Abstract

The WRF-simulated changes in clouds and climate due to the increased anthropogenic aerosols for the summers of 2002–08 (vs the 1970s) over eastern China were used to offline calculate the radiative forcings associated with aerosol–radiation (AR) and aerosol–cloud–radiation (ACR) interactions, which subsequently facilitated the interpretation of surface temperature changes. During this period, the increases of aerosol optical depth (ΔAOD) averaged over eastern China range from 0.18 in 2004 to 0.26 in 2007 as compared to corresponding cases in the 1970s, and the multiyear means (standard deviations) of AR and ACR forcings at the surface are −6.7 (0.58) and −3.5 (0.63) W m−2, respectively, indicating the importance of cloud changes in affecting both the aerosol climate forcing and its interannual variation. The simulated mean surface cooling is 0.35°C, dominated by AR and ACR with a positive (cooling) feedback associated with changes in meteorology (~10%), and two negative (warming) feedbacks associated with decreases in latent (~70%) and sensible (~20%) heat fluxes. More detailed spatial characteristics were analyzed using ensemble simulations for the year 2008. Three regions—Jing-Jin-Ji (ΔAOD ~ 0.63), Sichuan basin (ΔAOD ~ 0.31), and middle Yangtze River valley (ΔAOD ~ 0.26)—at different climate regimes were selected to investigate the relative roles of AR and ACR. While the AR forcing is closely related to ΔAOD values, the ACR forcing presents different regional characteristics owing to cloud changes. In addition, the surface heat flux feedbacks are also different between regions. The study thus illustrates that ACR forcing is useful as a diagnostic parameter to unravel the complexity of climate change to aerosol forcing over eastern China.

1. Introduction

Aerosol climate effects consist of the initial forcing that perturbs atmospheric radiation through both the direct radiative effect and the modulation of radiatively important cloud microphysics, and the subsequent changes in atmospheric states (e.g., instability and circulation) that could further affect cloud macro- and microphysics (Tao et al. 2012; Fan et al. 2016). These processes are not only sensitive to aerosol properties (e.g., concentration, refractive index, size distribution, and water solubility), but are also affected by the background climate settings like water vapor abundance and surface albedo. In particular, cloud adjustment effect, which includes both the aerosol–cloud microphysics interactions and the influences of meteorological changes, strongly depends on cloud types and atmospheric conditions (Fan et al. 2016), while the resultant radiative impacts also depend on the altitude where cloud changes occur (Stephens 2005). These complexities make aerosols a dominant contributor to uncertainties in current studies on climate changes (IPCC 2013). Therefore, it is essential to conduct model simulations focusing on specific regimes in order to better understand aerosol climate effects (Stevens and Feingold 2009).

East Asia has been a hot spot for studies of aerosol–climate interactions because of the unique aerosol and climate conditions in this region. The aerosol emissions in China increased rapidly in the past few decades due to the unprecedented industrialization and urbanization and only reached a peak recently, while aerosols decreased significantly in North America and western Europe due to strict emission regulations (Smith et al. 2011; Klimont et al. 2013; Li et al. 2017). To date, eastern China is still one of the most polluted regions on the globe (X. Zhang et al. 2012). The aerosol forcing and climate responses in this region are closely coupled with the East Asia monsoon, which not only affect the atmospheric aerosol loadings (Zhang et al. 2010), but also control the transport of water vapor that affects the aerosol optical properties (Li et al. 2014) and aerosol–cloud interactions. In addition, cloud types and vertical structures show strong spatial characteristics (Li et al. 2004) and interannual and intraseasonal variations (Guo and Zhou 2015; Pan et al. 2015) owing to the monsoon climate, which may induce unique aerosol–cloud interactions. Because of the intricate couplings between aerosols and monsoon systems, as well as influences from other systems like the western North Pacific subtropical high (WNPSH), climate responses to aerosol forcing in this region are very complex and warrant intensive investigations (Z. Li et al. 2016).

Previous studies have taken diverse approaches to highlight different aspects of aerosol–climate interactions over East Asia, such as the influences of aerosols on local climate (e.g., Liu et al. 2011), effects caused by different aerosol types (e.g., Jiang et al. 2013), aerosol impacts during strong and weak summer monsoon years (S. Li et al. 2016), and fast and slow responses of the summer monsoon to aerosol perturbations (Wang et al. 2017). Specifically, studies have been conducted to address the aerosol direct and microphysical effects, as well as the quantification of related radiative forcing. For example, the observational (e.g., Li et al. 2010) and modeling (e.g., H. Zhang et al. 2012) studies estimated the radiation perturbation caused by aerosol–radiation interactions; the studies by Huang et al. (2006) and Fan et al. (2013) studied aerosol–cloud microphysics interactions using regional climate and cloud-resolving model simulations, respectively. Most results show that more aerosols lead to smaller cloud droplets and increased cloudiness and liquid water, which, together with increased aerosols, reflect more solar radiation back into space. In response to the reduced surface solar radiation, the surface temperature decreases over the East Asia continent, which weakens the summer monsoon and reduces precipitation (e.g., Liu et al. 2011; Guo et al. 2013; Song et al. 2014). However, the spatial pattern of surface temperature changes does not always match that of aerosol loading (e.g., Tsai et al. 2016), which might be related to the cloud adjustment effect, as well as the adjustments of surface heat fluxes. To our knowledge, the relative roles of aerosol direct radiative effect and cloud adjustment effect in modulating the regional surface temperature have not been addressed yet.

Therefore, the motivation of this study is to quantify and compare the aerosol direct radiative and cloud adjustment effects over eastern China and use this knowledge to interpret the responses of the surface energy budget and temperature. Adjustments of latent heat (LH) and sensible heat (SH) fluxes are also discussed owing to their significant roles in damping the effect of radiation perturbations. To achieve these objectives, we analyze the WRF Model simulations of changes in clouds and climate caused by anthropogenic aerosol increases over the summers of 2002–08 (vs the 1970s) by Chen et al. (2018a). We choose the summer season because of large aerosol optical depth (AOD) as well as strong insolation and large amounts of water vapor and clouds, which can induce strong aerosol–radiation and aerosol–cloud interactions. These simulations have been used to investigate the intraseasonal responses of summer rainfall to increased aerosols (Chen et al. 2018a). This study is focused on the aerosol direct and cloud adjustment effects, which are examined by the offline calculation of radiation perturbations associated with aerosol–radiation (AR) and aerosol–cloud–radiation (ACR) interactions, respectively.

The rest of this paper is arranged as follows. Section 2 describes the data of WRF simulations including the aerosol forcing and associated cloud changes. Section 3 presents the method of offline radiation calculation and the comparisons between aerosol direct radiative and cloud adjustment effects. Section 4 shows the responses of surface energy balance and temperature to the aerosol radiative forcing. Section 5 gives the conclusions and discussion. For the results in sections 3 and 4, the interannual variations of eastern China mean are first shown, followed by detailed spatial characteristics from the year 2008 simulations.

2. Data description

a. WRF Model simulations

The study by Chen et al. (2018a) used an aerosol-aware WRF Model to simulate the East Asia summer (June–August) climate in 2002–08 with two aerosol loadings (control and sensitivity) that are representative of the current and the 1970s situations, respectively. Readers are referred to the paper for details on the model configuration and experiment setup, and only a brief description is given here.

