East Asian Study of Tropospheric Aerosols and their Impact on Regional Clouds, Precipitation, and Climate (EAST‐AIRCPC)

Aerosols have significant and complex impacts on regional climate in East Asia. Cloud‐aerosol‐precipitation interactions (CAPI) remain most challenging in climate studies. The quantitative understanding of CAPI requires good knowledge of aerosols, ranging from their formation, composition, transport, and their radiative, hygroscopic, and microphysical properties. A comprehensive review is presented here centered on the CAPI based chiefly, but not limited to, publications in the special section named EAST‐AIRcpc concerning (1) observations of aerosol loading and properties, (2) relationships between aerosols and meteorological variables affecting CAPI, (3) mechanisms behind CAPI, and (4) quantification of CAPI and their impact on climate. Heavy aerosol loading in East Asia has significant radiative effects by reducing surface radiation, increasing the air temperature, and lowering the boundary layer height. A key factor is aerosol absorption, which is particularly strong in central China. This absorption can have a wide range of impacts such as creating an imbalance of aerosol radiative forcing at the top and bottom of the atmosphere, leading to inconsistent retrievals of cloud variables from space‐borne and ground‐based instruments. Aerosol radiative forcing can delay or suppress the initiation and development of convective clouds whose microphysics can be further altered by the microphysical effect of aerosols. For the same cloud thickness, the likelihood of precipitation is influenced by aerosols: suppressing light rain and enhancing heavy rain, delaying but intensifying thunderstorms, and reducing the onset of isolated showers in most parts of China. Rainfall has become more inhomogeneous and more extreme in the heavily polluted urban regions.


Motivation and Background
As the most populated and fastest developing region of the world, East Asia has drawn much attention regarding the impact of human activities on the natural environment and climate. East Asia, especially China, suffers from severe dust storms and heavy air pollution from various emission sources (e.g., S. Chen et al., 2017;Guo, Deng, et al., 2016;Li, Guo, et al., 2017;Qian et al., 2003). With its sustained and steady economic growth, China's urbanization has stepped into the stage of full-scale acceleration. By the end of 2014, 55% of the total population lived in urban areas, a dramatic increase from 26% in 1990. Accompanied by the migration of people is rapid industrialization with dramatic increases in, for example, the number of factories, power plants, roads, and vehicles. At such a fast pace of urbanization and industrialization, the potential impacts on the environment, including atmospheric composition, climate, and the hydrological cycle, are so strong and fast that the impacts must be understood in order to understand any anthropogenic impact on Earth's climate and environment over a wide range of spatial and temporal scales (Z. . Aerosol particles, along with other pollutants in the gas phase and in water, have been primary environmental concerns in East Asia. Aerosols can cause respiratory and cardiovascular diseases (Brook et al., 2010) and reduce atmospheric visibility (Watson, 2002;. The heavy loadings of sulfuric, carbonaceous, and mineral dust aerosols over East Asia further induce complicated effects on radiative fluxes and climate changes (Z. Li et al., 2007, Li et al., 2011Qian et al., 2006). Increases in aerosol concentration increase concentrations of cloud condensation nuclei (CCN) or ice nuclei (IN) upon which cloud droplets or ice particles form. More but smaller cloud droplets and more reflection of solar energy to space for a given amount of water vapor results. By reducing cloud droplet sizes, which feeds back into cloud thermodynamics and dynamics, aerosols modulate cloud macrophysics and precipitation (Givati & Rosenfeld, 2004). The suppression of warm rain by aerosols causes more condensates to ascend and more latent heat to be released either through deposition or freezing or both. This may invigorate convection, resulting in increased cloud heights and thicknesses for deep convective clouds (DCCs; Li et al., 2008;Rosenfeld et al., 2008), which is the so-called "cold-phase" invigoration. The large number of ultrafine aerosol particles produced by industrialization can be activated in DCCs due to high supersaturation well above cloud base, enhancing condensation and invigorating convective intensity and precipitation in a more significant way than can the "cold-phase" invigoration (e.g., Sheffield et al., 2015;; "warmphase" invigoration). Aerosol microphysical effects on DCCs drastically enlarge cloud anvils (J. Fan et al., 2013;Yan et al., 2014). W. K. Tao et al. (2012), J. Fan et al. (2016), and Z. Li et al.  have comprehensively reviewed the influences of aerosols on radiation, clouds, and precipitation from different perspectives and scopes. Moreover, it was proposed that light-absorbing aerosols within the boundary layer can induce a thermodynamic invigoration of convective clouds by increasing the buoyancy near the cloud bottom and elevating convective available potential energy (CAPE) accordingly (Wang, Khalizov, et al., 2013). The effect of increased aerosols on cloud optical, microphysical, and macrophysical properties and associated radiative forcing is the most uncertain component of historical radiative forcing over the globe (IPCC, 2013).
As such, aerosol-related studies have increased nearly exponentially worldwide since the late 1980s ( Figure 1). In 2017, the total number of aerosol-related publications exceeded 3,000, an increase by a factor of 10. In China, the wave of aerosol-related studies began much later but accelerated more quickly. Prior to 2000, the number of aerosol-related publications (limited to only Chinese journals) was less than 10. However, in 2017, the number of such publications jumped to 1,000, a nearly hundredfold increase in less than two decades. This explosive interest in aerosol science was and continues to be driven by concerns about climate change and the air pollution resulting from fast economic development.
Cloud-aerosol-precipitation interactions (CAPI) have been investigated intensively in an effort to reduce uncertainties in the prediction of climate changes. In recent years, many CAPI studies have focused on regional scales because of the impact on regional precipitation and weather extremes such as torrential rain (e.g., Fan et al., 2015;Lee et al., 2018;. In particular, it is well known that East Asia has experienced dramatic changes in aerosol properties (e.g., loading) since industrialization began in that part of the world. It is thus expected that East Asia has experienced significant changes in clouds and precipitation, partially induced by the changing aerosol properties since industrialization. This merits the examination of CAPI in East Asia.
One major uncertainty of CAPI stems from the source of aerosols and CCN. The implementation of the "Atmospheric Pollution Prevention and Control Action Plan" in 2013 has led to significant improvement in air quality in the North China Plain (NCP) and Yangtze River Delta (YRD) regions of China. However, severe haze episodes characterized by high concentrations of fine particles (PM 2.5 ) still occur from time to time in both urban and rural areas during all seasons, though more frequently during wintertime Sun et al., 2014;. While the composition, sources, and evolution processes of aerosol particles have been extensively studied for megacities in the NCP and the YRD, the level of understanding of aerosol chemistry and new particle formation (NPF) in suburban and rural areas is low. Air pollution has proven to be a regional issue, and NPF is also recognized as a regional phenomenon. However, their influences may go far beyond local scales due to their transport and evolution along with large-scale atmospheric circulations.
East Asia is also an important monsoon area characterized by cold, dry winters and hot, humid summers with prevailing northerly and southerly winds in winter and summer, respectively. Analyses of long-term meteorological records have revealed that the East Asian monsoon (EAM) has experienced weakening trends in terms of both precipitation and circulation since the late 1970s (Ding et al., 2015). While the causes for the anomalies and trends of the EAM are diverse and highly uncertain, there appears to be a link with environmental changes, especially those induced by aerosols (Z. Wu et al., 2016). Also under debate is whether greenhouse gases or aerosol forcing is responsible for the observed declining trend in light precipitation during the past few decades over eastern China , besides natural variability. There is an urgent need to quantify the uncertainty of aerosol effects on climate in association with the uncertainties in aerosol emissions (external factors) and in processes related to aerosol-cloud interactions (ACI; internal factors).
In light of the dramatic and rapid changes in the atmospheric environment and potential impact on regional climate in East Asia, especially China, as well as their interactions, numerous major research endeavors have been made in the region. Findings have been reported in a series of special collections in the Journal of Geophysical Research: Atmospheres entitled "East Asian Studies": the East Asian Study of Tropospheric Aerosols: an International Regional Experiment (EAST-AIRE; Z. Li et al., 2007) and the East Asian Studies of Tropospheric Aerosols and their Impact on Regional Climate (EAST-AIRC; Li et al., 2011). Papers published in the current issue entitled "East Asian Study of Tropospheric Aerosols and Impact on Cloud and Precipitation (EAST-AIRcpc)" follow previous studies attempting to gain deeper insights into pertinent fronts but with a focus on the impact of aerosols on clouds and precipitation. Hence, the present review paper provides a synopsis of the new findings in those papers, centering on CAPI. The chemical formation and transformation of haze pollution, as well as the influence from emissions and meteorological conditions in East Asia, have recently been reviewed systematically by others (e.g., An et al., 2019;, so the present work is an important complement to them. This paper is structured as follows: section 2 summarizes the major methodologies employed in the papers of the current special issue, including groundbased, airborne, and spaceborne observations, as well as numerical models; section 3 reviews the studies pertaining to aerosol physicochemical properties; section 4 discusses the research about process-level understanding of aerosol effects on cloud and precipitation properties; section 5 reviews the large-scale effects of aerosols on the regional climate over East Asia; and section 6 summarizes the new findings in the current special issue and future challenges and directions.

