The tropospheric atmosphere chemistry model (TACM), including a condensed gaseous chemical mechanism and an inorganic aerosol thermodynamic equilibrium submodel, was coupled with the regional climate model (RegCM3) to build a Regional Climate Chemistry Modeling System (RegCCMS), which was applied to investigate the spatial and temporal distribution of anthropogenic nitrate aerosol, radiative forcing, and also its climatic effect over China. Modeling results show that the annual average surface concentration and column burden of nitrate are 2.19 μg/m3 and 5.05 mg/m2, respectively. The countrywide annual average direct radiative forcing, first indirect radiative forcing, and total radiative forcing at the top of atmosphere induced by nitrate are −0.88 W/m2, −2.47 W/m2, and −2.52 W/m2, respectively. Numerical experiments indicate that surface air temperature decreases and precipitation reduces when nitrate aerosol is included in the coupled modeling system. Changes in annual surface air temperature due to direct effect, first indirect effect, second indirect effect, and combined effect are −0.04°C, −0.11°C, −0.68°C and −0.78°C, respectively. The corresponding precipitation reduction is −0.05 mm/d, −0.10 mm/d, −0.42 mm/d, and −0.52 mm/d. Variations of surface temperature and precipitation due to the combined effect are less than the sum of individual direct, first indirect, and second indirect effect, showing a strong nonlinearity between radiative forcing and climate change. The indirect effect of nitrate is stronger than its direct effect. These preliminary results suggest that nitrate aerosol has unignorable effect on regional climate in China compared to sulfate and carbonaceous aerosols.
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.
 Nitrate is a typical secondary inorganic aerosol which is chemically formed in atmosphere and its precursors are nitrogen oxides (NOx = NO + NO2), dinitrogen pentoxide (N2O5), ammonia (NH3) and nitric acid (HNO3). Existing in both fine mode and coarse mode [Kadowaki, 1977], nitrate is one of the important components of atmospheric aerosols. In recent years, economic expansion and urbanization in China have resulted in significant growth in NOx and NH3 emissions. Anthropogenic NOx emissions associated with fossil fuel burning and NH3 emission associated with agricultural activities have grown as there experienced a period of rapid economic development and industrial/agricultural expansion in the last three decades [Streets and Waldhoff, 2000; Ohara et al., 2007]. The increase of NOx and NH3 emissions leads to the rise of nitrate proportion in the fine particles as PM2.5. In some cities, content of nitrate in the aerosols even has exceeded that of sulfate.
 Similar to the other types of aerosols, nitrate also has influence on climate through both direct and indirect ways. The direct effect refers to that aerosols change the net solar radiative flux and affect the radiative balance of Earth-atmosphere system by scattering or absorbing solar radiation. The indirect effect includes two aspects. The first indirect effect means that being cloud condensation nuclei (CCN), aerosol increases the number of cloud droplets and reduce the effective radius of cloud droplets, thus affecting the cloud optical thickness and scattering properties [Twomey, 1974]. The second indirect effect refers to the decrease of precipitation and increase of cloud lifetime because of the reduction of cloud droplet effective radius and coagulation [Albrecht, 1989]. Veefkind et al.  suggested that nitrate existing in the form of ammonium was a very important aerosol species in the optics active submicron size, suggesting its strong ability in scattering solar radiation and affecting cloud properties. Both IPCC TAR [Intergovernmental Panel on Climate Change (IPCC), 2001] and AR4 [IPCC, 2007] pointed out that the role of nitrate aerosol on climate and climate change should not be neglected.
 Some studies on global radiative forcing and climatic effect of nitrate aerosol have been carried out in recent years [Adams et al., 2001; van Dorland et al., 1997; Jacobson, 2001; Liao and Seinfeld, 2005; Bauer et al., 2007]. IPCC AR4 [IPCC, 2007] estimated that the direct forcing of nitrate was in the range of −0.10 ± 0.10 W/m2 and pointed out the large uncertainty due to the lack of related research. China has been the primary contributor of NOX emission in East Asia [Streets et al., 2003]. Therefore, more attention had been paid to the environmental effect of tropospheric nitrate [An et al., 2002; Zhang and Han, 2003; Zhang et al., 2007]. However, few studies have been found focusing on the direct and indirect climatic effect of nitrate over China, through more works are conducted concerning sulfate and carbonaceous aerosols. Therefore, it is necessary to quantitatively evaluate the radiative forcing and climatic effect on tropospheric nitrate aerosol over China.