The model reads in prescribed mass mixing ratios of multicomponent aerosol fields to simulate aerosol direct and cloud adjustment effects. Therein, aerosol radiative properties are calculated with subroutines imported from the WRF-Chem Model; the National Taiwan University (NTU) two-moment cloud microphysical scheme (Cheng et al. 2007, 2010; Chen et al. 2015; Chen and Wang 2016) calculates cloud nuclei numbers from the mass of soluble aerosol species based on empirical equations so that aerosols can affect the number and size of cloud hydrometeors and, thus, regulate cloud radiative properties. The NTU scheme includes a simple aerosol module to track interstitial and cloud-borne aerosol masses, which allows cloud nuclei (CN) to be depleted by cloud and precipitation processes and recover to the prescribed level when the grid cell is cloud free. In this way, CN removal due to clouds and precipitation is accounted for to some degree, avoiding nucleating too many droplets. Cloud fraction is diagnosed using the scheme by Xu and Randall (1996), where subgrid cloudiness is calculated with grid-mean relative humidity and masses of cloud water and cloud ice. This scheme predicts 100% cloudiness when the grid cell is saturated with respect to water or ice, so it might overestimate high-level clouds where high supersaturation is not uncommon (e.g., Gierens et al. 2000; Spichtinger et al. 2003; Khvorostyanov et al. 2006). Other major physical parameterization schemes used in the simulations are shown in Table 1, and the cumulus radiative feedback is deactivated. The horizontal resolution is 30 km, and vertically there are 50 levels with the model top at 50 hPa.

Table 1.

Major physical parameterization schemes used in the WRF simulations (Chen et al. 2017).

Major physical parameterization schemes used in the WRF simulations (Chen et al. 2017).
Major physical parameterization schemes used in the WRF simulations (Chen et al. 2017).

With the current horizontal resolution, microphysical parameterization can only resolve large-scale stratiform clouds that dominate cloud radiative effect, whereas convective clouds smaller than the grid size, which have small influences on radiation due to short lifetimes, are excluded. Meanwhile, note that the sun–Earth orbit and greenhouse gas concentrations are fixed in the simulations, and radiative processes above the level of 50 hPa are ignored, as anthropogenic aerosols are located mainly in the lower troposphere. These simplifications should have negligible effects on the results.

For each summer, a control case and a sensitivity case were conducted, which were representative of aerosol loadings in the present day (polluted) and in the 1970s (clean), respectively. In the control cases, aerosol fields were from the CAM-Chem aerosol data (Lamarque et al. 2012) and calibrated by AOD (550 nm) observed by the Aqua MODIS instrument (v051, Level 3, monthly). The CAM-Chem aerosol data provide monthly mixing ratios of 14 components for five aerosol species, including sulfate, black carbon (BC), organic carbon (OC), sea salt, and dust. Because the AOD calculated using the CAM-Chem data shows large deviations from satellite observations (Lamarque et al. 2012), the aerosol concentrations in each column were tuned by multiplying a factor (constant in the vertical direction) to constrain the column AOD with the MODIS observation. In other words, MODIS AOD provides information of aerosol horizontal distribution, while the CAM-Chem data provide detailed aerosol vertical distribution and compositions (Chen et al. 2018a). The sensitivity case had identical aerosol fields as the corresponding control case, except that the concentrations of anthropogenic aerosols (sulfate, BC, and OC) were reduced by 75%, considering that the SO2 emission rate in the 2000s over East Asia is about 4 times that in the 1970s (Smith et al. 2011; Wang et al. 2015). Except for the aerosol fields, the control and sensitivity cases are identically configured. This study is focused on the effects of aerosol increases, and to exclude the effects of temporal variability of aerosol perturbation, the aerosol loadings in both simulations are kept constant as the summer mean (June–August) throughout the simulation period.

There is one simulation for each case (control and sensitivity) of 2002–07 and three ensembles for each case of 2008. All simulations were driven with meteorological forcing from ERA-Interim (Dee et al. 2011) and the daily SST from the NCEP RTG_SST analysis data (Thiébaux et al. 2003), and no reinitializing or nudging was imposed. The WRF simulation data provide the vertical profiles of atmosphere and cloud properties and surface climate states including SH and LH fluxes every 3 h. Note that the simulation domain of Chen et al. (2018a) includes a part of South Asia in addition to East Asia, but analyses in this study hereafter are focused on eastern China (23°–43°N, 105°–122°E; see the bounded region in Fig. 2a) only. Below, we analyze results from all years to show the interannual variations and then concentrate on the ensemble-mean results of the year 2008 to emphasize spatial characteristics.

b. Changes in aerosol optical depth

Compared with the sensitivity case, AOD in the control case is increased by 0.23 on average over eastern China during 2002–08, with a range between 0.18 in 2004 and 0.26 in 2007 (see Table 3). The interannual variation is due to the variations in anthropogenic aerosol emissions and meteorology (e.g., circulation, precipitation, and humidity). The AOD increase (ΔAOD) occurs all over eastern China with large values over North China (peak over 0.5) and the Yangtze River valley (Fig. 1a), where large cities and industries are densely sited. The increase of AOD is larger than estimates based on AOD retrieved from ground station visibility data (Wu et al. 2014), which shows increases of ~0.1 in the annual-mean AOD in North China and middle to lower portions of the Yangtze River from the 1970s to 2000s, so aerosol forcing in this study might be overestimated to some degree. The interannual variation of ΔAOD has a similar spatial pattern to the mean of this period. Despite having the same regional-mean ΔAOD of 0.25 over eastern China in 2002 and 2008, the spatial distributions of the ΔAOD are very different (Figs. 1b,c): the year 2002 has a larger ΔAOD in the Yangtze River valley and South China and a smaller ΔAOD in North China as compared to the 2002–08 mean, while the year 2008 has an opposite spatial pattern. The regional difference in ΔAOD between these two years can reach up to 0.15. The different spatial patterns of aerosol forcing can induce different climate responses, which will be discussed in section 4a.

Fig. 1.

Changes of aerosol optical depth (ΔAOD) at 550 nm caused by aerosol increases (control minus sensitivity; color shading): (a) the mean of 2002–08 and the anomalies of (b) 2002 and (c) 2008 relative to the mean of 2002–08. Contours in (a) indicate the interannual standard deviation of AOD changes.

Fig. 1.

Changes of aerosol optical depth (ΔAOD) at 550 nm caused by aerosol increases (control minus sensitivity; color shading): (a) the mean of 2002–08 and the anomalies of (b) 2002 and (c) 2008 relative to the mean of 2002–08. Contours in (a) indicate the interannual standard deviation of AOD changes.

As the spatial characteristics of radiative forcing and climate responses in our study are based on the 2008 simulations, here we show the spatial pattern of the AOD from the 2008 control case and its difference from the sensitivity case (Fig. 2). In the control case, the AOD value averaged over eastern China is 0.55, with peaks larger than 1.0 over the North China plain. The ΔAOD is 0.25 over eastern China, which is around half of the control AOD, and peaks in similar regions as the AOD in the control case. The primary contributor to ΔAOD (more than 55%) is the increase in sulfate aerosols (contours in Fig. 2b).

Fig. 2.

(a) AOD at 550 nm from the 2008 control case and (b) its difference from the 2008 sensitivity case (ΔAOD; control minus sensitivity). Contours in (b) indicate the contribution of sulfate to ΔAOD. The number in the top right of each panel is the regional-mean value over eastern China [23°–43°N, 105°–122°E; red box in (a)]. Grids over the ocean and Taiwan are excluded here and calculations afterward.

