Atmospheric Environment Vol. 32, No. 8, pp. 1383—1395, 1998 ( 1998 Elsevier Science Ltd. All rights reserved Printed in Great Britain S1352–2310(97)00134–9 1352—2310/98 $19.00#0.00
AIR QUALITY IMPACTS AS A RESULT OF CHANGES IN ENERGY USE IN CHINA’S JIANGSU PROVINCE YOUNG-SOO CHANG,*,- RICHARD L. ARNDT,‡ GIUSEPPE CALORI,° GREGORY R. CARMICHAEL,‡ DAVID G. STREETS and HAIPING SU - Argonne National Laboratory, 9700 S. Cass Ave., Argonne, IL 60439, U.S.A.; ‡ Department of Chemical and Biochemical Engineering, Center for Global and Regional Environmental Research, University of Iowa, Iowa City, IA 52242, U.S.A.; and ° Polytechnic di Milano, Milano, Italy (First received 24 October 1996 and in final form 26 February 1997. Published April 1998) Abstract—Potential changes in air quality associated with increases in SO emissions resulting from 2 industrial growth for Jiangsu Province and the Shanghai Municipality in China are evaluated for the years 1990 (current) and 2010 (future) with a ‘‘no further control’’ scenario. Two long-range transport models are used to estimate airborne concentrations and deposition of SO and sulfate, based on available data on 2 emissions originating from area and major point sources. Modeling results demonstrate how human health and ecological impacts are related to the projected scenario of industrial growth, in terms of the critical level (or concentration) and critical loads. The current approach provides an effective and valuable tool for preliminary assessment of potential impacts of these changes on human health and critical ecosystems in the region for use in integrated energy and environmental risk assessments. ( 1998 Elsevier Science Ltd. All rights reserved. Key word index : Air quality impacts, Jiangsu Province, Shanghai, UR-BAT, human health and ecological impacts.
Along with many southeast Asian countries, China is experiencing very high economic growth. This economic growth improves the standard of living in China. However, the unprecedented economic growth results in various significant problems, such as resource depletion, deterioration of air and water quality, and intense land-use conflicts. In particular, the projection is that the growing energy demands associated with industrial and economic growth will require substantial increases in the use of indigenous coal as a primary fuel source; such use is a major culprit in air and water pollution. High levels of agricultural and industrial production and rapid population growth are also placing heavy burdens on the natural carrying capacity in the region. Streets et al. (1995) showed that rapid urbanization, industrialization, and demographic shifts have considerable ramifications for agriculture and land-use patterns in this part of China. By using remote sensing data for southern Jiangsu Province during the period 1976—1984, their analysis revealed patterns of anthropogenic disturbance that can provide early warning
* This work was not funded through Argonne National Laboratory.
signals of problems in resource decline and land-use management. Unless coherent land-use, population, and environmental control planning and practices are in place, rapid growth in the region may result in serious impacts on overall environmental quality in the foreseeable future. One region where these issues are particularly apparent is the Yangtze River delta. Jiangsu Province, hailed as a paragon of economic reform in the People’s Republic of China, is a nationally important agricultural and industrial center. After the end of the Cultural Revolution, this region experienced unprecedented industrial and economic growth. The juxtaposition of heavy industry, agricultural production, and human habitation poses a severe threat to human health. Jiangsu Province has recorded a disproportionate incidence of cancer. One-tenth of all cancer deaths in China occur in Jiangsu Province, although it accounts for less than 6% of China’s population (Chicago Tribune, 1995). Moreover, the province, which has long had China’s highest death rate from cancer, experienced a 30% increase over the past 30 yr. A high population density, a heavy concentration of rural factories, and associated serious air and water pollution contribute to a worsening environment in Jiangsu Province. Shanghai, a major industrial and commercial center with one of the largest ports in the world, is envisioned to dominate business activity in China in
Y.-S. CHANG et al.
the next century. Although Shanghai has enjoyed the fruits of industrial growth since the Cultural Revolution along with Jiangsu Province, it has been plagued by severe ambient air pollution problems. For example, concentrations of SO and suspended particulate 2 matter frequently exceeded the World Health Organization (WHO) guidelines at most monitoring sites during the period 1981—1989 (Earthwatch, 1992). In addition, the lung cancer incidence is the highest within urban districts in China. Because of the absence of industrial zoning, many small industrial facilities are located in residential areas, with an accompanying direct personal exposure to industrial air pollutants. In this paper, we present a preliminary assessment of the potential air quality impacts due to increases in SO emissions resulting from industrial growth in 2 terms of human health and ecological impacts in the RAINS-Asia framework. Air quality modeling was performed for the year 1990 as the base year and for
2010 as the future year with a ‘‘no further control (NFC)’’ (or ‘‘status quo’’) scenario and was applied to China’s Jiangsu Province and Shanghai.