 In this paper, a coupled regional climate chemistry modeling system (RegCCMS) is set up to simulate the spatial distribution and radiative forcing of anthropogenic nitrate aerosol and to investigate its climatic effect on regional climate over China.
 The dynamical core of RegCM3 is based on the hydrostatic version of the fifth-generation Pennsylvania State University–National Center for Atmospheric Research (PSU–NCAR) Mesoscale Model. It includes a large-scale cloud and precipitation scheme that accounts for the subgrid-scale variability of clouds, new parameterizations for ocean surface fluxes, and a cumulus convection scheme. The CCM3 radiative transfer package was used in this model, which includes new features such as the effect of additional greenhouse gases (N2O, CH4, CFCs), atmospheric aerosols, and cloud ice.
2.2. Tropospheric Atmosphere Chemistry Model
 The tropospheric atmosphere chemistry model (TACM) contains complex atmospheric physical and chemical processes affecting the distribution of air pollutants, including gases and inorganic aerosols. The thermodynamic equilibrium model ISORROPIA was coupled with TACM to deal with the volatile nitrate and other inorganic aerosols. The gas chemistry model was used to describe transformation of precursors for nitrate formation. The dry deposition submodel uses a three layer resistance analogy method to compute deposition velocities of gases and aerosols. The aqueous chemistry and wet scavenging submodel consists of a 1-D stationary cloud physics model and a cloud-rain chemistry model, the latter considers not only the soluble gas absorptions by cloud and its precipitation but also the aqueous oxidation for SO2 and NOx. The rainout and washout of aerosols are parameterized in terms of precipitation intensity. More information can be found in the previous work of T. Wang et al. [2003, 2004a] and Li et al. .
2.2.1. Thermodynamic Equilibrium Model
 The largest difference between nitrate and sulfate is that the former is volatile. For volatile aerosols, since their production reaction is reversible and the reaction rate is difficult to determine, the methodology based on the chemical dynamics lost its effect in dealing with such a problem and the accurate aerosol content is unable to be obtained. Therefore, it is necessary to calculate the concentration under the state of equilibrium from the viewpoint of chemical thermodynamics.
 In order to calculate the concentration and composition of aerosol, the chemical equilibrium state is usually assumed among the volatile species (gas or aerosol). This assumption is true in many situations. However, under some conditions, the time needed for the chemical equilibrium is longer than that needed for gas-particle touching. Under such situation, the equilibrium method is invalid and the transport process should be considered [Wexler and Seinfeld, 1991]. However, this is only limited to coarse particles and low temperature [Meng and Seinfeld, 1996]. The experiment indicated that the nonequilibrium state is existent [Allen et al., 1989]. For non-sea-salt particle and/or the warm environment, the thermodynamic equilibrium is considered to be valid [Hildemann et al., 1984; Quinn et al., 1992].
 The thermodynamic equilibrium model ISORROPIA used in this paper was developed by Nenes et al. . ISORROPIA includes 15 equilibrium reactions and the species investigated in this model are as follows: (1) gas phase: NH3, HNO3, HCl, H2O; (2) liquid phase: NH4+, Na+, H+, Cl−, NO3−, SO42−, HSO4−, OH−, H2O; and (3) solid phase: (NH4)2SO4, NH4HSO4, (NH4)3H(SO4)2, NH4NO3, NH4Cl, NaCl, NaNO3, NaHSO4, Na2SO4, and H2SO4.
 The input parameters of the model are total concentrations of Cl, Na, NH3, HNO3, and H2SO4 as well as the relative humidity and air temperature of the environment. For continental aerosols, the chlorine and sodium need not be considered.