Ground-Based, Airborne, and Spaceborne Observations and Modeling Studies
There is a wide range of observations and model simulations employed in the investigations published in this special section that are summarized as follows.

Ground-Based Observations
To understand formation mechanisms and sources of severe haze pollution, real-time measurements of submicron aerosol species and sizes, as well as precursor gases, were made using a suite of instruments deployed at five sites in China with different background aerosol sources and characteristics (Table 1).
The most comprehensive field experiment conducted was the Atmosphere-Aerosol-Boundary Layer-Cloud Interaction Joint Experiment held at a suburban site in Xingtai in Hebei Province. Xingtai is located near the western edge of the NCP (Figure 2).
These field experiments had the following objectives: 1. Investigation of aerosol and boundary-layer interactions using a variety of measurements of atmospheric profiles; 2. Investigation of cloud microphysical properties and development using aircraft in situ measurements and ground-based observations; 3. Investigation of ACI using coincident aerosol, cloud, and atmospheric observations. To achieve these objectives, extensive measurements were made using a combination of ground-based, airborne, and spaceborne instruments. Five observation systems were deployed on the ground, each of which was composed of a suite of instruments to measure or infer (1) atmospheric profiles of temperature, humidity, wind using radiosondes, a Raman lidar, an infrared hyperspectral radiometer, and a microwave radiometer; (2) aerosol physical, chemical, optical, and hygroscopic properties, in combination with CCN; (3) cloud liquid water path (LWP), droplet size, and vertical profiles; (4) broadband and spectral radiative quantities; and (5) surface heat and turbulent fluxes. Ample measurements made in these experiments have been employed to conduct a wide range of studies toward achieving the above objectives whose findings are reviewed in the following sections.

Airborne and Spaceborne Studies
As a primary index of air pollution, fine-particle (i.e., PM 2.5 ) loading in China is particularly heavy and highly variable in time and space (Zhang & Cao, 2015). Although the number of ground monitoring stations has rapidly increased during the past five years, those at suburban, rural, and remote regions are still limited. As a result, characterization of the spatial distribution of PM 2.5 , even with data assimilation, can have large uncertainties. The satellite-retrieved aerosol optical depth (AOD), a gross proxy of PM 2.5 loading, has advantages in studying the spatial and temporal variations of particulate pollutants . However, the accuracy of satellite retrievals of PM 2.5 strongly depends on the relationship between AOD and PM 2.5 . Although AOD-PM 2.5 relationships have been established for the United States and Europe, it is far from completely understood in China, particularly before 2004 when there was no dedicated national monitoring network in place (Kong et al., 2016;Xin et al., 2007Xin et al., , 2014Xin et al., , 2016. It is, therefore, critical to investigate the relationships between AOD and PM 2.5 across different regions and seasons in China. The AOD-PM 2.5 relationship in China was for the first time elucidated using widely available satellite-retrieved AODs, combined with China Atmosphere Watch Network operated by China Meteorological Administration (Guo et al., 2009). Recently, Xin et al. (2016) analyzed PM 2.5 and AOD data from 23 stations as part of the "Campaign on atmospheric Aerosol REsearch network of China" (CARE-China; Xin et al., 2015). They found grossly consistent changes in AOD and PM 2.5 in three precipitable water vapor ranges. However, the linear regression functions (PM 2.5 = A × AOD + B) exhibited large differences in different regions and seasons. To tackle this problem, T. Su et al. (2018) examined the multiple relationships between AOD, the planetary boundary layer (PBL), PM 2.5 , meteorological variables, and topography. Many of these quantities are intercorrelated and change with season and location. For example, there is a much stronger relationship between PM 2.5 and PBL in northern China in the winter than in southern China in the summer. Normalizing PM 2.5 by AOD generally leads to a stronger relationship with the PBL. Such findings help improve the estimation of PM 2.5 from remote sensing. More importantly, it has been recognized that the AOD of fine-mode aerosols retrieved from space is more useful in estimating surface PM 2.5 . Note that it was not until recently that the fine-mode fraction could be estimated from satellites with reasonable accuracy Yan, et al., 2019). A higher accuracy (R 2~0 .85 or higher) in the estimation of PM 2.5 was recently achieved by means of machine learning using the inputs of either satellite-derived AOD (Wei et al., 2019) or satellite-measured radiances (J. Liu et al., 2019), together with ancillary data such as meteorology and topography.
By applying two techniques widely used in computer science for image processing, Zhao et al. (2014) developed a new method to detect aerosol and cloud layers using the micropulse lidar (MPL). First, they used a semidiscretization processing technique to inhibit the impacts of increasing noise with distance from the MPL instrument. Second, they applied the value distribution equalization method to reduce the magnitude of signal variations with distance. Third, they defined threshold values to determine whether the signals originated from clouds or aerosols. Compared to empirical algorithms, this method can detect aerosols and clouds more accurately. The new method can identify more high-level clouds which were previously wrongly classified as aerosols by the standard method used by the U.S. Department of Energy Atmospheric Radiation Measurement program. Moreover, the new method can accurately identify locations with sharp signal changes at cloud boundaries without smoothing, making the cloud bases not biased low and cloud tops not biased high as what happens with other methods. Without the ability to detect high-level ice clouds, AOD retrievals from satellite and their radiative effects could be highly overestimated, creating issues with understanding the Earth's radiation budget. This is particularly important in urban areas where frequent haze events occur. Haze attenuates the lidar signal and makes distinguishing between aerosols and clouds difficult. Shang et al. (2017) further proposed a new method to detect haze aerosols in the lower atmosphere in eastern China. Most current satellite-based cloud and aerosol detection algorithms make use of radiative properties and cloud geometrical textures to discriminate between cloudy and clear areas or between cloudy and hazy areas (Goodman & Henderson-Sellers, 1988;Remer et al., 2005). Shang et al. (2017) made use of digital elevation model data to adjust the threshold values of brightness temperatures used to detect aerosols and clouds, which can more accurately detect aerosols and clouds under heavy pollution conditions. The main idea proposed by Shang et al. (2007) is that haze occurs near the ground, whereas clouds typically exist at high altitudes.
The errors in aerosol retrievals come not only from the classification of aerosols and clouds but also from the impacts of particle optical properties and surface conditions. M. Tao et al. (2017) evaluated Moderate Resolution Imaging Spectroradiometer (MODIS) Deep Blue (DB) Collection 6 aerosol retrievals in the desert regions of East Asia based on surface measurements from the China Aerosol Remote Sensing Network (CARSNET). They found that MODIS DB aerosol retrievals are low biased with most values falling outside the expected error envelope ± (0.05 + 20% AOD CARSNET ). Two reasons for this were given: (1) the overestimated scattering ability of airborne dust over East Asia and (2) the overestimation of surface reflectance. The study by M. Tao et al. (2017) implies that MODIS DB retrievals can be improved by applying more realistic aerosol models and surface characterizations over East Asia based on more surface measurements.
The vertical structure of aerosols in China and their radiative properties, especially their absorbing capability, were assessed using long-term satellite observations from the Cloud-Aerosol Lidar with Orthogonal Polarization (CALIOP) and ground-based lidar observations and Aerosol Robotic Network data. P. Tian et al. (2017) reported that more than 80% of the column aerosols are distributed within 1.5 km above the ground in the winter when the aerosol extinction lapse rate exhibits a maximum seasonal average in most regions of China except for the Tibetan Plateau (TP). The aerosol extinction lapse rates in polluted regions are higher than those of less polluted regions, indicating a stabilized atmosphere due to absorptive aerosols in the polluted regions. Tian et al. (2018) also found a strong enhancement of absorption due to a mixing of dust (mineral absorbing) and anthropogenic pollution (black carbon, or BC, absorbing), as revealed by the spectral behavior of the absorbing AOD, single scattering albedo, and imaginary refractive index.

