere since the start of the Industrial Revolution in about 1750. Despite scientists’ tremendous efforts to study aerosols and aerosol effects on climate, large uncertainties stubbornly persist in estimates of how much of the change in radiative flux has been caused by aerosols generated from human activities (i.e., anthropogenic aerosol radiative forcing).

Discrepancies in anthropogenic aerosol radiative forcing estimates among the Earth system models (ESMs) in the Intergovernmental Panel on Climate Change’s Fifth Assessment Report are responsible for the large spread of estimates—1.1 to 3.3 watts per square meter—of total anthropogenic radiative forcing (from GHGs plus aerosols) over the industrial era. This spread leads to large uncertainties in model simulations of present-day climate and projections of future climate. Therefore, there is a pressing need to reduce uncertainty in anthropogenic aerosol radiative forcing to improve the predictability of ESMs.

Working Toward a Solution

A sound approach to reduce the aerosol radiative forcing uncertainty in ESMs is to identify the key aerosol processes and properties responsible for the uncertainty and then to improve representations of these processes and properties in light of new observations. This approach was the goal of the 18th CTWF International Symposium on Aerosol and Climate Change last July, a forum (the “F” in CTWF) sponsored by the Chinese Academy of Sciences (C), The World Academy of Sciences (T), and the World Meteorological Organization (W).

Symposium participants discussed the roles of aerosols in the climate system, observations of aerosols from in situ and remote sensing platforms, and representations and modeling of aerosols in ESMs, and they examined key scientific questions and future research directions. We summarize those discussions here.

Aerosol-Cloud Interactions

Interactions between aerosols and clouds represent the largest source of uncertainty in estimates of aerosol radiative forcing on climate.

Interactions between aerosols and clouds represent the largest source of uncertainty in estimates of aerosol radiative forcing on climate. This large uncertainty has traditionally been attributed to poor simulations of the response of planetary boundary layer clouds (i.e., clouds in the lowest layer of the atmosphere) to aerosol perturbation. A recent observational analysis revealed that oceanic boundary layer clouds are even more sensitive to aerosol levels than previously reported. This finding suggests that aerosol effects might also be behind positive radiative forcing on other types of clouds (e.g., convective, mixed-phase, and cirrus clouds). Current ESMs represent these effects poorly in the case of mixed-phase and cirrus clouds, and they ignore them completely in the case of convective clouds.

For example, anthropogenic aerosol particles could suppress warm rains to allow more freezing of cloud water at subfreezing temperatures or increase condensation on more numerous cloud droplets in deep convective clouds. Either of these processes could cause larger releases of latent heat. This heat release could then intensify vertical motions in the clouds, producing larger anvil clouds with more radiative warming.

Ice-nucleating particles from naturally and anthropogenically emitted aerosols can induce warming effects by causing ice crystals to form rapidly and glaciating low-level mixed-phase clouds (i.e., clouds composed of both liquid droplets and ice crystals) or by promoting the formation of cirrus clouds. However, the mechanisms by which aerosols affect such ice-bearing, or “cold,” clouds and the resulting directions (positive or negative) and magnitudes of radiative forcing on climate are largely unknown.

Integrated approaches that combine observations and modeling are needed to tackle key aerosol uncertainties in ESMs. Satellite observations could be combined with in situ observations of changes in clouds downwind of industrial pollution (e.g., clouds over the Northwest Pacific in the path of East Asian pollution outflows), in marine clouds influenced by ship tracks, in stratocumulus clouds over the southeast Atlantic influenced by biomass burning smoke from South Africa, or in clouds influenced by volcanic plumes like those produced by Iceland’s Holuhraun eruption in 2014–2015. Comparing models with such observational analyses could help to identify shortcomings of model-simulated cloud responses to aerosols.

Aerosol Modeling and Representation in ESMs

High-resolution modeling will be vitally important for improving simulations of aerosol life cycles and aerosol-cloud interactions in ESMs. Coarse resolutions in ESMs (100–200 kilometers in horizontal dimensions) cannot represent fine-grained variability in cloud dynamics and thermodynamics in response to aerosol perturbations, which is an important factor contributing to the large uncertainty in aerosol-cloud interactions.

ESMs with increased resolutions are expected to more realistically simulate the dynamics and thermodynamics of clouds and the three-dimensional distributions of aerosols in the atmosphere.

Numeric approximations, or parameterizations, used to represent turbulence are mainly responsible for underestimations of stratocumulus and cumulus clouds in the tropics and subtropics by coarse-resolution models, for example. Parameterizations of convection also lead to significant problems in simulating convective systems.