Fig. 2.

(a) AOD at 550 nm from the 2008 control case and (b) its difference from the 2008 sensitivity case (ΔAOD; control minus sensitivity). Contours in (b) indicate the contribution of sulfate to ΔAOD. The number in the top right of each panel is the regional-mean value over eastern China [23°–43°N, 105°–122°E; red box in (a)]. Grids over the ocean and Taiwan are excluded here and calculations afterward.

c. Changes of cloud properties caused by aerosol increases

Table 2 summarizes the summer-mean cloud properties averaged over eastern China from the 2008 control and sensitivity cases. The total cloud fraction in eastern China is 68% in control, which is slightly smaller than the International Satellite Cloud Climatology Project (ISCCP) observation (74%) and presents a similar spatial distribution (figure not shown). High clouds dominate the simulated cloud fraction, while low and middle clouds have relatively small fractions. This is mostly due to the coarse horizontal resolution (30 km), which can only resolve the radiatively important large-scale stratus clouds, while the small-sized cumulus clouds in lower levels could not be captured. Compared with the sensitivity case, the control case has a larger cloud droplet number Nc, a smaller cloud droplet effective radius Re (decreased by 2.6 μm), and a larger liquid water path (LWP; increased by 2.5 g m−2), which are consistent with the general understanding of aerosol indirect effects. The model also simulates a small increase in low- and high-cloud fractions and a decrease in middle-cloud fraction. The increases in LWP and low-cloud fraction are more than 7%, compared to sensitivity. Note that all these changes should be attributed to not only the aerosol microphysical effects, but also the changes in circulation caused by aerosol increases as discussed in the following section. The changes are mostly small for the regional mean over eastern China, except for cloud droplet number, but they show remarkable spatial variations associated with large-scale meteorological changes.

Table 2.

Comparisons of cloud properties averaged over eastern China between control and sensitivity cases of 2008 summer. The total, low-, mid-, and high-level cloud fractions are calculated with the maximum/random overlap assumption (identical to the overlap assumptions used for radiation calculations in WRF simulations and offline evaluations), and low-, mid-, and high-level clouds are defined as clouds below 680 hPa, between 680 and 440 hPa, and above 440 hPa, respectively. The column-mean cloud droplet number concentration Nc and effective radius Re are weighted with cloud liquid water content. The observed total cloud fraction from ISCCP is about 74%.

Comparisons of cloud properties averaged over eastern China between control and sensitivity cases of 2008 summer. The total, low-, mid-, and high-level cloud fractions are calculated with the maximum/random overlap assumption (identical to the overlap assumptions used for radiation calculations in WRF simulations and offline evaluations), and low-, mid-, and high-level clouds are defined as clouds below 680 hPa, between 680 and 440 hPa, and above 440 hPa, respectively. The column-mean cloud droplet number concentration Nc and effective radius Re are weighted with cloud liquid water content. The observed total cloud fraction from ISCCP is about 74%.
Comparisons of cloud properties averaged over eastern China between control and sensitivity cases of 2008 summer. The total, low-, mid-, and high-level cloud fractions are calculated with the maximum/random overlap assumption (identical to the overlap assumptions used for radiation calculations in WRF simulations and offline evaluations), and low-, mid-, and high-level clouds are defined as clouds below 680 hPa, between 680 and 440 hPa, and above 440 hPa, respectively. The column-mean cloud droplet number concentration Nc and effective radius Re are weighted with cloud liquid water content. The observed total cloud fraction from ISCCP is about 74%.

Figure 3 presents the spatial distribution of changes in cloud micro- and macrophysical properties from the 2008 simulations. For cloud microphysics, the aerosol increases lead to larger Nc and smaller Re over nearly all of eastern China (Figs. 3a,b) by acting as cloud nuclei. The pattern of ΔNc is consistent with that of ΔAOD (Fig. 2b). The changes of cloud fraction at low and middle levels and high levels exhibit similar spatial patterns (Figs. 3e,f): both show increases over the northwest flank of the WNPSH and decreases over the rest of eastern China, except that low and middle clouds are also increased over southeast coastal China. The changes in LWP (Fig. 3c) are consistent with the changes in low- and middle-cloud fraction. In contrast, ice water path in the control case is larger than the sensitivity case over all of eastern China, which could be because more cloud droplets are elevated to higher levels to form cloud ice due to smaller droplet sizes.

Fig. 3.

JJA-mean changes of cloud properties caused by aerosol increases from 2008 simulations (control minus sensitivity): (a) column-averaged in-cloud cloud droplet number concentration Nc (cm−3), (b) column-averaged cloud droplet radius Re (μm), (c) liquid water path (g m−2), (d) ice water path (g m−2), and cloud fraction at (e) low and middle and (f) high levels. Dotted areas in (e) and (f) indicate regions where water vapor content is increased at the respective levels, and vectors superimposed in (e) represent horizontal circulation changes at 850 hPa (m s−1).

Fig. 3.

JJA-mean changes of cloud properties caused by aerosol increases from 2008 simulations (control minus sensitivity): (a) column-averaged in-cloud cloud droplet number concentration Nc (cm−3), (b) column-averaged cloud droplet radius Re (μm), (c) liquid water path (g m−2), (d) ice water path (g m−2), and cloud fraction at (e) low and middle and (f) high levels. Dotted areas in (e) and (f) indicate regions where water vapor content is increased at the respective levels, and vectors superimposed in (e) represent horizontal circulation changes at 850 hPa (m s−1).

The pattern of changes in cloud fraction and LWP greatly differs from that of aerosol increases but is consistent with changes in water vapor and circulation (hatching and vectors in Figs. 3e,f). As shown by Chen et al. (2018a; cf. their Fig. 8), the increases of anthropogenic aerosols cool the surface and cause a westward extension of the WNPSH. This, on the one hand, induces anomalous southwesterlies to transport more water vapor and surface air convergence at the northwest flank of WNPSH (Fig. 3e), which cause increased cloud fraction and precipitation at 30°–35°N, and on the other hand induces surface air divergence over southeastern China reducing clouds and precipitation. The westward extension of WNPSH has been noticed by observational studies (e.g., Zhou et al. 2009), and Chen et al. (2018b) further confirmed its link with anthropogenic aerosol climate effects. The increase of cloud fraction over southeastern coastal regions is caused by the increased atmospheric stability due to the surface cooling and mainly occurs in low-level clouds.

3. Comparisons between AR and ACR radiative forcing

This section compares the radiative forcing associated with AR and ACR interactions. The approach to offline calculate radiative forcing is first described. Then, AR and ACR radiative forcing averaged over eastern China during 2002–08 is presented to give a general perspective and show the interannual variations, and a more detailed spatial pattern is illustrated using the 2008 simulations.

a. Offline radiation calculation

In the model simulations, the radiative effects of changes in aerosols, clouds, and meteorology are coupled together and difficult to separate. Thus, we developed an offline radiation calculation package to evaluate the contributions of AR and ACR to the radiation perturbations, with the direct radiative effect of meteorological changes (e.g., temperature and humidity) excluded. The offline radiation calculation package is based on RRTMG radiation schemes in WRF, and the process is identical to the WRF online radiation calculation, so its results can be compared with the WRF output. This package reads profiles of atmospheric state properties (temperature, pressure, and humidity), cloud properties (fraction, hydrometeor mass, and size), and surface properties (solar albedo, longwave emissivity, and temperature) from the WRF dataset (3 hourly) and aerosol loadings. At each model level, subroutines from the WRF-Chem Model are used to calculate aerosol radiative properties (optical depth, single-scatting albedo, and asymmetry factor) with aerosol loadings and relative humidity. The RRTMG radiation scheme calculates cloud radiative properties (same parameters as aerosols) using cloud hydrometeor mass and size, and then the radiative transfer with output of radiation fluxes at the top of the atmosphere and the surface.