In this section, the modeling domain and source emission distribution are described. Two air quality models are presented, along with the input parameters and meteorological data used in the study. 2.1. Description of modeling domain The region under study is the eastern part of China, covering Jiangsu Province and Shanghai, as shown in Fig. 1. The region is bordered by the Pacific Ocean to the east and by Shandong, Anhui, and Zhejiang Provinces to the north, west, and south, respectively. Most of Jiangsu Province is flat, at an elevation of less than 50 m above mean sea level. The Yangtze River
Fig. 1. Map of the study region in eastern China, including Jiangsu Province and Shanghai.
Air quality impacts
flows through the southern part of the province, while the Grand Canal extends across the province. Table 1 presents general information for Jiangsu Province and Shanghai Municipality for calendar year 1993 (Sinton, 1996). Jiangsu Province has the highest population density among the provinces, and
Shanghai is one of the world’s most densely populated urban areas. Population is also heavily concentrated along the Yangtze River and the Grand Canal. Jiangsu Province and Shanghai are also an important industrial and agricultural center in China. As shown in Fig. 2, Jiangsu Province produced approximately
Table 1. General information for Jiangsu Province and Shanghai Municipality for the Year 1993 Jiangsu Province Population, millions Area, 1000 km2 Population density, persons km~2 GDP, billion 1990 yuan Per capita GDP, 1990 yuan
69.67 102.6 679 215.94 3100
(5.9%)! (1.1%)! (3)" (8.8%)! (7)"
Shanghai Municipality 13.49 6.2 2176 118.5 8785
(1.1%)! (0.1%)! (1)" (4.8%)! (1)"
China 1185.17 9586.0 124 2460.06 2076
! Percentage of the national total. " National ranking out of 30 provinces in total (including 5 autonomous regions and 3 municipalities).
Fig. 2. Annual production trends of rice, wheat, and corn in Jiangsu Province.
Y.-S. CHANG et al.
10% of China’s rice and wheat, the staple foods in China, although the province accounts for only 1% of the land and less than 6% of the population in China. In addition, Shanghai and Jiangsu Province ranked first and seventh, respectively, in terms of per capita gross domestic product (GDP), an important economic indicator. The modeling grid system under study includes most of Jiangsu Province, Shanghai, and parts of neighboring provinces (118—122°E on 30—32°N, 118—121°E on 32—34°N, and 117—121°E on 34—35°N). Calculations that take into account both the effects due to emissions from within this study domain and those due to emissions from source areas outside this targeted study area are performed. The modeling grid system for these background contributions to the domain under study is the same as that used in other applications of the RAINS-Asia model (Arndt and Carmichael, 1995). 2.2. Distribution of SO emissions 2 To simulate the background contributions to the study domain, area source emissions provided at 1° resolution from the RAINS-Asia module were used. The emission location for surface area sources is the center of the grid square. The large point sources (LPSs) are electric power plants and industrial
sources which meet certain emissions criteria. Because sources of elevated emissions are individual facilities (e.g. electrical power plants), their emissions are designated at their actual locations (at 0.01° resolution). The LPS data files are subdivided into existing and planned sources. Planned LPSs are those facilities that are not operational in 1990 but are expected to become operational by the year 2010. The modeling grid system for background contributions to the domain under study is the same as that for use in the original RAINS-Asia model. The emission strengths were assigned by 1°]1° grids, covering the Asian domain bounded by 10°S to 55°N and 60°E to 150°E. Within the domain under study, area emission figures from the RAINS-Asia emission module were used to construct the modeling emission fields with a 0.1° resolution. The distribution of area emissions was based on the 1994 population distribution data with a 0.08° resolution (Tobler et al., 1995), as shown in Fig. 3. The distribution of area emissions with a 0.1° resolution for the year 1990 are shown in Fig. 4a. The LPSs for the years 1990 and 2010 are presented in Fig. 4b, in which the relative size of the symbols indicates the relative magnitude of the emission sources. Total SO emissions for the study domain 2
Fig. 3. Spatial distribution of population for the year 1994.
