We estimate the contributions of natural and anthropogenic forcings to the observed summer cooling over the Eastern China through a set of controlled experiments of an atmospheric general circulation model. The control run is forced by historical natural and anthropogenic agents, as well as by the Hadley Center global sea surface temperature. The model suggests a weak influence by the solar radiation change on this cooling, but a significant contribution by the sulfate aerosol to this cooling, through both dynamic and thermal processes. The inclusion of the sulfate aerosol induces a positive gradient of air temperature in the middle-upper troposphere, which results in a northward shift in the 200 hPa East Asian westerly jet stream and an increase of the East Asian summer monsoon, leading to more cloud cover and precipitation in the Eastern China therefore surface cooling over the region.
If you can't find a tool you're looking for, please click the link at the top of the page to "Go to old article view". Alternatively, view our Knowledge Base articles for additional help. Your feedback is important to us, so please let us know if you have comments or ideas for improvement.
 While the major concern nowadays is global warming, a surface cooling trend from spring to summer has been observed in the Central Eastern China (27–36°N, 102.5–122.5°E) since 1950s [Li et al., 1995; Chen et al., 1998; Hu et al., 2003]. The springtime cooling has been attributed to the decadal variability related to the North Atlantic Oscillation [Yu and Zhou, 2004; Li et al., 2005; Xin et al., 2006], but explanations on the mechanisms responsible for the summer cooling have not been agreed on. Since cooling trends are generally associated with wet trends, the summer cooling may be partly caused by excessive rainfall due to the recent warming in the tropical Pacific and Indian Oceans [Nitta and Hu, 1996; Gong and Ho, 2002; Hu et al., 2003; Zhou and Huang, 2003]. Other studies emphasized contributions by natural and decadal variability. Sultan and Gaofa  believed that the solar-cycle was responsible for the climate variation along the lower-middle Yangtze River Valley, while Yu et al.  showed that the strong tropospheric cooling trend in the East Asia during July–August resulted in the tendency towards increased droughts in the Northern China and flood in the Yangtze River Valley. Modeling studies also indicated that the summer cooling might be partly caused by the increase in atmospheric aerosol loadings [Menon et al., 2002; Qian et al., 2003]. The negative radiative effect of sulfate aerosols over the Eastern China was suggested to exceed the warming effect of greenhouse gasses (GHGs) in summer [Xu, 2001]. The most recent study on the variations of surface air temperature (SAT) over China based on nineteen coupled climate models of the Intergovernmental Panel on Climate Change (IPCC) 4th Assessment Report (AR4) found that few models could reproduce the cooling trend even if both natural and anthropogenic forcings including sulfate aerosols were prescribed [Zhou and Yu, 2006]. The causes for this bias are highly complex, including the bias of ocean-atmosphere coupled models, other processes and interaction among them. As the first step, we use an atmospheric general circulation model (AGCM) forced by observed sea surface temperature (SST), natural (solar irradiance) and anthropogenic (GHGs and sulfate aerosols) forcings, according to the coordinated experiment of Phase II of the Climate Variability and Predictability (CLIVAR) International Climate of the Twentieth Century Project (C20C) [Folland et al., 2002]. The goal of this study is to estimate the contributions of both natural and anthropogenic forcings to the summer cooling over the Eastern China by comparing the results under different forcings.
2. Model and Simulation Strategy
 The model used in this study is the Grid Atmospheric Model of IAP LASG, Version 1.1 (hereinafter GAMIL1.1). Its key dynamical part includes a finite difference scheme that satisfies conservation laws of total mass and effective energy for solving the primitive hydrostatic equations of baroclinic atmosphere [Wang et al., 2004], and a two-step shape-preserving advection scheme for the moisture equation [Yu, 1994]. The model employs a hybrid horizontal grid, with Gaussian grid of 2.8° between 65.58°S and 65.58°N and weighted equal-area grid poleward of 65.58°. There are 26 vertical σ-layers from surface to 2.19 hPa. It mainly uses the physical parameterizations from National Center for Atmospheric Research (NCAR) Community Atmospheric Model, Version 2 (CAM2) [Collins et al., 2003], except for a modified Tiedtke convective scheme [Li et al., 2007]. Only the direct effects of aerosols on shortwave flux and heating rate are considered; in another word, there is no treatment of the radiative effects of aerosols on longwave flux and heating rate [Collins et al., 2003]. The GHGs effects on longwave parameterization are included through specified concentration distributions [Kiehl et al., 1996].