Modeling Studies
Comprehensive modeling studies were conducted using a variety of models as summarized in Table 2.
Traditionally, the aerosol effect is quantified in the climate modeling community by looking at the differences between two numerical simulations, one with present (higher) aerosol loading and the other with historical (lower) loading. However, this approach is not feasible for studying how aerosols impact the response to the gradual change in aerosol emissions from low to high. H. Yan et al. (2015) adopted a novel parametric sensitivity analysis framework that integrates the quasi-Monte Carlo parameter sampling approach and a surrogate model to examine aerosol effects on the EAM climate using the Community Atmosphere Model (CAM5). A total of 256 CAM5 simulations  were conducted to quantify the model responses to uncertain parameters associated with cloud microphysics parameterizations and aerosol (e.g., sulfate, BC, and dust) emission factors as well as their interactions ( Figure 3). Interaction terms among parameters, which are usually ignored in modeling studies, are found to be important for quantifying the sensitivity of precipitation to the parameters. Their results also showed that sulfate, BC, and dust aerosols have very different impacts on the EAM climate through aerosol-cloud-radiation interactions. The climatic effects of aerosols do not always have a monotonic response to changes in emission factors. They also identified cloud microphysical parameters that have the most significant impacts on the climatic effect induced by different types of aerosols in East Asia.

Aerosol Formation, Nucleation, Optical Properties, and Radiative Effects
The importance of aerosol formation and transformation processes and their radiative effects in context with East Asia pollution have been identified and recognized in many previous studies. For example, NPF has been documented as a significant source of the aerosol population in China  and also an important source of CCN . The aerosol radiative effect, rapid aging, and enhanced absorption in polluted urban environments in China were characterized by chamber experiments (Peng, Hu, et al., 2016;Peng et al., 2017). Concerning aerosol growth, the formation of sulfate involving heterogeneous reactions  is crucial because these reactions largely regulate the aerosol hygroscopicity. These new findings from the EAST-AIRcpc are reviewed next.

Aerosol Formation and Growth
NPF is a ubiquitous phenomenon that has important implications in understanding the formation of secondary aerosols, air quality, and climate. However, NPF and particle growth in highly polluted environments remains less understood . Dai et al. (2017) made nearly two months of simultaneous measurements of particle number size   and Guo, Gao, et al. (2014) WRF with two-moment bulk-bin cloud microphysics and aerosol radiative effect distributions (4-750 nm) at an urban site and a regional background site in the polluted YRD region. They found that the formation and growth rates at the polluted urban site were always lower than those at the regional background site during seven regional NPF events. This implies that regional NPF intensities are likely substantially underestimated if measurements are solely available from urban sites. Dai et al. (2017) also found a type of strong local nucleation event at the urban site where the formation rate of particles (e.g., J5) was positively correlated with H 2 SO 4 and enhanced by anthropogenic volatile organic compounds from nearby industrial emissions. This finding highlights that local NPF events might be an important contributor to CCN in polluted urban environments. ) made real-time measurements of submicron aerosol species using an aerosol chemical speciation monitor in June 2013 at the suburban site Xianghe located~50 km southeast of Beijing. There was an overall similarity between the aerosol composition at urban and rural sites in the NCP which are dominated by organics (35-40%) followed by nitrate (20-25%) and sulfate (18-25%; J. Sun et al., 2010;Y. Sun et al., 2012, Sun, Jiang, et al., 2016. This highlights the regional characteristics of air pollution. Y. Sun et al. (Sun, Jiang, et al., 2016) also found a significant impact of biomass burning (BB) on air quality in June. The average PM 1 concentration was 126 μg/m 3 during BB periods, which is more than twice that during non-BB periods (61 μg/m 3 ). In particular, BB organic aerosols (OA), which account for 21% of the total OA, show the largest increase (by a factor of 10) during BB periods. In addition to BB, Y. Sun et al. (Sun, Jiang, et al., 2016) also highlighted two major sources affecting air quality in the NCP: (1) regional transport dominated by secondary inorganic aerosols and secondary OA and (2) local source emissions dominated by organics and BC.
Particle growth plays a critical role in the deterioration of visibility for which aerosol hygroscopicity is a key factor. During field campaigns carried out in the NCP, this property has been investigated using in situ and remote sensing approaches . The former was achieved using the humidified tandem differential mobility analyzer and many collocated measurements, as shown in Table 1. The campaign in Beijing took place around the time of the 2015 China Victory Day Parade during which many drastic measures were implemented to reduce emissions in Beijing and surrounding regions. Comparisons before and after the parade (polluted and clean conditions, respectively) showed significant changes in aerosol species and associated aerosol properties, that is, hygroscopic growth. During the clean period, the dramatic reduction in inorganic gases and aerosols (especially secondary aerosols) rendered particles more hydrophobic and volatile, leading to a further improvement of air quality. Comparatively, measurements made at Xingtai, one of the most polluted regions in China, showed that aerosol particles are more hydrophilic and thus have a stronger swelling effect associated with the internal mixing of secondary aerosol species . By combining aerosol chemical composition measurements,  further analyzed the roles of different aerosol species in affecting aerosol hygroscopic growth in Xingtai. However, most measurements of hygroscopicity are currently limited to the ground level. Until recently, Raman lidar retrievals were used to infer the hygroscopicity of aerosol particles aloft, as demonstrated by the campaign in Xinzhou (Lv et al., 2016).

CCN
By serving as CCN, aerosol particles play a key role in the formation of clouds and their evolution during which they alter both the water and energy cycles in the Earth's climate system. CCN is thus a dominant factor in ACI, which can further alter the energy balance of a cloud. To better understand ACI and their impact on precipitation and radiation, it is important to gain a good knowledge of the ability of aerosol particles to form CCN at the typical supersaturations found in the atmosphere. Due to the large spatial variability of aerosol types and compositions, the CCN activation efficiency varies greatly over different regions, especially in such a large and fast-developing country like China.
To this end, numerous field experiments have been conducted in China in recent years involving measurements of CCN, aimed at better characterizing CCN and its influential factors (e.g., Li, Zhang, et al., 2017;Ren et al., 2018;Wang, Li, Zhang, et al., 2018;. These studies provide different perspectives on the influence of aerosol particle size and composition on CCN activity. For example, F. Zhang et al. (2014) demonstrated that the 30-40% uncertainties in the number of CCN (N CCN ) are mainly associated with changes in the particle composition. Also, the impacts of OA on CCN activity were investigated based on field measurements made at a suburban site in northern China. A strong dependence of N CCN on the volume fraction of organic material and aerosol oxidation level was reported . NPF events were also found to enhance N CCN during the "leveling-off" stage but restrain CCN activation at the "growth" stage during which larger particle diameters are needed to reach an activation diameter . Using a comprehensive data set including size-resolved CCN activity, size-resolved hygroscopic growth factor, and chemical composition, F. Zhang et al. (2017) found that the effect of aging/coating on particle hygroscopicity for aged aerosols leads to an underestimation of~22% of N CCN . Such a coating effect has not been considered properly in most models. The underestimation of N CCN is at par with the underestimation of aerosol hygroscopicity for aged aerosol particles (Wang, Li, Zhang, et al., 2018). For fresh aerosol particles, however, the CCN activity tends to be overestimated largely because of inaccurate assumptions about the particle mixing state and the variability in chemical composition over the particle size range of activation. A much better agreement between observed and model-predicted CCN can be achieved by accounting for size-resolved chemical compositions, as was done by Ren et al. (2018).
Despite the significant accomplishments of these studies, long-term measurements of those aerosol properties at more sites are needed to get more accurate CCN predictions under different atmospheric environments. This would improve the parameterization of N CCN in models, which is critical for estimating ACI (Sotiropoulou et al., 2007). Novel remote sensing approaches that use clouds as cloud chambers , together with satellite-derived updraft speeds (e.g., Zheng et al., 2015), may help in this regard.