2.2.2. Gas-Phase Chemistry Model
 A condensed gas-phase chemistry model based on CBM4 [Wang, 1996] is used in this paper in order to describe the conversion rates from NOx, SO2 to HNO3, H2SO4. According to the carbon bond state, hydrocarbon compounds are classified into four groups: PAR, OLE, ARO and CAR. The chemical mechanism contains 36 reactions (4 photolysis reactions) and 20 species. The quasi-steady-state-approximation (QSSA) scheme is chosen to solve the chemical kinetic equations. The condensed chemical mechanism and high-efficiency algorithm are suitable for coupling with the regional climate model.
2.2.3. Parameterization of Indirect Effect
 Parameterization schemes were chosen for illustrating the first and second indirect effect of nitrate. In order to investigate the first indirect effect of nitrate aerosol, a parameterization scheme is introduced into RegCM3. According to Ghan et al. :
where Nc is the number concentration of cloud droplets, Na is the total aerosol number concentration, w is vertical velocity, α depends on aerosol properties and vertical velocity. Following Chuang et al. , their expressions of αl and αo are given below. Over land
where γ is the ratio of anthropogenic sulfate mass loading (in μg m−3) to the total aerosol number concentration (in 1000 cm−3) and β is the ratio of the fraction of anthropogenic sulfate converted by the aqueous pathway to the mean value used previously (75%). Then cloud droplet effective radius re can be obtained from Nc and cloud liquid water content L.
where ρw is the water density and k is the cube of ratio of average volume radius and effective radius which is 0.67 over land and 0.8 over ocean.
 To evaluate the second indirect effect of nitrate, we refer to Beheng's  work,
where P is the autoconversion rate of rainwater, n(= 10) is the width parameter of the initial cloud droplet spectrum, γ1 (= 150) is a tunable parameter, ρa is air density, qc is cloud liquid water content, and b is the cloud cover fraction.
2.3. Numerical Experimental Design
 To understand the formation mechanism of nitrate aerosol, the gas-phase chemistry model and thermodynamic equilibrium model were linked together, and several sensitivity tests were conducted [Wang et al., 2006]. TACM and RegCM3 were coupled to build up a regional climate chemistry modeling system (RegCCMS), in which the two models are coupled in an online way at each time step.
 The emission inventory is based on Streets et al.'s  work. The annual total emissions of SO2, NOx and NH3 over China are 26.56 × 109 Kg, 14.47 × 109 Kg and 8.85 × 109 Kg, respectively. Since the focus of this study is on the climatic effect of nitrate formed from anthropogenic emissions, biogenic emissions and biomass burning emissions are not included. The horizontal resolution is set as 50 km with 120 × 89 grid points and the vertical resolution is set as 18 levels (sigma values are 1.0, 0.99, 0.98, 0.96, 0.93, 0.89, 0.84, 0.78, 0.71, 0.63, 0.55, 0.47, 0.39, 0.31, 0.23, 0.16, 0.10, 0.05 and 0.0). The pressure at the model top is 50 hPa. The model domain, covering the mainland China and its surrounding regions, is on a Lambert map projection. Holtslag's planetary boundary layer scheme, relaxation boundary condition and Grell's cumulus convection parameterization were selected in this study.
 To investigate the direct and indirect effect of nitrate on climate, five numerical experiments (Table 1) were conducted spanning a period from November 2002 to November 2003, with the first month serving as spin-up for the subsequent 1 year simulation. The control run Experiment A (EA) uses the standard model in which only background aerosol is included. In Experiment B (EB), Experiment C1 (EC1) and Experiment C2 (EC2), direct, first indirect and second indirect effect of nitrate, respectively, is taken into account. The direct, first and second indirect effect of nitrate are all included in Experiment D (ED). NCEP reanalysis data with resolution of 2.5° × 2.5° is used as initial and boundary conditions. In order to investigate the first indirect effect, the Nc in formula (7) is forced to be influenced by only background aerosol in EC1. Therefore the second indirect effect will be shielded. Similarly, the Nc in formula (6) is forced to be influenced by only background aerosol in EC2 to consider the second direct effect. Both the direct and indirect radiative forcings of nitrate aerosol are estimated by calculating the net downward solar radiative flux at the top of the atmosphere twice in one simulation, once with nitrate aerosol effect included, and once without nitrate aerosol effect. The radiative forcing of nitrate aerosol can then be determined from the difference of solar radiative flux between the two calculations. The climatic effect of nitrate aerosol could be assessed by comparing each experiment with EA.