Three sets of offline calculations were conducted to separate the aerosol radiative forcing into radiation perturbations caused by change of aerosol–radiation interactions only (AR) and by cloud changes in response to aerosol increases only (ACR). The first reads all data from the control case to provide a reference; the second (third) reads aerosols (clouds) from the sensitivity case and the remaining input data from the control case. Thus, the radiation fluxes under all-sky from the first minus those from the second (third) gives the forcing contributed by the AR (ACR) interaction with fixed meteorology fields. The direct radiative and cloud adjustment effects are calculated every 3 h for each pair of WRF simulations, and the analyses are focused on the summer-mean (JJA) radiative forcing at the surface to facilitate the interpretation of surface temperature changes.

b. Interannual variations

Table 3 shows the year-to-year variation of ΔAOD and offline-calculated AR and ACR radiative forcings at the surface. The ΔAOD has mean and standard deviation of 0.23 and 0.03, respectively. Increased AOD reduces the shortwave (SW) radiation reaching the surface and causes surface radiative cooling of −6.7 W m−2. This magnitude is smaller than the estimated aerosol direct radiative effect of −15.7 W m−2 from Li et al. (2010) because they considered the radiative effects of all aerosols in the atmosphere. The AR forcing in individual years is consistent with ΔAOD and has a standard deviation of ~10%. The total radiative forcing [SW plus longwave (LW); −3.5 W m−2] caused by cloud changes (ACR) is also negative, dominated by the SW effect (−3.2 W m−2), and is consistent with the cloud changes as described in section 2c: larger cloud fraction and optical depth (owing to larger LWP and smaller Re) reflect more SW radiation and reduce SW reaching the surface. The small magnitude of ACR forcing, compared to that of AR forcing, is mainly due to the small changes in cloud fraction and liquid water path. The estimation of ACR forcing is similar to those from previous studies [e.g., −3.2 W m−2 from Huang et al. (2006) and −6.8 W m−2 from Fan et al. (2013)], but with some differences due to different assumptions and variations of study periods and regions. Meanwhile, the ACR forcing has a larger standard deviation of ~20% than the AR forcing and presents different year-to-year variations from ΔAOD. For instance, the largest ACR forcing occurs in 2004, when the smallest ΔAOD occurs. This indicates that the changes of cloud properties play important roles in affecting both the aerosol radiative forcing and its interannual variation.

Table 3.

Interannual variations of aerosol forcing and climate responses at the surface during 2002–08. The aerosol forcing includes changes in AOD at 550 nm and offline-calculated radiative forcings at the surface (W m−2) associated with AR and ACR interactions. The WRF-simulated responses at the surface include changes in radiation (RAD; net SW + downward LW), LH, and SH fluxes (W m−2) and changes in surface temperature Ts (°C). Downward energy fluxes have positive values. Mean and standard deviation σ for the period are also given.

Interannual variations of aerosol forcing and climate responses at the surface during 2002–08. The aerosol forcing includes changes in AOD at 550 nm and offline-calculated radiative forcings at the surface (W m−2) associated with AR and ACR interactions. The WRF-simulated responses at the surface include changes in radiation (RAD; net SW + downward LW), LH, and SH fluxes (W m−2) and changes in surface temperature Ts (°C). Downward energy fluxes have positive values. Mean and standard deviation σ for the period are also given.
Interannual variations of aerosol forcing and climate responses at the surface during 2002–08. The aerosol forcing includes changes in AOD at 550 nm and offline-calculated radiative forcings at the surface (W m−2) associated with AR and ACR interactions. The WRF-simulated responses at the surface include changes in radiation (RAD; net SW + downward LW), LH, and SH fluxes (W m−2) and changes in surface temperature Ts (°C). Downward energy fluxes have positive values. Mean and standard deviation σ for the period are also given.

The interannual variations of regional characteristics in AR and ACR forcings are illustrated in Fig. 4, showing the standard deviations of summer-mean AR and ACR forcings at the surface during 2002–08. While the AR forcing has smaller variations and a pattern similar to the ΔAOD distribution (see Fig. 1a), the ACR forcing presents larger variations and also a pattern different from ΔAOD, indicating the sensitivity of ACR to climate regime. This aspect will be addressed in detail in sections 3c and 4b for the year 2008. We also examined the interannual variations of changes in cloud properties and found that the pattern of ACR variation is similar to the patterns of variations in LWP and low-cloud-fraction changes (figures not shown). This suggests that the interannual variation of ACR forcing can be largely attributed to different responses in LWP and low cloud fraction within different atmospheric states.

Fig. 4.

Interannual standard deviation of aerosol-caused JJA-mean radiative forcings (W m−2) associated with (a) AR and (b) ACR interactions at the surface during 2002–08.

Fig. 4.

Interannual standard deviation of aerosol-caused JJA-mean radiative forcings (W m−2) associated with (a) AR and (b) ACR interactions at the surface during 2002–08.

c. Spatial characteristics

The online results of spatial patterns in AR and ACR forcings were analyzed and discussed by Chen et al. (2018a), so here only the offline results are given. Figure 5 shows the spatial characteristics of total radiative forcings (SW + LW) associated with AR and ACR interactions using the ensemble-mean results of the 2008 simulations. For AR, the total radiative forcing is the same as the SW radiative forcing because the LW forcing is negligible. The AR and ACR forcings caused by anthropogenic aerosol increases respectively decrease the total radiation at the surface by 6.9 and 3.5 W m−2, indicating the dominance of AR interactions in the regional mean. Meanwhile, AR and ACR forcings exhibit different spatial patterns. First, the AR radiative cooling is spatially in line with ΔAOD (Fig. 2b) and has peaks over the Jing-Jin-Ji area, whereas the ACR cooling is mostly significant over the northwest flank of the WNPSH caused by large increases in cloud fraction and LWP (Figs. 3c,e). Second, in contrast to negative AR forcing all over eastern China, the ACR radiative forcing is positive over some regions caused by SW effect. For example, over the middle Yangtze River valley, the low- and middle-cloud fraction and LWP are decreased, and consequently more SW radiation can reach the surface. It can be inferred that the effect of decreased low- and middle-cloud fraction and LWP over these regions overwhelms the effect of smaller Re.

Fig. 5.

Spatial distribution of total radiative forcing (SW + LW; W m−2) at the surface associated with (a) AR and (b) ACR interactions from the 2008 simulations. The boxes show boundaries of three regions: the Jing-Jin-Ji area (37°–40°N, 115°–119°E; box A), the Sichuan basin (28°–31°N, 103°–107°E; box B), and the middle Yangtze River valley (27°–31°N, 112°–115°E; box C).

Fig. 5.

Spatial distribution of total radiative forcing (SW + LW; W m−2) at the surface associated with (a) AR and (b) ACR interactions from the 2008 simulations. The boxes show boundaries of three regions: the Jing-Jin-Ji area (37°–40°N, 115°–119°E; box A), the Sichuan basin (28°–31°N, 103°–107°E; box B), and the middle Yangtze River valley (27°–31°N, 112°–115°E; box C).