Visibility, Aerosol Optical Properties, and Radiative Effects
Visibility, a widely used proxy of air quality, has been measured as a standard meteorological variable at virtually all weather stations in China. Chen and Wang (2015) conducted a comprehensive analysis of the 1960-2012 visibility trend in northern China. They found that the poorest visibility tends to occur in the morning in wintertime. The most dramatic deterioration in visibility occurred from the 1960s to the 1970s and continued into the 2000s at a more moderate rate. Significant differences were noted between urban and rural areas, with the former being worse than the latter. Relative humidity, a key factor influencing visibility, is virtually identical between urban and nearby rural regions, signifying that the difference originates from the emissions of urban pollutants. Bi et al. (2014) examined several serious atmospheric haze events that occurred over eastern and northern China in January 2013. They found that the volume size distribution and median radius of the fine-mode aerosol particles generally increased with AOD and discovered a significant amount of absorbing aerosols during these haze episodes. Correspondingly, they showed a strong influence of the aerosol direct radiative (ADR) effect on the atmospheric thermal structure: Up to half of the incident solar radiation was absorbed in aerosol layers, causing exceptionally strong heating of the atmosphere and cooling of the surface. The relationship between daily mean aerosol radiative forcing (ARF) and AOD at 500 nm at the surface and at the top of the atmosphere (TOA) was studied. The slopes of the best linear fittings were 11.02, 11.23, and 11.27 at the surface and 4.14, 3.57, and 2.56 at the TOA for three sites in Beijing. The ARF values at the surface are about twice those at the TOA. The daily mean heating rate in the aerosol layer over Beijing ranged from 0.12 to 0.81 K/day. Liu et al. (2011) examined the aerosol optical properties and their associated radiative effects based on skyradiometer and surface solar radiation data in the spring over the Loess Plateau in northwest China. Y. G.  and Jia et al. (2018) examined aerosol transport to the TP and the dust-induced heating effect over the TP using satellite data, a radiative transfer model, and the Spectral Radiation-Transport Model for Aerosol Species. The aerosol index observed by the Ozone Monitoring Instrument shows that absorbing aerosols are distributed over the northern and southern slopes of the TP, corresponding to dust and carbonaceous aerosols, respectively. Model simulations show that high values of dust AOD are primarily distributed over the northern slope of the TP, whereas high AOD values due to anthropogenic aerosols, including carbonaceous and sulfate aerosols, are distributed over the southern slope and east of the TP. Dust aerosols significantly affect the radiative energy budget and thermodynamic structure of the air over the TP, mainly by altering the shortwave radiation budget. The instantaneous heating rate can be as high as 5.5 K/day depending on the density of the dust layers. Wang, Zhang, et al. (2016) examined the effect of ADR on East Asian surface radiation and air quality using Weather Research and Forecasting model coupled with chemistry (WRF-Chem) simulations during the EAST-AIRE campaign. The ADR effect decreased surface downwelling radiation, the surface air temperature, the boundary layer height (BLH), and the temperature lapse rate. The PM 2.5 concentrations increased accordingly by 4.4%, 10%, 2.3%, and 9.6% in eastern China, southern China, western China, and the Sichuan Basin, respectively. Another observation-based estimation by  found that the ADR effect can strengthen the total aerosol amount by~12% and~15% through reductions in the near-surface wind and BLH in Beijing, respectively, which are even larger than model simulation results.  further analyzed the sensitivity of the ADR effect to aerosol type in China and found that the net effects of aerosols on the surface solar radiation and corresponding climatic effects are dependent on both the aerosol amount and optical properties. Note that the aforementioned studies are mostly of short duration. Interestingly, radiosonde-derived BLHs have shown a trend reversal (from increasing to decreasing) in recent decades in China. This is found, however, to be correlated more with increases in soil moisture than with the aerosol trend . Therefore, the long-term impact of ADR, especially on the PBL structure, remains inconclusive and merits further analyses by combining observations with model simulations where land-atmosphere coupling should be considered.