Table 1. List of Numerical Experiments
Experiment A (EA)
Standard model without the climatic effect of nitrate
Experiment B (EB)
Standard model with the direct climatic effect of nitrate only
Experiment C1 (EC1)
Standard model with the first indirect climatic effect of nitrate only
Experiment C2 (EC2)
Standard model with the second indirect climatic effect of nitrate only
Experiment D (ED)
Standard model with direct and indirect climatic effect of nitrate
3. Results and Discussions
3.1. Performance Evaluation on RegCCMS
 In order to evaluate the performance of RegCCMS, the annual average meteorological fields from experiment EA was compared with the observations, which were shown on Figure 1. NCEP reanalysis data of pressure, wind, temperature and humidity are used for comparison. Generally speaking, the modeling results are consistent with the observations concerning both spatial distribution and magnitude. The model performance on pressure and wind is good, however the modeling system presents lower temperature in central and south China and lower humidity.
 The observational precipitation data are from CPC Merged Analysis of Precipitation (CMAP) in NOAA. For regional climate modeling, the biggest challenges lay in precipitation. In this work, difference is obvious between simulation and observation as precipitation is concerned. Model precipitation is strong compared to observation. Abnormal precipitation is found in Tibetan area, showing the influence of the topography on regional climate modeling. The observational cloud data are from International Satellite Cloud Climatology Project (ISCCP). The modeling cloud amount is calculated according to the random overlap assumption, showing underestimation in the east over ocean and overestimation in high latitudes. The comparisons above indicate the general ability of the model on regional climate modeling over East Asia. While RegCCMS performs well on pressure, temperature, humidity and wind simulation, more works are necessary to be done to improve cloud and precipitation modeling, which is a challenge for regional climate modeling.
3.2. Surface Concentration and Vertical Loadings
Figures 2a and 2b depict the annual average surface concentration and column burden of nitrate aerosol. It shows that nitrate generally concentrated over the industrial regions. Surface concentration of nitrate is high in Henan, Shandong, Hebei, Beijing and Sichuan, with a maximum of 12 μg/m3. Column burden of nitrate is large in Sichuan, Shanxi, Henan and Shandong, with a maximum of 20 mg/m2. The maximum of nitrate surface concentration and column burden have exceeded those of sulfate as reported by T. Wang et al. , in which 1995 emission inventory was used. Comparatively speaking, there exists difference between the spatial patterns of surface concentration and column burden, since the former is mainly affected by source emissions while the later is much more affected by atmospheric transport and diffusion processes. At 500 hPa, nitrate can be transport at long distance under the westerly. Although the concentration is very low at 200 hPa, nitrate maximum also can be found over Tibet region due to vertical convection.
 The observed nitrate ion concentrations in precipitation in some cities were compared with the simulations, showing in Figure 3. The observational data is obtained from the Acid Deposition Monitoring Network in East Asia (EANET: http://www.eanet.cc/). These stations are located in Chongqing, Xian, Xiamen of China, Oki of Japan and Kanghwa, Cheju of Korea. Table 2 summarizes the observed nitrate concentration in some cities of China and the simulated results in this study. In general, the simulations are consistent with the observations. However, in some urban areas the simulations are less compared to the observations. This deviation may be attributed to several aspects. The most important factor is the uncertainty of emission inventory. NOx emission in some cities may be underestimated. Another reason is that only the anthropogenic sources of NOx are considered, while the natural sources (e.g., soil NO emission) are not included, leading to the lower concentrations of nitrate precursors. In addition, the observational data are more affected by local sources, while the simulations represent the average in the model grid.