4. Responses in surface energy balance and temperature

a. Interannual variations

To interpret the surface temperature response, we analyzed the changes of surface energy fluxes (radiation, LH, and SH) because the surface temperature changes to keep surface energy balance. In the Noah land surface scheme (Chen and Dudhia 2001) used in WRF simulations of Chen et al. (2018a), the surface is warmed by the net (downward minus reflected) SW and downward LW radiation and cooled by the emission of LW radiation, LH, and SH fluxes. Therefore, the surface heat fluxes and temperature have to respond accordingly to offset the aerosol-induced radiation changes. The changes of surface energy balance can be expressed as

 
formula

where and are the LW emissivity and SW reflectivity of the surface, respectively; is the Stefan–Boltzmann constant; and are the downward shortwave and longwave radiation fluxes reaching the surface, respectively; LH, SH, and G are the outgoing latent heat, sensible heat fluxes to the atmosphere, and ground flux to the subsurface, respectively; and implies the control case minus the sensitivity case. Note that besides AR and ACR interactions, radiation changes here also include the effect of meteorological changes (e.g., changes of atmospheric temperature and humidity) as part of the climate response. The change of ground flux is very small and thus omitted in our results.

The model-simulated changes in surface energy fluxes and temperature are listed in Table 3. Note that downward fluxes are assigned with positive values, and thus positive LH and SH changes represent decreases in the magnitude of LH and SH fluxes, which transport energy from the land surface to the atmosphere in both control and sensitivity cases. It is clear that the cooling associated with the decreases in radiation at the surface (net SW plus downward LW; RAD) is significantly compensated by the decreases of LH and SH fluxes. For example, for the year 2003, 77% of RAD changes (−12.3 W m−2) caused by ΔAOD (0.24) are offset by the decreases of LH (7.9 W m−2) and SH (1.6 W m−2). Note that the surface radiative cooling is dominated by the decreases of downward solar flux associated with AR and ACR interactions, but it is 1.1 W m−2 cooler than the sum of AR and ACR forcings because of responses in meteorology (e.g., LW cooling effect due to decreased atmospheric temperature).

In response to the negative surface energy perturbation [right part of Eq. (1)], the surface temperature decreases in all years, with mean and standard deviation values of −0.35° and 0.11°C, respectively. It is worth noting that although the aerosol forcing in 2005 (ΔAOD ~0.21) is larger than that in 2004 (ΔAOD ~ 0.18), the magnitude of surface cooling is smaller in 2005 (−0.17°C) than that in 2004 (−0.33°C). This is attributed to the small ACR forcing and strong LH responses in 2005, as well as different meteorological changes. The different responses of surface temperature to ΔAOD also demonstrate the significance of the spatial pattern of aerosol forcing and the regional characteristics. For example, with nearly identical magnitudes of ΔAOD in 2002 and 2008, the surface temperature changes differ by almost a factor of 2, with values of −0.47° and −0.25°C, respectively. This appears to be related to different spatial patterns of ΔAOD (Figs. 1b,c), as well as different regional surface climate responses (e.g., LH and SH) associated with various surface characteristics.

b. Spatial characteristics

Figure 6 presents detailed spatial distributions of changes in surface radiation, LH, SH, and temperature during the summer of 2008. As in Table 3, positive values mean downward fluxes, and thus positive LH and SH changes represent decreases in the magnitude of upward heat fluxes, and vice versa. The decrease of radiation flux is most significant in the North China plain and the western regions (Fig. 6a), corresponding to the peaks of AR and ACR forcings, respectively. Meanwhile, LH decreases significantly over regions with large radiation decreases because of less energy for evaporation, and also over South China (Fig. 6b). There is an LH increase over the Loess Plateau. The significant LH decrease in the middle Yangtze River valley and increase in Loess Plateau are consistent with the changes in soil moisture (figure not shown). The SH flux decreases at 30°–40°N (Fig. 6c), with centers corresponding to the surface temperature cooling centers. This is because the SH highly depends on the vertical temperature gradient near the surface, and the temperature decreases more significantly at the surface than in the atmosphere. In the end, the changes of surface temperature (Fig. 6d) coincide well with the total surface energy perturbation (sum of RAD, LH, and SH; figure not shown). It has three cooling centers at 30°–40°N: the Jing-Jin-Ji area, Sichuan basin, and Loess Plateau. The first two centers can be attributed to the significant radiative cooling, while the third centeris due to the larger LH flux caused by more precipitation and larger surface soil moisture. In contrast, the surface temperature change over large parts of South China is relatively small as the radiative cooling is mostly offset by the decrease of LH. The temperature increase in the middle Yangtze River valley will be discussed later. The different spatial patterns between surface temperature response and initial ΔAOD suggest that the cloud adjustment effect, as well as the changes in surface heat flux, is important in modulating the surface temperature response.

Fig. 6.

JJA-mean changes in surface energy fluxes and temperature caused by aerosol increases from the year 2008 simulations (control minus sensitivity): (a) radiation fluxes (RAD; net SW + downward LW), (b) LH, (c) SH, and (d) surface temperature Ts. Positive values mean downward fluxes. Grids over lakes are omitted in (b) because the land surface model does not consider lakes.

Fig. 6.

JJA-mean changes in surface energy fluxes and temperature caused by aerosol increases from the year 2008 simulations (control minus sensitivity): (a) radiation fluxes (RAD; net SW + downward LW), (b) LH, (c) SH, and (d) surface temperature Ts. Positive values mean downward fluxes. Grids over lakes are omitted in (b) because the land surface model does not consider lakes.

Based on the spatial patterns of AR and ACR forcings, three regions are selected to address the contrast between AR and ACR forcings and surface climate responses in different climate regimes (boundaries shown in Fig. 5a). These three regions are located at different positions relative to the WNPSH and have distinct climate characteristics. The Jing-Jin-Ji area (region A), part of the North China plain, is located at the north flank of the WNPSH and is heavily polluted; the Sichuan basin (region B) is located at the northwest flank of the WNPSH; and the middle Yangtze River valley (region C) is over humid South China and is occupied by the WNPSH in summer.

Table 4 gives aerosol forcing and surface climate responses averaged over these three regions. In the Jing-Jin-Ji area, the AR cooling (−16.1 W m−2) is significant because of the large ΔAOD (0.63), whereas the ACR cooling is relatively small (−1.5 W m−2) due to small cloud changes. In the Sichuan basin, the ACR cooling plays a more important role and is comparable to the AR cooling (−8.8 vs −8.6 W m−2) because of the significant increases in cloud fraction and LWP, which are caused by the increased water vapor and anomalous convergent flows as discussed in section 2c. For both regions, the radiative cooling is strong, while negative feedbacks from decreased LH and SH offset the cooling by up to 80%, which leaves negative surface energy perturbation of more than −3 W m−2 and thus decreased surface temperatures of 0.43° and 0.51°C, respectively. In the middle Yangtze River valley, where the WNPSH dominates, the ACR presents a slight warming (0.2 W m−2) due to the slight decrease in cloud fraction and liquid water (Figs. 3c,e) associated with the strengthened WNPSH, whereas the AR enhancement (−8.1 W m−2) still dominates owing to the large ΔAOD (0.26). Meanwhile, there is significant warming (12.3 W m−2) from LH decrease due to reduced radiation and decrease in soil moisture caused by less precipitation (−2.8 mm day−1), which overwhelms the radiative cooling and yields a surface warming of 0.17°C. The decrease of precipitation is because of the surface air divergence associated with the WNPSH intensification, as illustrated by Chen et al. (2018a).