As the computational power and resources available to scientists increase, including the new generation of exascale computers that can perform a quintillion calculations per second, horizontal resolutions in next-generation ESMs are approaching 10–25 kilometers or even higher: ESMs with resolutions of about 3 kilometers now enable scientists to resolve convection phenomena. ESMs with increased resolutions are expected to have much better capability in realistically simulating the dynamics and thermodynamics of clouds as well as the three-dimensional distributions of aerosols in the atmosphere. As ESMs achieve higher resolutions, the accuracy of turbulence and cloud microphysics parameterizations used in the models becomes increasingly critical. Thus, improvements and developments of model physics should continue to be emphasized in research efforts.

Aerosol Observations

Future observational programs should focus on the detailed multiphase chemistry of atmospheric aerosols and on routine measurements of aerosol profiles in the free troposphere (the part of the lower atmosphere that is not affected by Earth’s surface) and the upper troposphere. Such observations will help to improve understanding of aerosol processes and properties in the atmosphere, and the gained knowledge can be implemented into ESMs to improve their representation of aerosols.

Several specific topics that aerosol observational efforts can shed light on include the following:

  1. The newly identified mechanisms of sulfate formation on the surfaces of preexisting aerosol particles (e.g., dust, black carbon) are one area to study. These aerosol productions have been found to contribute to rapid haze formation in heavily polluted regions (e.g., eastern China).
  2. Secondary organic aerosols (SOA) and their roles in the formation and growth of new particles are also important. SOAs produced by forests and biomass burning, which represented the dominant source of submicrometer particles in preindustrial times, still contribute significantly to the formation and growth of new aerosol particles in both clean and polluted air.
  3. Wildfires, which are becoming more frequent and intense in some regions (e.g., North America) with climate change, exert important influences on radiation, clouds, and biogeochemistry in the Earth system through releases of GHGs and aerosols. However, wildfires and smoke aerosols are poorly represented and predicted in ESMs.
  4. Aerosol processes at different levels in the atmosphere need to be studied further. Current field measurements mostly focus on aerosols in the boundary layer near Earth’s surface, but more measurements of aerosols at upper levels are needed. Wildfires can inject smoke above the boundary layer, as can emissions from aircraft and volcanoes. Long-range transport of aerosols follows a high-altitude route. Aerosols at high levels can be entrained downward, providing an important source of aerosols in the boundary layer, especially in remote regions. And observations of free tropospheric aerosols—which can, for example, seed formation of cold clouds—could provide critical constraints on how convective transport and aerosol scavenging are treated in models.

Coordinating Observations and Models

Coordinated efforts across different disciplines, agencies, and countries are needed to make the sort of rapid progress in aerosol and aerosol-climate studies that is critical for improving ESMs and future climate projection. Traditionally, aerosol measurements have not been used adequately to inform models.

An integrative approach could be used to tie together the different research communities studying clouds and aerosols.

An integrated approach is thus needed. Model developers should identify shortcomings in models and inform experimentalists of these issues. For example, modelers could identify weaknesses in current representations of aerosol parameters and processes or regions that have the largest model biases specifically related to aerosols. Experimentalists could then use this information to collect focused measurements for use in improving model constraints.

A similar integrative approach could be used to tie together the different research communities studying clouds and aerosols. In situ aircraft measurements of clouds are encouraged to include aerosol measurements (e.g., of size distribution and chemical composition), thus making it possible to interpret aerosol-cloud interactions. Developing well-coordinated, international model intercomparison studies (e.g., Aerosol Comparisons between Observations and Models (AeroCom)) of aerosol and aerosol-cloud interactions that involve both experimentalists and modelers is a key step in greatly accelerating model developments that reduce the uncertainties of aerosols in ESMs.

Author Information

Xiaohong Liu (, Department of Atmospheric Sciences, Texas A&M University, College Station; Zhaohui Lin, International Center for Climate and Environment Sciences, Institute of Atmosphere Physics, Chinese Academy of Sciences, Beijing;  and Minghua Zhang, School of Marine and Atmospheric Sciences, Stony Brook University, New York


Liu, X.,Lin, Z., and Zhang, M. (2020), Clearing the haze around aerosol assessments, Eos, 101, Published on 23 June 2020.

Text © 2020. The authors. CC BY-NC-ND 3.0
Except where otherwise noted, images are subject to copyright. Any reuse without express permission from the copyright owner is prohibited.