Table 2. Comparison of Surface Concentration of Nitrate From Simulation and Observation
Figure 4 shows the spatial distribution of the annual average of direct radiative forcing due to nitrate at the top of atmosphere (TOA), which has similar pattern as the column burden of nitrate. The negative forcing is strong in Sichuan, Shandong, Henan and Hebei, with extreme value of −3 W/m2. The countrywide annual average of direct radiative forcing of nitrate at TOA is −0.88 W/m2. Since there is no report on direct radiative forcing of nitrate from regional studies in China, comparisons were performed with other studies at global scale. The IPCC  reported that the global mean radiative forcing of nitrate is −0.1 ± 0.1 W/m2. Adams et al.  presented a global mean radiative forcing as strong as −0.22 W/m2. van Dorland et al.  and Jacobson  suggested relatively minor global mean radiative forcing of −0.03 and −0.05 W/m2, respectively. Liao and Seinfeld  estimated a global mean radiative forcing of −0.16 W/m2. It is evident that direct radiative forcing due to nitrate aerosol in China is stronger compared to the global average. Wu et al.  simulated the radiative forcing due to black carbon aerosol with a maximum of 4 W/m2 in Sichuan. T. Wang et al.  estimated that direct radiative forcing of sulfate was about −0.92 W/m2, and the forcing can be as strong as −7 W/m2 in central and east China during winter. Giorgi et al.  estimated that anthropogenic sulfate induces a negative radiative forcing at TOA which varies spatially from −1 to −8 W/m2 in winter to −1 to −15 W/m2 in summer. Qian et al.  simulated the direct radiative forcing due to various aerosols in China, which is in the range −1 to −14 W/m2 in autumn and summer and −1 to −9 W/m2 in spring and winter, showing substantial spatial variability at regional scale. Obviously, direct radiative forcing due to nitrate in China is weaker compared to sulfate and black carbon aerosols, but at the same magnitude.
Figures 5a and 5b depict the annual average changes of surface air temperature and precipitation due to the direct effect of nitrate, respectively. It can be seen from Figure 5a that surface temperature decreases in most areas, especially in eastern, central and southwestern China. The decrease is significant in Neimenggu, Yunnan and Hubei, with extreme value of −0.3°C. Notable increases in surface air temperature occur in Liaoning, Fujian and Xinjiang, reaching 0.2°C in some area. The change of surface temperature is obviously induced by the direct radiative forcing. However, it is also affected by the complex feedbacks in the climate system. The pressure and circumfluence fields also change because of the direct radiative forcing, leading to the decrease of water vapor and cloudiness as well as increase of solar radiation in some regions, which result in the surface temperature rise. It is shown in Figure 5b that the pattern of precipitation change is much more complex compared to surface temperature because the former is more affected by the transport and vertical convection of water vapor. The largest decrease appears in Hubei and Heinan, with extreme value of −1.5 mm/d and the most significant increase is found in Yunnan and Guizhou, with peak of 1 mm/d.
3.4. First Indirect Radiative Forcing and Climatic Effect
Figure 6 shows the spatial distribution of the annual average of the first indirect radiative forcing due to nitrate at TOA. As depicted in Figure 6, the negative forcing is significant in southwest China, especially in Chongqing, Sichuan and Guizhou provinces where −7 W/m2 was estimated. Unlike the direct radiative forcing which is stronger in Shandong and Henan, first indirect radiative forcing is relatively weak in those polluted regions, and remains large in Hunan and Fujian apart from southwest China, which is because that the distribution of first indirect radiative forcing is not only controlled by nitrate loading, but also affected by cloud cover, surface albedo and other factors. The countrywide annual average of the first indirect radiative forcing is −2.47 W/m2. In the report of IPCC , the first indirect radiative forcing from anthropogenic aerosol is −0.7[−1.1, +0.4] W/m2. The first indirect radiative forcing of sulfate is −1.35 W/m2 as estimated by Rotstayn and Penner . Thus, the first indirect radiative forcing of nitrate in China is at the same magnitude as sulfate and stronger compared to the global average.
Figures 7a and 7b show the changes of surface air temperature and precipitation due to the first indirect effect of nitrate. Figure 7a shows that in most areas of China, surface temperature decreases due to the negative radiative forcing resulting from the first indirect effect, especially in Yunnan, Guizhou, Hebei and Neimenggu, the maximum cooling can be −0.3°C. Notable increases in surface temperature occur in Neimenggu, Gansu and Xinjiang, with the maximum of 0.2°C. Surface temperature change is affected by the first indirect radiative forcing and climatic feedback mechanisms. The spatial distribution of precipitation change (Figure 7b) shows an irregular pattern. Significant decrease of precipitation appears in Chongqing, and this reduction can be as much as −2 mm/d. In addition, strong increase occurs in Henan, Hebei and Shandong, with maximum of 1 mm/d.