Table 4.

Aerosol forcing and climate responses at the surface over three regions in 2008 summer (control minus sensitivity): the Jing-Jin-Ji area (box A in Fig. 5a), the Sichuan basin (box B in Fig. 5a), and the middle Yangtze River valley (box C in Fig. 5a). The energy fluxes are decomposed into three parts (RAD, LH, and SH) as in Table 3, with positive values indicating downward fluxes.

Aerosol forcing and climate responses at the surface over three regions in 2008 summer (control minus sensitivity): the Jing-Jin-Ji area (box A in Fig. 5a), the Sichuan basin (box B in Fig. 5a), and the middle Yangtze River valley (box C in Fig. 5a). The energy fluxes are decomposed into three parts (RAD, LH, and SH) as in Table 3, with positive values indicating downward fluxes.
Aerosol forcing and climate responses at the surface over three regions in 2008 summer (control minus sensitivity): the Jing-Jin-Ji area (box A in Fig. 5a), the Sichuan basin (box B in Fig. 5a), and the middle Yangtze River valley (box C in Fig. 5a). The energy fluxes are decomposed into three parts (RAD, LH, and SH) as in Table 3, with positive values indicating downward fluxes.

The meridional distribution of aerosol-induced changes in surface energy fluxes and temperature from the year 2008 simulations is presented in Fig. 7 (thick lines). The decrease of radiation flux change is more than 10 W m−2 to the south of 40°N and peaks at 34°N. As a response, there is significant anomalous downward LH flux toward the surface to the south of 35°N, implying a decrease in the magnitude of upward LH flux, compared to the control case. The LH flux is sensitive to the radiation perturbation in this region because of large surface water content and strong evaporation. This also implies that the surface temperature is less sensitive to the radiation perturbation in the southern regions because of the strong negative feedbacks from LH. The meridional distribution of SH flux is similar to that of surface temperature changes, and SH changes also act to damp the temperature decrease. The negative total energy perturbation at the surface (RAD + SH + LH) is very small to the south of 30°N and to the north of 40°N and is largest in the middle, yielding surface cooling with similar meridional variation.

Fig. 7.

Meridional distributions of JJA-mean changes in surface energy fluxes (W m−2) and surface temperature (black; °C) zonally averaged between 105° and 122°E. The energy fluxes include radiation (net SW + downward LW; RAD; blue), LH (yellow), SH (purple), and summation of the three (red), where positive values mean downward fluxes. The thick lines are for the ensemble mean of 2008 simulations, while the thin lines and shading represent the mean and the range of plus and minus one standard deviation, respectively, for simulations of 2002–08.

Fig. 7.

Meridional distributions of JJA-mean changes in surface energy fluxes (W m−2) and surface temperature (black; °C) zonally averaged between 105° and 122°E. The energy fluxes include radiation (net SW + downward LW; RAD; blue), LH (yellow), SH (purple), and summation of the three (red), where positive values mean downward fluxes. The thick lines are for the ensemble mean of 2008 simulations, while the thin lines and shading represent the mean and the range of plus and minus one standard deviation, respectively, for simulations of 2002–08.

Also shown in Fig. 7 are the mean and standard deviation of the zonal-mean changes of surface energy and temperature during 2002–08 (thin lines and the shading). The multiyear-mean results resemble the year 2008 results. The interannual variations of energy fluxes and temperature are small at low and high latitudes and large between 30° and 40°N, suggesting that the aerosol climate effects have large uncertainties at midlatitudes. This is in line with the large climate variability at midlatitudes. Note that the 2008 results stay in the range of one standard deviation from the 2002–08 mean, which improves our confidence in the analyses of the 2008 simulations.

Because it takes about 6 weeks for the soil moisture to reach equilibrium in the simulations, we also examined the results averaged over the last 6 weeks of the simulations. It shows a smaller magnitude of LH changes, but the ratio of LH changes to radiation changes and the spatial characteristics of LH changes are similar to the summer-mean results above.

5. Conclusions and discussion

Based on WRF Model–simulated cloud and climate changes caused by anthropogenic aerosol increases, this study first quantifies and compares the radiative forcing of aerosol direct radiative and cloud adjustment effects using offline-calculated radiation perturbations associated with AR and ACR interactions, and then uses them to facilitate the interpretation of changes in surface energy balance and temperature over eastern China. Highlights of the findings are summarized below.

  1. An increase of AOD between 0.18 (2004) and 0.26 (2007) occurred during 2002–08 over eastern China. The mean radiative forcing of AR during this period is −6.7 W m−2 with a range between −5.5 (2004) and −7.2 W m−2 (2007), whereas ACR has a forcing of −3.5 W m−2 with a range between −2.4 (2005) and −4.2 W m−2 (2004). The negative ACR forcing is consistent with smaller cloud droplet size and, more importantly, increases in LWP and low- and middle-cloud fraction caused by changes in meteorology. Thus, changes of cloud properties play important roles in affecting both the aerosol radiative forcing and its interannual variation.

  2. The cloud adjustment effect shows a spatial pattern very different from the aerosol direct radiative effect and dominates in some regions. The surface radiative forcing of AR is cooling all over eastern China, corresponding to AOD increase with maximum in the Jing-Jin-Ji area, whereas the forcing of cloud adjustment effect has strongest cooling in the western regions due to large increases in cloud fraction and water, and small warming in the middle Yangtze River valley because of small decreases in clouds. This implies that cloud adjustment effect is important in modulating the spatial pattern of aerosol radiative forcing.

  3. Decreases in LH and SH fluxes compensate the surface radiative cooling by ~70% and ~20%, respectively, resulting to a surface cooling of −0.35°C. The pattern of surface temperature changes is different from the initial aerosol forcing, which clearly illustrates the roles of cloud adjustment effect and surface heat flux changes in modulating the spatial pattern of surface temperature changes.

This study has drawn some conclusions concerning the aerosol perturbation on the surface energy fluxes and temperature in eastern China summer. Three important issues must be emphasized. First, while AR, which is directly related to aerosol loading, still dominates the surface cooling for the mean of eastern China, the ACR, which accounts for cloud responses, is important on the regional changes and the interannual variations. The close relationships between the ACR effect and the changes in cloud fraction and liquid water that are dominated by the large-scale circulation responses imply that a better understanding of the changes in atmospheric states, in particular the circulation and humidity, in response to aerosol forcing is warranted. Second, the negative feedbacks of LH and SH fluxes play important roles in modifying the effects of AR and ACR interactions, especially for the spatial pattern of surface temperature response. Clearly, the representation of surface characteristics and the parameterization of land surface processes in the models could introduce substantial uncertainties in quantifying the aerosol climate effects. Third, it is worthy of further investigation using higher horizontal resolution and more realistic treatment of CN depletion and replenishment and convection invigoration.

Acknowledgments

This study is part of Song's doctoral dissertation at SUNYA. The research was supported by grants from the Office of Sciences (BER), U.S. Department of Energy (Grant DEFG0292ER61369), and U.S. National Science Foundation (1545917) in support of the Partnership for International Research and Education project at the University at Albany.