Impact of Aerosols on Clouds
Ground-based and spaceborne remote sensing techniques have proven indispensable for learning more about the spatiotemporal variations of cloud and precipitation. Comparisons between cloud properties retrieved from the surface and spaceborne platforms are commonly made to validate satellite retrievals (e.g., Dong et al., 2002Dong et al., , 2008. It is important to understand and reconcile any systematic differences that arise before attributing any discrepancies to satellite retrieval errors. Comprehensive surface-based retrievals of cloud optical and microphysical properties were made at Taihu, a highly polluted site in the central YRD region, which were further compared to corresponding MODIS-Terra and MODIS-Aqua cloud property retrievals (J. Liu et al., 2013). Large discrepancies between surface-and satellite-retrieved cloud properties were found, suggesting that MODIS cloud products suffer from large uncertainties in this highly polluted region. The degradation in agreement likely stems from the influence of absorbing aerosols, which can also impinge on the retrieval of cloud properties (Coddington et al., 2010;Wilcox & Platnick, 2009). Z.  developed a ground-based retrieval algorithm for the simultaneous retrieval of cloud optical depth (COD) and cloud droplet effective radius over the same sites as that in J. Liu et al. (2013) to examine the influences of aerosols on both surface and satellite retrievals. They found stronger sensitivities in the retrieval of COD from spaceborne sensors to aerosol absorption and vertical distribution than that from ground-based retrievals. For absorbing aerosols mixed with clouds, COD tends to be underestimated from space, but overestimated from the ground, leading to a very poor agreement between ground-based and MODIS retrievals. This difference increases with increasing COD. Such a significant bias may be misconstrued as the semidirect effect of aerosols; that is, COD decreases with increasing AOD due to the cloud droplets undergoing evaporation by aerosol-induced absorption. Caution must thus be exercised when examining ACI using both satellite and ground retrievals over regions with an abundance of strongly absorbing aerosols.
Aerosols can alter cloud microphysics, which is concerned with the amount, phase, and size of cloud droplets and ice crystals. The first indirect effect (FIE) describes the reduction in cloud droplet effective radius with an increase in aerosol concentration under fixed LWP conditions (Feingold, 2003;Kim et al., 2008;Liu, Chen, et al., 2016). This implies an increase in the albedo of a cloud, resulting in enhanced reflection and a cooling effect (Twomey, 1977). Several studies using space-, surface-, and ship-based measurements have revealed the strong impact of aerosols on cloud droplet size, COD, and cloud reflection and emission (Feingold, 2003;Feingold et al., 2006;Kim et al., 2008;Liu, Chen, et al., 2016;Liu & Li, 2018a), indicating that the magnitude of the FIE differs drastically depending on aerosol properties, meteorological conditions, analysis methods, and so on (Liu & Li, 2018b;Liu, Chen, et al., 2016;Rosenfeld & Feingold, 2003;. Under highly polluted conditions over eastern China, a positive FIE was observed; that is, cloud droplet effective radii increased with AOD ( Huang et al., 2015;Wang, Guo, et al., 2015) and the aerosol swelling effect (Liu & Li, 2018b), or both. Additional nucleation of cloud droplets aloft may also contribute because it decreases the effective radius with height. Meteorology plays a significant role in observational studies of the FIE. A positive relationship between the FIE and atmospheric stability was derived from surface-retrieved cloud and aerosol information, contradicting the negative relationship found from satellite data. This contrast originates from differences in the retrieved cloud microphysical properties at different levels from satellite and surface sensors, which likely explains some major discrepancies in many past studies concerning ACI . The positive indicated FIE is likely unreal because the increase in AOD under moist conditions is not generally associated with an increase in CCN.
Aerosol particles like mineral dust, soot, organic matter, and biological aerosols can be efficient IN, helping to form ice crystals at temperatures below 0°C. Increases in IN could lead to the so-called glaciation indirect effect for mixed-phase clouds (Ovchinnikov et al., 2011). Satellite observations have documented the relationships between aerosols and the glaciation temperature (T g ) in growing convective clouds over East Asia and adjacent oceans (Rosenfeld et al., 2011). They showed that the most important factor in the determination of T g is the cloud droplet effective radius at the −5°C isotherm. The effective radius is smaller with greater AOD, which leads to smaller T g (see also Zhu et al., 2015). The cause for the increased T g with less AOD is the larger droplets which freeze more quickly, mainly by the ice multiplication mechanism (Hallett & Mossop, 1974). For clouds with similar droplet effective radii at −5°C, desert dust and urban air pollution cause higher T g compared to other kinds of aerosols. Long-range transported dust particles from Asia were found to enhance snow precipitation in California orographic mixed-phase clouds . Nevertheless, studies of cloud thermodynamic phase changes and the vertical structures of mixed and ice-phase clouds over East Asia from both observations and model simulations are still sparse (Fan et al., 2012;J. Liu et al., 2013). The impacts of aerosols on cold-phase cloud thermodynamic phase changes over East Asia were studied using four years of combined measurements from the CloudSat radar and the Cloud-Aerosol Lidar and Infrared Pathfinder Satellite Observation (CALIPSO) lidar . Seasonal anomalies of glaciated and mixed-phase clouds correlate well with the seasonal variations in the frequency of dust occurrence. In addition, cold-phase clouds associated with mineral dust contain approximately 5%, 10%, and 20% larger glaciated (smaller mixed-phase) relative cloud fractions (CFs) than those associated with polluted dust, smoke, and background aerosols at any given cloud-top temperature . The study by Zhang, Liu, et al. (2015) illustrates the impact of dust on subfreezing clouds over East Asia by providing abundant effective IN to glaciate mixed-phase clouds. Zhu et al. (2015) found that for convective clouds in the developing stage, the cloud droplet size increases with height, but the rate of change varies with aerosol loading, that is, faster under clean conditions than under dirty conditions. The cloud droplet size must reach a critical size of 13-14 μm for the initiation of the conversion of cloud droplets to raindrops and thus of precipitation . This leads to a positive correlation (with a correlation coefficient of 0.83) between AOD and cloud thickness, which is required to initiate precipitation. In a heavily polluted environment (e.g., AOD >0.6), clouds do not precipitate unless their thicknesses are on the order of a few kilometers. Clouds that are a few hundreds of meters thick can rain out if the AOD is less than 0.1. This was also measured directly in relation to cloud-base CCN and drop concentrations in India (Konwar et al., 2012) and in the Amazon (Braga et al., 2017). This finding has important implications for weather forecasting. Neglecting aerosol effects can lead to global rainfall forecast errors whose spatial pattern bears a close resemblance to the spatial pattern of aerosol loading (Jiang, Feng, et al., 2017).
The influence of aerosols on cloud droplets and ice crystals can induce changes in cloud lifetime and CF. Both positive and negative effects have been reported, arising from various mechanisms for different cloud regimes. A strong positive relationship between AOD and shallow CF has been widely observed over the Atlantic Ocean (e.g., Kaufman & Koren, 2006), the Amazon (e.g., Kaufman et al., 2005), the southeastern Pacific Ocean (e.g., X. Zheng et al., 2010), and eastern China (e.g., Wang, Guo, et al., 2014). This likely originates from aerosol microphysical and lifetime effects because the smaller cloud droplets resulting from more CCN are less likely to rain out and more likely to grow. Direct relationships between marine boundary layer cloud-base droplet concentration, cloud cover, and cloud radiative effects show much stronger relationships than with AOD (Rosenfeld et al., 2019). This is so because clouds are most sensitive to changes in aerosols under clean conditions when AOD does not produce measurable signals at CCN <100 cm 3 (Shinozuka et al., 2015). Meanwhile, the reduction in cloud droplet size may lead to increased evaporation and a reduction in CF (e.g., Small et al., 2009).
The aerosol microphysical effect also invigorates DCCs, elevates cloud tops, and induces increases in the anvil cloud cover of DCCs, leading to an overall warming effect in a polluted environment (e.g., Xu et al., 2017;. Two main mechanisms have been proposed for the aerosol convective invigoration: a Enhanced latent heat induced from freezing of more condensates: Rosenfeld et al. (2008) have shown that the initiation of warm rain due to added CCN can lead to a greater amount of supercooled cloud water, which can freeze aloft and convert to ice hydrometeors, releasing more latent heat as a result of enhanced freezing and deposition. This would not happen if the same precipitation occurred as warm rain lower in the cloud. The added latent heating invigorates the clouds. b Enhanced latent heat of condensation:  have shown that added ultrafine aerosol particles (UAP), which are generally too small to be activated at cloud base, can invigorate deep convection through being activated well above cloud base where in-cloud supersaturation is high because coalescence reduces the number of droplets for condensation. With a low concentrations of aerosols larger than 50 nm (such as several hundred cubic centimeter or less) and a humid condition in Amazon, the coalescence can lead to a vapor supersaturation exceeding 10% in strong updrafts when UAPs are scarce. When UAPs are present, they are activated and form much more additional cloud droplets, strengthening condensation and thereby latent heat release. As a result, the supersaturation can be lowered to 2-3% in the strong updrafts. This process can also occur at the supercooled levels and lead to nucleation of a larger number of smaller ice crystals, which form anvils with larger amounts of smaller ice crystals . This promotes the longevity of the anvils.
This aerosol convective invigoration increases cloud-top heights, thicknesses, or both, which is most significant for DCCs with warm cloud bases and relatively weak wind shear (e.g., Peng, Li, et al., 2016).

Impact of Aerosols on Precipitation
The initiation and development of convection are considerably affected by atmospheric dynamics and thermodynamics except for cloud microphysics. Aerosols, via the two main pathways of the radiative effect (i.e., aerosol-radiation interactions, or ARI) and the microphysical effect (i.e., ACI), can significantly change precipitation properties, including intensity, frequency, and initiation time. ARI tend to alter the atmospheric circulation by changing the thermodynamic profile of the atmosphere, which, in turn, changes the occurrence of clouds and precipitation. For example, BC is believed to account for the summertime "southernflood-northern-drought" pattern over recent decades in China (Menon et al., 2002). The BC produced by biomass burning has proven to dramatically heat the atmosphere and stabilize air, making it difficult for clouds to form (Koren et al., 2004). The burning effect can also make clouds evaporate faster (Ackerman et al., 2000). By contrast, the quantification of the impact of ACI on precipitation appears challenging. The magnitude of ACI varies according to different cloud and precipitation regimes and depends on the large variation in meteorology. Even the sign of ACI changes or reverses abruptly, likely due to the buffered system embedded in the aerosol-cloud-meteorology system (Stevens & Feingold, 2009). The buffering effect explains the insensitivity to aerosols after system adjustment. The different signs of ACI are mainly due to different dynamics and thermodynamics.
As AOD increases, both rainfall amount and frequency from DCCs increase at first and then decrease, which was noted in numerous previous studies as reviewed by Z. Li et al. (2017a). To date, there have been many studies on the impact of aerosols on precipitation around the world and in China, as shown in Figures 4 and  5, respectively.
The rainfall suppression by aerosols has more to do with the aerosol radiative effect, whereas the enhancement is often fueled by the latent heat release due to both condensation and freezing (Li, Rosenfeld, et al., 2017). Downwelling solar radiative fluxes at the surface decrease as AOD increases. This decrease in solar radiation leads to a decrease in ground heat fluxes and CAPE, which is unfavorable for the development of convective clouds and precipitation (Guo, Deng, et al., 2016;Jiang et al., 2016). Cloud droplets tend to form more easily when aerosols increase due to increases in coalescence and collision induced by more CCN. By neglecting ACI, operational weather forecasts tend to overforecast light rain and underforecast heavy rain (Jiang, Feng, et al., 2017). This ACI effect leads to suppression of shallow liquid clouds and invigoration of warm-base mixed-phase clouds. The orography effect on precipitation cannot be ignored either (e.g., Dong et al., 2018;Fan et al., 2015;Guo, Deng, et al., 2014). In addition to changes in the prevailing circulation and moisture transport patterns, a decrease in precipitation over mountains has occurred mainly because increases in the local water vapor saturation capacity, which scales with temperature, have outpaced the available moisture supply. This reduces the relative humidity, suppressing precipitation . The coupling effect of aerosols and orographic lifting has proven to be crucial for precipitation forecasts Rosenfeld et al., 2007).
Following previous pioneering studies (e.g., Qian et al., 2009;Zhao et al., 2006), intensive follow-up studies of the impact of aerosols on precipitation in China further confirmed that the increases in aerosol loading in recent decades are closely associated with changes in frequency and intensity of precipitation over the Guanzhong Plain of central-western China (Y. Wang et al., 2011, Wang, Allen, et al., 2016, the NCP , and southeast China (Jiang, Feng, et al., 2017;. Typically, light rain has decreased, and heavy rain has increased, based on long-term contemporaneous aerosol and precipitation observations. No matter if rainfall is produced by orographic-forced shallow clouds or by mesoscale stratiform clouds, light rain is significantly inhibited, consistent with simultaneously observed increases in aerosol loading, which seems to be a testimony to the aerosol effect . From the perspective of precipitation diurnal variability, aerosol pollution is observed to suppress the initial convection over the Pearl River Delta region of China, leading to more frequent light rain prior to peak precipitation. This is followed by much delayed and considerably enhanced convection afterward (Guo, He, et al., 2016). Model simulations confirm this phenomenon, which is attributed to competing mechanisms induced by aerosol radiative and microphysical effects .
Analyses using hourly gauge data combined with long-term visibility measurements  point to the widespread declining trend in the frequency of summertime local-scale convective precipitation in eastern China. Its intensity, however, has increased, which can be due to the accelerated release of energy caused by the aerosol invigoration effect (Fan, Rosenfeld, et al., 2018;Rosenfeld et al., 2008). This effect was observed for the first time from the perspective of the spatial scale over which precipitation events occur. By contrast, the frequency of large-scale precipitation changed little, as opposed to the increasing trend in frontal precipitation (Day et al., 2018).
The advent of state-of-the-art satellites with active sensors (e.g., CloudSat, CALIPSO, and the Tropical Rainfall Measuring Mission satellite) allows us to take a closer look at the response of clouds or precipitation to aerosols from the vertical perspective (Wall et al., 2014).  investigated the climatology of various cloud regimes in eastern China and connected relative changes in the vertical structures of clouds to aerosols. Among the cloud regimes, convective clouds were found to be significantly invigorated in the upper troposphere. By focusing on localized precipitation events, Guo et al. (2018) elucidated the responses of vertical precipitation structures to aerosols. One intriguing phenomenon is that the enhancement of convective precipitation was largely limited to the upper troposphere at the expense of lower tropospheric precipitation. Overall, the connection between changes in localized cloud or precipitation systems and aerosols seems to be more significant than that for large-or synoptic-scale systems. In the future, more focus will be placed on the initial stages of localized clouds or precipitation, which can only be accomplished through the combined use of geostationary observations and convection-permitting and -resolving model simulations.
Despite the increasing evidence of aerosol effects on precipitation, such effects seem to account more for rainfall variations than long-term trends in rainfall changes. Y.  found that the accumulation of greenhouse gases, not the changes in aerosol properties since industrialization, is responsible for the shifts in precipitation intensity on a global scale. However, in East Asia, dramatic increases in anthropogenic aerosols appear to account for most of the observed light precipitation suppression since the 1950s. Under a warming climate induced by greenhouse gases, the enhanced ascending motions primarily lead to a decrease in the light precipitation frequency and increases in moderate and heavy precipitation frequencies over the tropics. There is no significant change in ascending motions over East Asia solely due to greenhouse gas forcing.