3.5. Second Indirect Climatic Effect
 The change of autoconversion rate of rainwater (Pauto) due to the second indirect effect of nitrate is depicted in Figure 8a. In the modeling domain, Pauto shows strong decreases in north and central China compared to south China, with extreme value of −3 × 10−8 Kg/Kg/s in Jilin and Tibet. The change of Pauto is mainly influenced by both nitrate loading and cloud amount. In area with much nitrate and cloud, reduction of Pauto is more significant.
Figure 8b shows the annual average impacts on cloud amount of the second indirect effect of nitrate. In most areas of China, cloud amount increases because of the second indirect effect, large increase occurring in north and northeast China, especially in Jilin and Neimenggu, with a maximum of 4%. However, notable decreases appear southeast and northwest China, with the maximum of 0.5% occurring in Xinjiang. The variation of cloud amount is well corresponded to that of Pauto since increase of cloud amount results from the increase of cloud water which is caused by the reduction of Pauto.
 The impacts of the second indirect effect on surface air temperature and nonconvective precipitation are shown in Figures 8c and 8d, respectively. In most parts of China, surface temperature decreases due to the second indirect effect, strong cooling occurs in north and southwest China, for example −1.2°C in Neimenggu. Relative weak increase appears in Xinjiang, and the maximum warming is only about 0.1°C. The variation of surface temperature is resulted from that of solar radiation because of the cloud amount change. The largest increase in nonconvective precipitation occurs in southwest China, with a maximum of 1.2 mm/d, while significant decrease is found in central and southeast China, with extreme value of −1.2 mm/d. The existence of nitrate leads to the decrease of effective radius of cloud droplets, inhibiting the transformation from cloud water to precipitation. Moreover, the decrease of surface temperature would reduce evaporation and weaken the upward flow, which would also reduce precipitation.
3.6. Combined Radiative Forcing and Climatic Effect
Figure 9 shows the annual average combined radiative forcing at TOA due to nitrate. In this experiment, both direct and indirect effects are considered. The estimated radiative forcing for the combined effect is −2.52 W/m2. The negative radiative forcing is stronger in south than north, especially in Guizhou, Chongqing and Sichuan, with peak of −7 W/m2. The combined radiative forcing is not the simple sum of direct and first indirect radiative forcing, indicating strong nonlinearity in climate system. Since the first indirect radiative forcing is much stronger than the direct radiative forcing, the spatial distribution of the combined radiative forcing has a similar pattern to that of the first indirect radiative forcing.
Figures 10a and 10b illustrates the variation of surface air temperature and precipitation resulting from the combined effect of nitrate at annual scale. Surface temperature decreases in most regions of China, especially in Neimenggu, with strong cooling of −1.5°C. The distribution of surface temperature change due to the combined effect is similar to the second indirect climatic effect, suggesting that the second indirect effect is a predominant factor in controlling surface temperature change. Decrease of precipitation is found to be the strongest in Sichuan and Hubei, with peak reaching −2 mm/d.
Table 3 lists a comparison of radiative forcing and climatic effect between this work and some previous studies. The direct and first indirect radiative forcing of nitrate in China is at the same magnitude as sulfate, suggesting that nitrate also plays an important role in regional climate over China apart from sulfate, which should be included in the coupling aerosol-chemistry-climate model. The regional forcing of nitrate is also stronger compared to the global average, therefore, regional modeling is especially important in East Asia due to strong inhomogeneousness in emission, land use and land cover. There are differences in various simulation results on the climatic effects especially in precipitation, showing that large uncertainty exists in numerical modeling of climatic effects of different aerosol and muck work should be done in model and parameterization improvements.
Table 3. Comparison of Radiative and Climatic Effects Between This Work and Previous Studies
 The statistical significance of the climatic effects have been calculated to examine that our results are statistically significant or due to the model internal variability. Linear correlations between the simulated annual average column burden of nitrate and temperature or precipitation changes due to direct, first indirect, second indirect, and combined effects are statistically significant at the 95% confidence level.