REFERENCES

REFERENCES
Chen
,
F.
, and
J.
Dudhia
,
2001
:
Coupling an advanced land surface–hydrology model with the Penn State–NCAR MM5 modeling system. Part I: Model implementation and sensitivity
.
Mon. Wea. Rev.
,
129
,
569
585
, https://doi.org/10.1175/1520-0493(2001)129<0569:CAALSH>2.0.CO;2.
Chen
,
G.
, and
W.-C.
Wang
,
2016
:
Aerosol–stratocumulus–radiation interactions over the southeast Pacific: Implications to the underlying air–sea coupling
.
J. Atmos. Sci.
,
73
,
2759
2771
, https://doi.org/10.1175/JAS-D-15-0277.1.
Chen
,
G.
,
W.-C.
Wang
, and
J.-P.
Chen
,
2015
:
Aerosol–stratocumulus–radiation interactions over the southeast Pacific
.
J. Atmos. Sci.
,
72
,
2612
2621
, https://doi.org/10.1175/JAS-D-14-0319.1.
Chen
,
G.
,
J.
Yang
,
Q.
Bao
, and
W.-C.
Wang
,
2018a
:
Intraseasonal responses of the East Asia summer rainfall to anthropogenic aerosol climate forcing
.
Climate Dyn.
,
51
,
3985
3998
, https://doi.org/10.1007/s00382-017-3691-0.
Chen
,
G.
,
W.-C.
Wang
, and
J.-P.
Chen
,
2018b
:
Circulation responses to regional aerosol climate forcing in summer over East Asia
.
Climate Dyn.
,
51
,
3973
3984
, https://doi.org/10.1007/s00382-018-4267-3.
Cheng
,
C.-T.
,
W.-C.
Wang
, and
J.-P.
Chen
,
2007
:
A modelling study of aerosol impacts on cloud microphysics and radiative properties
.
Quart. J. Roy. Meteor. Soc.
,
133
,
283
297
, https://doi.org/10.1002/qj.25.
Cheng
,
C.-T.
,
W.-C.
Wang
, and
J.-P.
Chen
,
2010
:
Simulation of the effects of increasing cloud condensation nuclei on mixed-phase clouds and precipitation of a front system
.
Atmos. Res.
,
96
,
461
476
, https://doi.org/10.1016/j.atmosres.2010.02.005.
Dee
,
D. P.
, and Coauthors
,
2011
:
The ERA-Interim reanalysis: Configuration and performance of the data assimilation system
.
Quart. J. Roy. Meteor. Soc.
,
137
,
553
597
, https://doi.org/10.1002/qj.828.
Fan
,
J.
,
L. R.
Leung
,
D.
Rosenfeld
,
Q.
Chen
,
Z.
Li
,
J.
Zhang
, and
H.
Yan
,
2013
:
Microphysical effects determine macrophysical response for aerosol impacts on deep convective clouds
.
Proc. Natl. Acad. Sci. USA
,
110
,
E4581
E4590
, https://doi.org/10.1073/pnas.1316830110.
Fan
,
J.
,
Y.
Wang
,
D.
Rosenfeld
, and
X.
Liu
,
2016
:
Review of aerosol–cloud interactions: Mechanisms, significance, and challenges
.
J. Atmos. Sci.
,
73
,
4221
4252
, https://doi.org/10.1175/JAS-D-16-0037.1.
Gierens
,
K.
,
U.
Schumann
,
M.
Helten
,
H.
Smit
, and
P. H.
Wang
,
2000
:
Ice-supersaturated regions and subvisible cirrus in the northern midlatitude upper troposphere
.
J. Geophys. Res.
,
105
,
22 743
22 753
, https://doi.org/10.1029/2000JD900341.
Guo
,
L.
,
E.
Highwood
,
L.
Shaffrey
, and
A.
Turner
,
2013
:
The effect of regional changes in anthropogenic aerosols on rainfall of the East Asian summer monsoon
.
Atmos. Chem. Phys.
,
13
,
1521
1534
, https://doi.org/10.5194/acp-13-1521-2013.
Guo
,
Z.
, and
T.
Zhou
,
2015
:
Seasonal variation and physical properties of the cloud system over southeastern China derived from CloudSat products
.
Adv. Atmos. Sci.
,
32
,
659
670
, https://doi.org/10.1007/s00376-014-4070-y.
Hong
,
S.-Y.
,
Y.
Noh
, and
J.
Dudhia
,
2006
:
A new vertical diffusion package with an explicit treatment of entrainment processes
.
Mon. Wea. Rev.
,
134
,
2318
2341
, https://doi.org/10.1175/MWR3199.1.
Huang
,
Y.
,
R. E.
Dickinson
, and
W. L.
Chameides
,
2006
:
Impact of aerosol indirect effect on surface temperature over East Asia
.
Proc. Natl. Acad. Sci. USA
,
103
,
4371
4376
, https://doi.org/10.1073/pnas.0504428103.
Iacono
,
M. J.
,
J. S.
Delamere
,
E. J.
Mlawer
,
M. W.
Shephard
,
S. A.
Clough
, and
W. D.
Collins
,
2008
:
Radiative forcing by long-lived greenhouse gases: Calculations with the AER radiative transfer models
.
J. Geophys. Res.
,
113
, D13103, https://doi.org/10.1029/2008JD009944.
IPCC
,
2013
: Climate Change 2013: The Physical Science Basis. Cambridge University Press, 1535 pp., https://doi.org/10.1017/CBO9781107415324.
Jiang
,
Y.
,
X.
Liu
,
X.-Q.
Yang
, and
M.
Wang
,
2013
:
A numerical study of the effect of different aerosol types on East Asian summer clouds and precipitation
.
Atmos. Environ.
,
70
,
51
63
, https://doi.org/10.1016/j.atmosenv.2012.12.039.
Kain
,
J. S.
,
2004
:
The Kain–Fritsch convective parameterization: An update
.
J. Appl. Meteor.
,
43
,
170
181
, https://doi.org/10.1175/1520-0450(2004)043<0170:TKCPAU>2.0.CO;2.
Khvorostyanov
,
V. I.
,
H.
Morrison
,
J. A.
Curry
,
D.
Baumgardner
, and
P.
Lawson
,
2006
:
High supersaturation and modes of ice nucleation in thin tropopause cirrus: Simulation of the 13 July 2002 Cirrus Regional Study of Tropical Anvils and Cirrus Layers case
.
J. Geophys. Res.
,
111
,
D02201
, https://doi.org/10.1029/2004JD005235.
Klimont
,
Z.
,
S. J.
Smith
, and
J.
Cofala
,
2013
:
The last decade of global anthropogenic sulfur dioxide: 2000–2011 emissions
.
Environ. Res. Lett.
,
8
,
014003
, https://doi.org/10.1088/1748-9326/8/1/014003.
Lamarque
,
J. F.
, and Coauthors
,
2012
:
CAM-chem: Description and evaluation of interactive atmospheric chemistry in the Community Earth System Model
.
Geosci. Model Dev.
,
5
,
369
411
, https://doi.org/10.5194/gmd-5-369-2012.
Li
,
C.
, and Coauthors
,
2017
:
India is overtaking China as the world’s largest emitter of anthropogenic sulfur dioxide
.
Sci. Rep.
,
7
,
14304
, https://doi.org/10.1038/s41598-017-14639-8.
Li
,
J.
,
W.-C.
Wang
,
Z.
Sun
,
G.
Wu
,
H.
Liao
, and
Y.
Liu
,
2014
:
Decadal variation of East Asian radiative forcing due to anthropogenic aerosols during 1850–2100, and the role of atmospheric moisture