Journal of Geophysical Research: Atmospheres
Mesoscale convective systems (MCSs) are responsible for the majority of precipitation. Hence, there is a very important implication in weather and climate by studying aerosol impacts on various forms of MCSs. However, there is limited work on this. Based on a summer convection event in East China, Q.  found that the strong wind shear at different vertical levels changes the convective system organization, leading to the formation of different kinds of MCSs and total precipitation amounts over the domain. They found that aerosols induce stronger updrafts and downdrafts in all forms of MCSs. The stronger updrafts and enlarged convective core area contribute to larger vertical mass fluxes and enhance precipitation, indicating convective invigoration. The accumulated rainfall and mean rain rate are increased, with an increased occurrence frequency of heavy rain. This frequently involves decreases in the frequency of relatively weaker rain, leading to negligible changes in the rainfall amount. Both convective and stratiform regions show an increase in the rain rate with a more significant increase in the convective rain rate. The unanimous invigoration of MCSs by aerosols under various wind shear conditions revealed by this study has an important implication for weather and climate in warm and humid East Asia.
The impacts of the long-range transport of natural and anthropogenic aerosols on marine clouds and precipitation have received increasing attention. The long-range transport of anthropogenic aerosols from Asia in recent decades has been attributed to climatically significant changes in winter storm patterns over the Pacific. The observed increasing DCCs and precipitation, as well as a decadal variation in midlatitude cyclones, may be at least partially attributed to CAPI (G. Li et al., 2008;Zhang et al., 2007). A recent observational analysis showed a significant correlation between enlarged rainfall areas of tropical cyclones over the northwestern Pacific with increasing anthropogenic aerosols .

Impact of Aerosols on Regional Climate
It is important to detect and diagnose the regional climate impacts of aerosols (both anthropogenic and natural) for improving physical parameterizations and the validation of simulations. The impact of aerosols on the East Asian climate in the premonsoon season (spring) has attracted considerable attention. Previous studies Hu & Liu, 2013;Jiang et al., 2015;Jiang, Yang, et al., 2017) have shown that an elevation in anthropogenic aerosols could lead to a significant reduction in precipitation in south China during spring. Jiang et al. (2015) further found that the distinct climatic background of East Asia could modulate aerosol effects. Large amounts of middle-and low-level clouds in cold seasons favor a strong aerosol indirect effect that leads to a significant spring drought in southern China. The summer monsoon circulation transports more aerosols to northern China, which explains why the summer drought takes place in northern China .

Impact on Temperature and Its Diurnal Range
Temperature is an important climate indicator and important for human life owing to its physical links to the radiative and heat flux budgets. Aerosols may alter both temperature  and the diurnal temperature range (DTR) by the direct effect or by indirectly modulating clouds or both. Z.  argued that the cooling trend from the 1960s to the 1990s in central-eastern China in the midst of widespread warming across China (Ding et al., 2007) was caused at least partially by the rapid increase in air pollution, consistent with the long-term trends of changes in surface solar radiation  and visibility .
Note that the DTR is also affected by other factors such as clouds, the greenhouse effect, and the urbanization heat effect (X. . A dominant feature of DTR changes in East Asia is its steady deceasing trend during the past five to six decades (L. Liu et al., 2018). Over heavily polluted northern China, the trend in DTR from 1960 to 2014 is −0.30°C per decade, much faster than that under clear-sky conditions (−0.17°C per decade; Xue et al., 2018). Over cloudy southeastern China, the observed DTR reduction under overcast conditions is even more significant (Xia, 2013). Previous observations and simulations have shown that both anthropogenic greenhouse gases and aerosols contribute to the decreasing DTR, especially over eastern Asia (Liu, Li, Yang, et al., 2016) and central China . Through its direct effects, aerosols could dampen the DTR, which may explain the decreasing trend in DTR under clear-sky polluted conditions. In northern China, the DTR tends to decrease as the aerosol dry extinction for a given specific humidity increases (Xue et al., 2018). The sunshine duration, a proxy of solar radiation, likely plays a role in modulating the maximum temperature seen in the weak DTR trend over eastern China since 1990 (Shen et al., 2014). Station surface incident solar radiation has consistently declined from 1994 to 2010 (−1.06 W/m 2 per decade; Wang & Wild, 2016). Such covariations between the DTR and solar radiation dimming/brightening are observed on monthly to decadal time scales (Wang & Dickinson, 2013). Recent efforts (e.g., L. Liu et al., 2018) suggest that changes in the large-scale circulation under global warming may at least partly account for the spatially uniform, declining pattern of the DTR.