3.7. Seasonal Variations of Radiative Forcing and Climatic Effect
 In this study, the seasonal concentration, column burden of nitrate and the corresponding climatic effects over China are also investigated, which are listed in Table 4. Seasonal column burden of nitrate, radiative forcings and changes of temperature and precipitation in winter (DJF) and summer (JJA) are illustrated in Figure 11. They show that indirect effect is strong in summer (JJA) and weak in winter (DJF). For direct effect, it is strong in spring (MAM) and weak in summer (JJA) in terms of surface air temperature because of increase of temperature in some areas caused by strong climatic feedback in summer. Seasonal variations of precipitation are in contrast to surface temperature. They are highly dependent on monsoon climate in east China, especially the wind, precipitation, humidity and cloud, which controls the column burden of nitrate as well as the direct and indirect radiative forcings. From the perspective of the annual average, the indirect effect is much stronger compared to the direct effect both on surface temperature and precipitation. In addition, the second indirect effect on regional climate change is stronger than the first indirect effect, the former is prevailing in the combined effect.
Table 4. Seasonal Variations of Nitrate Concentration and Its Climatic Effect Over China
 In this paper, the regional climate model (RegCM3) and a tropospheric atmosphere chemistry model (TACM) was coupled to construct a regional climate chemistry modeling system (RegCCMS), which was applied to investigate spatial distribution of nitrate aerosol, radiative forcing as well as its climatic effect over China. RegCCMS was run continuously for 13 months from November 2002 to November 2003 with the first month as a spin-up period. One year simulations show that nitrate is mainly distributed in east and north China as well as Sichuan basin, with the maximum of 12 μg/m3.
 Annual average of direct radiative forcing of nitrate at TOA is −0.88 W/m2, in some region it can be −3 W/m2. Spatial distribution of direct radiative forcing is similar to that of nitrate surface concentration. Besides, it is also influenced by cloud cover, surface albedo and other factors. Owing to the direct effect, the countrywide annual average changes of surface air temperature and precipitation are −0.04°C and −0.05 mm/d, respectively.
 Annual average of the first indirect radiative forcing of nitrate at TOA is −2.47 W/m2; it reaches −7 W/m2 in some regions. Spatial distribution of indirect radiative forcing is related to that of nitrate concentration as well as cloud cover. Because of the first indirect effect, the countrywide annual average changes of surface air temperature and precipitation are −0.11°C and −0.10 mm/d, respectively.
 The second indirect effect of nitrate would increase countrywide cloud amount, thus reducing solar radiation, decreasing surface air temperature and precipitation. Annual average changes of surface air temperature and precipitation due to the second indirect effect are −0.68°C and −0.42 mm/d, nonconvective precipitation reducing by −0.29 mm/d. The second indirect effect is much stronger than both direct effect and the first indirect effect.
 The annual average of total radiative forcing of nitrate at TOA is −2.52 W/m2. Owing to the combined effect, the countrywide annual average changes of surface air temperature and precipitation are −0.78°C and −0.52 mm/d, respectively, and the reduction of nonconvective precipitation is −0.33 mm/d. The changes of surface temperature and precipitation due to the combined effects are less than the sum of each individual effect, showing a strong nonlinearity between climate change and radiative forcing. It should be noted that the simulation in this work is for 1 year only, and not for a climatology, so the errors may be relatively higher than in other studies looking at climatologies for multiple years. The preliminary results in this paper suggest that as an important anthropogenic aerosol, the impact of nitrate on regional climate change in China cannot be neglected compared to sulfate and carbonaceous aerosols. More emphasis should be concentrated on the improvements on parameterization of nitrate-induced cloud condensation nuclei and autoconversion rate of rainwater, thus to reduce the uncertainty in estimation of indirect effect of nitrate. In addition, the performance of regional climate model need to be improved to reduce uncertainties associated with simulated cloud and precipitation which could influence radiative forcing and climate response significantly.
 This work was supported by the Ministry of Science and Technology (2010CB428503 and 2006CB403706), the National Key Technology R & D Program (grant 2007BAC03A01), and the National Special Fund for the Weather Industry (GYHY200806001–1 and GYHY2007-6-36). Special thanks are given to A. Nenes for providing his ISORROPIA model.