 The C20C provides a standard set of forcing fields [Folland et al., 2002]. The anthropogenic forcings include four kinds of GHGs (CO2, CH4, N2O and CFC) and sulfate aerosols (SUL) in the form of concentration data from the Boucher data set [Boucher and Pham, 2002]. Note the anthropogenic forcing varies with time. The natural forcing is referred to the solar irradiance (SOL). The time series of perturbations to the solar constant is taken from Lean et al. . The Hadley Center global SST data [Rayner et al., 2003] and the CRUTS2.1 SAT data are also used [Mitchell and Jones, 2005] as boundary conditions over oceans and for model validation purposes, respectively.
 One control run (CTL) and three sensitivity experiments are carried out (Table 1). The integrations cover the period from 1949 to 2002, and our analysis focuses on the period between 1951 and 2000. The CTL is the standard coordinated experiment of the C20C Phase II, which uses monthly SST, GHGs, SUL and SOL changes covering 1949–2002. The sulfate concentration is updated decade by decade, both the spatial and temporal changes of the sulfate concentration has been included. The sensitive experiments exclude GHGs changes (ExpA), GHGs changes and SUL (ExpB), anthropogenic forcings and SOL (ExpC), respectively. In ExpC, the GHGs concentrations and solar constant are fixed at the 1950's values (see Table 1 for details).
GHG, greenhouse gases; SUL, sulfate aerosol direct effects; SOL, solar irradiance. “Yes” denotes the inclusion of a time-varying forcing.
Figure 1 shows linear trends of summer (June–August) SAT from (a) CRUTS2.1, (b) CTL run and (c) the associated net surface solar radiation. Linear regression is used to estimate the trends, and signals shorter than 5 years are filtered. The CRUTS2.1 shows obvious cooling trends in the Central Eastern China and warming trends elsewhere (Figure 1a), consistent with meteorological station data [Hu et al., 2003; Zhou and Yu, 2006]. The CTL run agrees with the observation to some extent, although there is a northward shift in the simulated cooling (Figure 1b). The cooling trends are generally collocated with negative net surface solar radiations (Figure 1c), indicating a direct response of the SAT to radiation budget.
 The results of CTL run show a combined effect of GHGs, SUL and SOL on the cooling. In order to see the contribution of a specific agent, the differences between CTL and ExpA, ExpA and ExpB, and ExpB and ExpC are used to assess the role of GHGs, SUL, and SOL, respectively. It is not surprising to find that the inclusion of GHGs forcing warms most part of the Eastern China, except for the region along the coastline (Figure 2a). The coastal cooling is attributed to increased cloud cover and rainfall associated with strengthened easterly winds (not shown), which reflects more solar radiation and transports more latent heat through evaporation.
 The direct effect of SUL induces a cooling over the whole China, with an amplitude exceeding −1°C/50 yr in the central area (Figure 2b), matching the centers of net surface solar radiation (Figure 2e). The SOL changes, however, resulted in a pattern of “cooling in the north and warming in the south” in SAT response (Figure 2c), opposite to Figure 2b in sign south to 30°N. Since the centers of net surface solar radiation response (Figure 2f) do not exactly match those of SAT (Figure 2c), the contribution of SOL change may be explained via dynamic processes.
 The annual global solar radiation forcing due to anthropogenic SUL at the top of the atmosphere (TOA) in the GAMIL1.1 is −0.298 WM−2, which falls into the range of −0.26 ∼ −0.4 WM−2 by most AGCMs [Myhre et al., 1998]. The averaged reflection of solar radiation by sulfate aerosols over the Central Eastern China at TOA is about −2 WM−2 in summer, close to −2.7 WM−2 based on a coupled regional model [Qian and Giorgi, 1999; Giorgi et al., 2003], while it is 0.63 WM−2 for GHGs and −0.1 WM−2 for SOL. Hence, at the regional scale the effect of SUL is the same as, or slightly greater than, the sum of GHGs and SOL [Qian and Giorgi, 1999].