Interactions Between Aerosols and Monsoon Climate and Circulation
Numerous studies have used global climate models to investigate anthropogenic aerosol effects on the East Asian monsoon climate (e.g., Jiang, Huo, et al., 2017;Tian et al., 2018). Anthropogenic aerosol effects on the East Asian summer monsoon (EASM) have been intensively investigated because the variation in summer monsoon precipitation has substantial impacts on agriculture, the economy, and human life . Numerous model results suggest that anthropogenic aerosols tend to reduce the land-sea thermal contrast, weaken the EASM, and thus decrease rainfall over land regions Guo et al., 2013;Jiang et al., 2013;. The decrease in land monsoon precipitation can reduce the upper tropospheric condensational heat release (cooling effect) and further weaken the EASM (dynamical feedback; Jiang et al., 2013). The EASM weakening could be directly induced by aerosol impacts on radiation and clouds (fast response) or indirectly by sea surface temperature (SST) adjustments (slow response). The fast response plays a dominant role in the land rainfall decrease (X. Li et al., 2018), while the slow response is more important in reducing the land-sea thermal contrast and weakening the monsoon circulation (Wang, Xie, et al., 2017).
In a modeling study of interactions between the EASM and anthropogenic aerosols, Wang, Zhuang, et al. (2015) used a coupled regional climate-chemistry model, a similar framework used by Qian and Giorgi (1999) and Qian et al. (2001). They found that aerosols emitted from East Asia reduced the sunlight reaching the surface and cooled the surface air over land, thus decreasing the land-ocean air temperature gradient and weakening the summer monsoon. This weakened monsoon circulation is unfavorable for the dispersion of aerosols, resulting in a further increase in aerosol loading over East Asia. However, the aerosol feedback in the study by Wang, Zhuang, et al. (2015) may be underestimated because the SST is fixed, and the external forcing is constrained by the lateral boundary condition in the limited-area model they used.
Compared with the summer monsoon, anthropogenic aerosol effects on the East Asian winter monsoon (EAWM) are rarely investigated. Niu et al. (2010) found that atmospheric aerosols tend to heat the atmosphere and generate a cyclonic circulation anomaly over eastern-central China during winter. By analyzing meteorological observations during holidays, J. Zhang et al. (2018) reported that following a significant reduction in air pollution, there tends to be anomalous cyclonic circulation over East Asia-northwestern Pacific. On holidays, the precipitation frequency and intensity, both tightly linked to the anomalous cyclone, decreased over southern China. Jiang, Yang, et al. (2017) proposed a new mechanism responsible for anthropogenic aerosol effects on the EAWM, as illustrated in Figure 6. They found that BC-induced TP warming tends to intensify the northern mode of the EAWM (cooling in northern East Asia) by jet stream acceleration around 40°N and by the westward shifting of the East Asian major trough. Two positive feedback are underlined: the snow-albedo feedback over the TP and the transient eddy-mean flow feedback in the midlatitudes.  investigated the dual effects of the EAWM on haze-fog variations in eastern China.
Aerosol concentrations are also strongly modulated by monsoon circulation through surface winds, rain scavenging, and stratification stability Jeong & Park, 2017;Jia et al., 2018;Liu, Sheng, et al., 2017;. In summer, haze-fog events are related to the monsoon-induced changes in air flow transport and moisture convergence over eastern China (Liu, Sheng, et al., 2017;Wu et al., 2016). On a decadal time scale, a negative correlation between the summer monsoon strength and the aerosol pollution level in the Yangtze River valley was reported. The upper level westerlies and jets may impact the boundary layer pollutant concentration by dynamically modulating kinetic momentum and vertical motion . Warmer tropical SSTs during El Niño-Southern Oscillation events lead to an anomalous anticyclonic circulation over the western Pacific. Discernable aerosol feedback in observations are reported. In northern China, the aerosol-wind covariation, which tends to strengthen air pollution, is observed in case studies (Yang, Fan, Leung, et al., 2016). The significant weakening of surface wind over eastern China during the past decades is likely related to the heavy air pollution there because such negative trends in wind speed cannot be found at mountain stations (X. Yang et al., 2013). The decrease in surface wind speeds over Asia may reduce dust aerosol emissions.

Impact of Combined Land Cover Change and Aerosols on Urban Climate
Urbanization is a worldwide trend that usually has two concurrent anthropogenic effects: (1) land cover change (e.g., Figure 7) and the associated urban heat island effect (UHI) and (2) air pollution or the urban aerosol effect. Zhong et al. (2015) conducted ensemble simulations with one-way three nested domains for the Greater Beijing Metropolitan Area using the WRF-Chem model coupled with a single-layer urban canopy model. The model spatial resolution in the innermost domain is 4 km, which can resolve most clouds and partly considers cloud-aerosol interactions. Four sets of sensitivity experiments were performed to quantify the relative importance of UHI, ARI, and ACI. High-resolution urban area data were updated from NASA satellite nighttime light products. Their results show that ACI plays a dominant role in the region by modifying the rainfall pattern as aerosol particles suppress convection and rainfall in the upwind area and increase convection and rainfall in the downwind area. The findings from this study were derived from a single case study. The influence of aerosols on precipitation may be dependent on the environmental conditions of the atmosphere, such as cloud height and phase, and the geographical locations of cities. Further investigations are needed to gain an in-depth understanding of the impact of elevated aerosol concentrations on clouds and precipitation in other cases or other metropolitan areas with different weather conditions (Zhong et al., 2017). A relatively long-term simulation is also needed to understand and quantify the roles of both land-cover change and aerosols in modifying regional climate and air quality (Zhong et al., 2018) for various synoptic systems.

Impact of Dust Aerosols on Climate
Besides anthropogenic aerosols, natural aerosols like dust, which could be contaminated anthropogenically, can also have a significant impact on  climate. The Taklimakan desert is one of the largest global sources of atmospheric dust. Ge et al. (2014) combined data from the Multi-angle Imaging Spectroradiometer (MISR) and the CALIOP to investigate the three-dimensional distribution of dust over the Taklimakan desert and surrounding areas. They found that the dust extinction coefficient rapidly decreases in the vertical in the spring, while dust particles are relatively well mixed vertically in the summer. Using the daily AOD products from MISR and comparing lowand high-dust days, they also identified two meteorological conditions favorable for entrainment of dust from the dry desert surface: vertical lofting and horizontal transport, that is, a strong anticyclonic wind anomaly over the Taklimakan desert at 500 hPa and an enhanced easterly wind over the Tarim Basin at 850 hPa. Huang et al. (2014) summarize the typical transport paths of East Asian dust, which affects regional and global climates, and discuss the numerous effects of dust aerosols on clouds and precipitation primarily over East Asian arid and semiarid regions. Compared with Saharan dust, East Asian dust absorbs more solar radiation, and its direct radiative forcing at the TOA is nearly positive or nil. This means that East Asian dust can influence cloud properties by not only acting as CCN and IN (via the first indirect, second indirect, and invigoration effects) but also through changing the relative humidity and stability of the atmosphere (via the semidirect effect). By converting visible light to thermal energy, dust aerosols can burn off cloud droplets, incurring a warming effect on climate, which is opposite to the first and second indirect effects of aerosols. The net dust aerosol radiative effects are still unclear. In addition, dust can inhibit or enhance precipitation under certain conditions, thus impacting the hydrological cycle. Over Asian arid and semiarid regions, the positive feedback loop in CAPI may aggravate drought in its interior. Su and Fung (2015) studied the sensitivities of the WRF-Chem to the Air Force Weather Agency (AFWA) and Shao et al. (2011) (i.e., S11) dust emission schemes, and to various land surface properties generated from United States Geological Survey (USGS) and Beijing Normal University (BNU) soil data over East Asia for spring 2012. The dust emissions generated with the S11 scheme are 2-5 times that generated with the AFWA emission scheme, with emissions ranging from 0.2 to 1 T g /day over East Asia during the study period. The AFWA emission scheme omits almost all of the Gobi desert and produces low dust emissions, whereas large dust emissions in this region are produced with the S11 emission scheme regardless of whether USGS or BNU soil data are used. The discrepancy between the performances of the AFWA and S11 emission schemes is mainly due to the underestimation of dust emissions over the Gobi desert by the AFWA scheme, which directly scales dust emissions based on the erodibility factor. This suggests that the erodibility factor for the Gobi desert is highly underestimated, highlighting an urgent need to improve the erodibility data set.