.
Climate Res.
,
61
,
241
257
, https://doi.org/10.3354/cr01236.
Li
,
S.
,
T.
Wang
,
F.
Solmon
,
B.
Zhuang
,
H.
Wu
,
M.
Xie
,
Y.
Han
, and
X.
Wang
,
2016
:
Impact of aerosols on regional climate in southern and northern China during strong/weak East Asian summer monsoon years
.
J. Geophys. Res. Atmos.
,
121
,
4069
4081
, https://doi.org/10.1002/2015JD023892.
Li
,
Y.
,
R.
Yu
,
Y.
Xu
, and
X.
Zhang
,
2004
:
Spatial distribution and seasonal variation of cloud over China based on ISCCP data and surface observations
.
J. Meteor. Soc. Japan
,
82
,
761
773
, https://doi.org/10.2151/jmsj.2004.761.
Li
,
Z.
,
K. H.
Lee
,
Y.
Wang
,
J.
Xin
, and
W. M.
Hao
,
2010
:
First observation-based estimates of cloud-free aerosol radiative forcing across China
.
J. Geophys. Res.
,
115
,
D00K18
, https://doi.org/10.1029/2009JD013306.
Li
,
Z.
, and Coauthors
,
2016
:
Aerosol and monsoon climate interactions over Asia
.
Rev. Geophys.
,
54
,
866
929
, https://doi.org/10.1002/2015RG000500.
Liu
,
X.
,
X.
Xie
,
Z.-Y.
Yin
,
C.
Liu
, and
A.
Gettelman
,
2011
:
A modeling study of the effects of aerosols on clouds and precipitation over East Asia
.
Theor. Appl. Climatol.
,
106
,
343
354
, https://doi.org/10.1007/s00704-011-0436-6.
Monin
,
A.
, and
A.
Obukhov
,
1954
:
Basic laws of turbulent mixing in the surface layer of the atmosphere
.
Contrib. Geophys. Inst. Acad. Sci. USSR
,
151
,
e187
.
Pan
,
Z.
,
W.
Gong
,
F.
Mao
,
J.
Li
,
W.
Wang
,
C.
Li
, and
Q.
Min
,
2015
:
Macrophysical and optical properties of clouds over East Asia measured by CALIPSO
.
J. Geophys. Res. Atmos.
,
120
,
11 653
11 668
, https://doi.org/10.1002/2015JD023735.
Smith
,
S. J.
,
J.
van Aardenne
,
Z.
Klimont
,
R. J.
Andres
,
A.
Volke
, and
S.
Delgado Arias
,
2011
:
Anthropogenic sulfur dioxide emissions: 1850–2005
.
Atmos. Chem. Phys.
,
11
,
1101
1116
, https://doi.org/10.5194/acp-11-1101-2011.
Song
,
F.
,
T.
Zhou
, and
Y.
Qian
,
2014
:
Responses of East Asian summer monsoon to natural and anthropogenic forcings in the 17 latest CMIP5 models
.
Geophys. Res. Lett.
,
41
,
596
603
, https://doi.org/10.1002/2013GL058705.
Spichtinger
,
P.
,
K.
Gierens
, and
W.
Read
,
2003
:
The global distribution of ice-supersaturated regions as seen by the Microwave Limb Sounder
.
Quart. J. Roy. Meteor. Soc.
,
129
,
3391
3410
, https://doi.org/10.1256/qj.02.141.
Stephens
,
G. L.
,
2005
:
Cloud feedbacks in the climate system: A critical review
.
J. Climate
,
18
,
237
273
, https://doi.org/10.1175/JCLI-3243.1.
Stevens
,
B.
, and
G.
Feingold
,
2009
:
Untangling aerosol effects on clouds and precipitation in a buffered system
.
Nature
,
461
,
607
613
, https://doi.org/10.1038/nature08281.
Tao
,
W. K.
,
J. P.
Chen
,
Z. Q.
Li
,
C.
Wang
, and
C. D.
Zhang
,
2012
:
Impact of aerosols on convective clouds and precipitation
.
Rev. Geophys.
,
50
,
RG2001
, https://doi.org/10.1029/2011RG000369.
Thiébaux
,
J.
,
E.
Rogers
,
W.
Wang
, and
B.
Katz
,
2003
:
A new high-resolution blended real-time global sea surface temperature analysis
.
Bull. Amer. Meteor. Soc.
,
84
,
645
656
, https://doi.org/10.1175/BAMS-84-5-645.
Tsai
,
I. C.
,
W.-C.
Wang
,
H.-H.
Hsu
, and
W.-L.
Lee
,
2016
:
Aerosol effects on summer monsoon over Asia during 1980s and 1990s
.
J. Geophys. Res. Atmos.
,
121
,
11 761
11 776
, https://doi.org/10.1002/2016JD025388.
Wang
,
Y.
,
J. H.
Jiang
, and
H.
Su
,
2015
:
Atmospheric responses to the redistribution of anthropogenic aerosols
.
J. Geophys. Res. Atmos.
,
120
,
9625
9641
, https://doi.org/10.1002/2015JD023665.
Wang
,
Z.
,
L.
Lin
,
M.
Yang
,
Y.
Xu
, and
J.
Li
,
2017
:
Disentangling fast and slow responses of the East Asian summer monsoon to reflecting and absorbing aerosol forcings
.
Atmos. Chem. Phys.
,
17
,
11 075
11 088
, https://doi.org/10.5194/acp-17-11075-2017.
Wu
,
J.
,
J.
Luo
,
L.
Zhang
,
L.
Xia
,
D.
Zhao
, and
J.
Tang
,
2014
:
Improvement of aerosol optical depth retrieval using visibility data in China during the past 50 years
.
J. Geophys. Res. Atmos.
,
119
,
13 370
13 387
, https://doi.org/10.1002/2014JD021550.
Xu
,
K.-M.
, and
D. A.
Randall
,
1996
:
A semiempirical cloudiness parameterization for use in climate models
.
J. Atmos. Sci.
,
53
,
3084
3102
, https://doi.org/10.1175/1520-0469(1996)053<3084:ASCPFU>2.0.CO;2.
Zhang
,
H.
, and Coauthors
,
2012
:
Simulation of direct radiative forcing of aerosols and their effects on East Asian climate using an interactive AGCM-aerosol coupled system
.
Climate Dyn.
,
38
,
1675
1693
, https://doi.org/10.1007/s00382-011-1131-0.
Zhang
,
L.
,
H.
Liao
, and
J.
Li
,
2010
:
Impacts of Asian summer monsoon on seasonal and interannual variations of aerosols over eastern China
.
J. Geophys. Res.
,
115
,
D00K05
, https://doi.org/10.1029/2010JD014030.
Zhang
,
X.
,
Y.
Wang
,
T.
Niu
,
X.
Zhang
,
S.
Gong
,
Y.
Zhang
, and
J.
Sun
,
2012
:
Atmospheric aerosol compositions in China: Spatial/temporal variability, chemical signature, regional haze distribution and comparisons with global aerosols
.
Atmos. Chem. Phys.
,
12
,
779
799
, https://doi.org/10.5194/acp-12-779-2012.
Zhou
,
T.
, and Coauthors
,
2009
:
Why the western Pacific subtropical high has extended westward since the late 1970s
.
J. Climate
,
22
,
2199
2215
, https://doi.org/10.1175/2008JCLI2527.1.

Footnotes

© 2019 American Meteorological Society. For information regarding reuse of this content and general copyright information, consult the AMS Copyright Policy (www.ametsoc.org/PUBSReuseLicenses).