 Further analyses find that the inclusion of sulfate aerosols induces a positive meridional air temperature difference between 500 and 200 hPa (Figure 3a), warmer in the south and colder in the north. Previous study notes that the meridional temperature difference of the troposphere is responsible for the location of the East Asian Westerly Jet (EAWJ) [Zhang et al., 2006]. Any change in the troposphere temperature would result in corresponding change of the EAWJ and hence the East Asian Summer monsoon via momentum convergence [Yu et al., 2004]. This mechanism also applies here, as shown by the northward shift of the upper-level East Asian westerly jet stream and the enhancement of East Asian summer monsoon. Associated with the increased meridional wind over the Eastern China and the intensified western Pacific subtropical high, strong rising motions can be seen over the regions north to 30°N (Figures 3b and 3c). Consequently, excessive rainfall exists over the Northern and Southern China (Figure 3d). This leads to an enhanced surface cooling through more reflection and increased surface evaporation over these regions, which could be another potential reason for the northward shift of the simulated surface cooling by CTL (Figure 1b), in addition to possible model bias and interactions between sulfate aerosols and other forcings. In contrast to the dynamic response to SUL, the “warming in the south” pattern in Figure 2c indicates a stronger land-ocean thermal contrast, which leads to an increase of southerly winds over the Southern China. Correspondingly, over the Central and Northern China the “cooling in the north” causes a decrease of meridional winds (Figure 3e), which results in stronger rising motions over these regions and increases subsidence in the Southern China (Figure 3f). Hence, less cloud cover and precipitation are seen over the Southern China. A reversed condition is seen over the Northern China.
4. Summary and Discussion
 AGCM experiments are carried out to isolate the causes for the observed summer cooling over the Eastern China, using different historical natural and anthropogenic forcings. They reveal a significant contribution of the sulfate aerosol to the cooling, and a weak influence of the solar radiation change on the cooling. Both thermal and dynamic responses to different forcings are analyzed. A negative feedback over the Central China and a positive feedback over the Northern and Southern China occur, mostly in response to precipitation changes. The positive meridional gradient of air temperature averaged over the middle-upper troposphere is induced by the inclusion of sulfate aerosols, which results in a northward shift in the 200 hPa East Asian westerly jet stream and an increase of the East Asian summer monsoon, leading to less cloud cover and precipitation over the middle-lower Yangtze River Valley, more in the Eastern China therefore surface cooling over the region. The increased GHGs induce a warming over most part of East Asia. The feedback of increased rainfall along the Eastern China coastline also results in a weak cooling.
 Based on above analysis, we find that both anthropogenic and natural forcings contribute to the summer cooling over the Central Eastern China through thermal and dynamic processes. Our new results support those based on a coupled regional model but through analysis of different dynamical process [Qian and Giorgi, 1999; Qian et al., 2003]; they also support statistical correlation between surface climate change over the Eastern China and solar cycle [Sultan and Gaofa, 1994; Chen et al., 1998].
 Due to the limited experiments we have carried out so far, the assessment on the effect of a specific forcing is incomplete. First, different backgrounds may alter the relative effect of SUL and SOL comparing with that of GHGs. For example, both ExpB and ExpC do not consider the anthropogenic forcings (GHGs and SUL); therefore the potential interactions between solar constant and these agents are not included. Similarly, the assessment of the SUL' effect might not be as accurate as it could be if we had included GHGs. Second, the semi-direct and indirect radiative effect of aerosols is not included. Many studies indicate that the negative radiative forcing due to these indirect effects might be larger than the direct ones [Qian and Giorgi, 1999, 2000; Boucher and Pham, 2002], and that the positive radiative forcing of black carbon particles should not be negligible. These could potentially cause the northward shift of the simulated cooling in CTL run, due to the lack of the dynamic response of less rainfall in the Northern China [Menon et al., 2002; Huang et al., 2007]. Last, all the natural and anthropogenic external forcings are partly included in the observed SST, our AGCM's boundary condition. More experiments are being carried out to advance this research.
 This work is supported by the 973 Project (grant 2005CB321703), the National Natural Science Foundation of China (grant 40221503), and the Chinese Academy of Sciences International Partnership Creative Group, entitled “The Climate System Model Development and Application Studies.” The model integration is performed on the Lenovo DeepComp 6800 Supercomputer at the supercomputing Center of the Chinese Academy of Sciences. This work also contributes to the CLIVAR C20C project.