Concluding Remarks, Problems, and Future Directions
The changes in atmospheric composition, most notably the emissions of greenhouse gases and aerosols, have a significant impact on the climate system. Due to health concerns, the general public has increasingly paid more attention to severe pollution episodes and the degradation of air quality. Of importance is the strong coupling between the atmospheric environment and the climate system. Aerosols generally have a cooling effect, offsetting some of the warming induced by greenhouse gases. The former plays a much more complex role in the Earth's climate than does the latter. Radiative forcing by aerosols remains the largest source of uncertainties in simulating and predicting Earth's climate. To tackle this problem, studying the aerosol effects on climate in regions of the world characterized by heavy loadings of aerosols with diverse properties is critical. East Asia, especially China, is an ideal region for this effort. Three serial special sections focused on East Asian aerosols (Li et al., , 2011, and the current special section) in the Journal of Geophysical Research: Atmospheres attest to this.
The papers collected in this special section represent some recent progress on aerosols and their impacts on clouds, precipitation, and climate through aerosol-radiation interactions and aerosol-cloud interactions in the region. Understanding the mechanisms and identifying and quantifying the impacts are the main thrusts for the current special section, which involves two major topics, that is, the formation of aerosol particles and CAPI. Understanding and accounting for CAPI in the context of the complex Asian climate system involves the coupling of emissions, atmospheric physics and chemistry, and the dynamics of the Asian monsoon climate system. Many painstaking efforts have been devoted to untangling CAPI from a wide range of factors, 10.1029/2019JD030758

Journal of Geophysical Research: Atmospheres
leading to the identification and quantification of the roles of aerosols in changing surface temperature, clouds, precipitation, and monsoon circulation, among other factors. CAPI is found to be strongly contingent upon environmental conditions, that is, wind shear, stability, and humidity. Studies in this special section have revealed a wide range of climatic effects induced by aerosols through analyses of observational data acquired from both long-term meteorological observatories and short-term field experiments, and simulations by a hierarchy of models from cloud-resolving to global climate models.
Much effort has been made toward pollution abatement in China to improve air quality. As far as the aerosol direct radiative forcing is concerned, the improvement in air quality has tended to exacerbate the warming trend as the cooling effect of aerosols yields to global warming. During the twenty-first century, a higher frequency of hot summer days may ensue with accompanying changes in precipitation frequency and intensity, although the rainfall changes are much more uncertain than the temperature changes. As the rainfall variability increases, however, floods and prolonged droughts may increase. This would have a tremendous impact on properties and lives and on some critical sectors such as agriculture. Understanding the causes of these change and any connections between them are among the high priorities of environmental and climate sciences.

Uncertainties, Challenges, and Suggestions for Future Studies
Despite the tremendous progress made in our understanding of Asian aerosols and their climate effects, especially on clouds and precipitation, there still exist many uncertainties and challenges, from the observation of key variables to the better understanding of the complex CAPI processes. Understanding the true impact of aerosols on regional climate requires more thorough investigations as outlined here.
To make a leap in our understanding of CAPI and their broad impact on regional climate, it is important to obtain concurrent measurements of aerosol optical, physical, and chemical properties, CCN, cloud microphysical and dynamic properties, and their vertical profiles over a range of temporal and spatial scales. Aerosol properties have large spatiotemporal differences across Asia. With the dramatic development of industrialization and urbanization, aerosol emissions and properties have been changing quickly during the past 50 years in the region, which will continue in the foreseeable future. It is thus essential to build and maintain long-term observational networks for monitoring the spatiotemporal variation trends/patterns of aerosol optical properties, physical characteristics, chemical compositions, and aerosol gaseous precursors, as well as vertical distributions of aerosol and meteorological elements (e.g., Xin et al., 2015).
Major challenges in studying aerosol radiative effects are, among others, observing aerosol chemical compositions, which determines aerosol absorption, and water uptake, which modifies aerosol scattering. Measurements of aerosol size distributions are relatively easy to acquire. Changes in aerosol optical properties further make the quantification of aerosol radiative effects challenging. The response of air temperature to aerosol radiative effects is further complicated by the vertical distribution of aerosols, which is hard to measure. Air temperature is also affected by highly uncertain ACI. Elaborately designed field campaigns and numerical experiments are needed to identify the temperature changes caused by a specific type of aerosol and interaction mechanism. A key issue is to effectively distinguish the temperature response from the noisy, natural synoptic variability.
CCN and INP are essential to understanding CAPI. Despite the significant accomplishments achieved in studies regarding CCN, long-term measurements at more sites are needed to obtain more in situ CCN measurements and aerosol parameters used to estimate CCN in different atmospheric environments. This is because the ability for aerosols to serve as CCN and cloud droplets can vary greatly among regions with different atmospheric environments. For example, Fan, Liu, et al. (2018) have shown that ultrafine particles with diameters smaller than 50 nm can also serve as CCN in the Amazon region where the supersaturation is sometimes high enough to activate these very small particles, something once considered impossible to do. Currently, there is no single model that can predict CCN well under all environments. Models tend to underestimate the observed aerosol particle and CCN number concentrations (Fanourgakis et al., 2019). INP measurements in China are even less than CCN. A recent study indicated that anthropogenic aerosols in China may have enhanced ice formation by serving as INP and impacted cloud lifetime and radiative forcing (B. Zhao et al., 2019).
For any CAPI study, it is essential and most challenging to differentiate CAPI from other effects that are not physically related to aerosols but may covary with them (J. Fan et al., 2016;W. K. Tao et al., 2012). Separating the microphysical effects of aerosols from the dynamic effects of the atmosphere and surface is a general bottleneck, making the influence of aerosols on cloud properties particularly uncertain. Satellite-based retrievals of AOD have been widely used to study the influence of aerosols on clouds, but they have some inherent limitations such as large errors in estimating CCN from AOD (Andreae, 2009;. Comprehensive studies based on data from multiple platforms (surface, aircraft, and satellite; e.g.,  and model simulations are critically needed to investigate and differentiate the influences of various factors (e.g., CCN/INP, meteorology, swelling effects) for different types of aerosols and cloud regimes. Neglecting aerosol chemical composition and the ensuing swelling effect can lead to systematic biases in the estimate of ACI (Liu & Li, 2018b).
A possibility for disentangling meteorology and aerosol impacts on clouds is emerging with recently developed methodologies for retrieving cloud droplet number concentrations  and updrafts at the base of clouds (Zheng et al., 2015 from satellite observations. Having both cloud-base droplet number concentrations and updrafts allows calculating the cloud-base maximum supersaturation (S max ). This has made it possible to retrieve CCN(S) from satellites . By applying the CCN and updraft retrievals over oceans, it is possible to disentangle the effects of aerosols and meteorology there (Rosenfeld et al., 2019). Application of these methodologies with additional in situ validation over both land and oceans has the best potential for advancement, especially at climate-relevant scales in space and time.
The advent of new-generation geostationary meteorological satellite makes it possible to identify the life cycle of clouds in Asian monsoon regions . This may lead to a major leap in our understanding of the aerosol impact on cloud lifetimes, which is the least understood aerosol-cloud-interaction process. More importantly, clouds in the initial stage (cumulus stage) are more susceptible to aerosol pollution in the PBL, largely due to the coupling of the land surface and the atmosphere (Houze, 2018). As such, one of the top priorities of future CAPI studies should be the identification of the status of coupling between clouds and the boundary layer, as done by Zheng et al. (2018), as well as unraveling how clouds in the initial or developing stage respond to aerosols, which makes more sense compared with the bulk cloud responses.
There are still large uncertainties in modeling studies of the impact of aerosols on the monsoon climate, although much progress has been made in recent years . Current global climate models overestimate the indirect effect of aerosols, especially in highly polluted monsoon regions (e.g., East Asia and South Asia), which needs further examination using improved climate models and aerosol emission data sets. Furthermore, most current modeling studies are based on simulations with prescribed SSTs, where the slow response to aerosol forcing associated with SST changes is not considered. Fully coupled simulations are needed to include the air-sea coupling and feedback. To better understand the responses of the Asian monsoon climate to anthropogenic aerosols, the forcing-feedback paradigm needs to be accounted for in future investigations (Sherwood et al., 2014).
In Asia, there are some unique problems to tackle. First, aerosols are particularly inhomogeneous spatially, which may cause significant changes in precipitation by generating local circulations and associated changes in the spatiotemporal distributions of convective cells. It is necessary to account for interactions between aerosol advection and convective-cloud and precipitation processes in the process-level understanding of the effects of aerosol inhomogeneity on convective clouds and precipitation (W. K. Tao et al., 2012). High dust aerosol concentrations over East Asia lead to substantial effects on the energy budget and climate change through its interactions with clouds, radiation, and precipitation, a key process dictating the climate of semiarid regions in central Asia . However, the overall climatic effects of dust are still highly uncertain, particialrly in its ice-nucleating effect. More investigations on dust transport, size distribution, wavelength-dependent dust refractive index, and droplet and ice nucelating capabilities are crucially needed to account for their impact on the Asian climate.