Air quality impacts
Fig. 4. Spatial distributions of SO : (a) area source emissions in 1990, and (b) point source emissions in 2 1990 and 2010 (Relative size of symbols indicates the relative magnitude of LPS.).
Y.-S. CHANG et al.
in 1990 were estimated to be 3.2]106 t, with 13% and 87% from point and area sources, respectively. Emissions for a future scenario are also studied. For this purpose, the energy scenario generation module of RAINS-Asia has been applied to create emission projections for 2010. The hypothesis of an NFC scenario has been adopted, allowing evaluation of future emissions based on the projections of socioeconomic activities and under the hypothesis of no emission control measures adopted beyond the ones already implemented in 1995. Under this scenario, emissions are projected to increase by a factor of two over 1990 levels by the year 2010. This increase is due to projected growth in power plants and industrial production activities. In 2010, the proportion of point source emissions increases to about 24% of the total SO emissions of 6.7]106 t. 2 2.3. ¹rajectory models To calculate airborne concentrations and dry and wet deposition of SO and sulfate, two Lagrangian 2 long-range transport/deposition models were employed. The first model, referred to as the ATMOS model (Arndt and Carmichael, 1995), is designed to simulate long-range transport and deposition of sulfur in Asia, originating from area sources and LPSs. This model was used to estimate contributions of outside sources to the region under study. The second model, referred to as UR-BAT (URban-Branching Atmospheric Trajectory) (Calori and Carmichael, 1996), was used to estimate the contributions of sources in the region under study. The UR-BAT model is a derivation of the ATMOS model, designed to simulate the transport and deposition of sulfur around megacities. Basic concepts of the ATMOS and UR-BAT models are briefly summarized as follows. The ATMOS model is a three-dimensional Lagrangian model with three vertical layers (surface, boundary, and upper layers). During the daytime, two layers exist: the boundary layer, extending from the surface to the bottom of the temperature inversion; and the upper layer, extending from there to the fixed height of 6000 m. At night, one more layer (the surface layer) exists, extending from the ground to the fixed height of 300 m, although this layer is absorbed into the boundary layer due to daytime turbulence. The inversion height can be determined from the vertical temperature profiles. Emissions from a single location (grid cell) are represented by a series of puffs starting every three hours. At its source, each puff is placed in the appropriate vertical layer, depending on the source type (surface or elevated) and release time of day (day or night). Each puff is assigned a mass proportional to the source strength and is assumed to mix uniformly in the vertical throughout its assigned layer and to diffuse with a Gaussian distribution in the horizontal. Each puff is transported by the average horizontal winds for each layer, which were computed from upper sounding data.
Puffs can be pulled apart by vertical wind shears. As a result, the puff branches into independent trajectories, and this branching takes place at the day-tonight or night-to-day transitions, up to a number of 32 for each puff initially emitted. After branching, the mass is divided proportional to the layer thickness. Each puff is tracked for five days or before disappearing from the modeling domain. As the puffs are transported along their trajectories, SO is chemically converted to sulfate, and SO and 2 2 sulfate are removed from the atmosphere through dry and wet deposition mechanisms. The ATMOS model incorporates the mass balance equations by considering the first-order chemical conversion of SO to sulfate and the first-order processes of dry and 2 wet deposition. Note that emissions from the region under study were turned off when estimating the background contributions. The UR-BAT model is based on the ATMOS model, so only its differences will be described here. In order to characterize the urban environment (e.g. heat island), the UR-BAT model further subdivides the lower part of the atmosphere over urban grid cells with respect to the regional-scale model. Over a major urban area and at daytime, the boundary layer is split into two vertical layers: the urban boundary layer (UBL) and the boundary layer over the urban area (BOU). The UBL extends from the ground to 300 m, whereas the BOU extends from 300 m to the bottom of the temperature inversion. Every hour, a new puff is released from each source. In the ATMOS model, the entire domain is subdivided into 1°]1° ‘‘land’’ and ‘‘sea’’ grid cells. In the UR-BAT model, land is further characterized to realize concentration peaks near the megacities. The location and extent of the urban area is specified in terms of ‘‘urban’’ and ‘‘rural’’ grid cells of an appropriate urban terrain mask, generally specified at a higher resolution (say 0.1°). On the basis of a procedure recommended by the U.S. Environmental Protection Agency (U.S. EPA, 1995), the area is classified as urban if the population density is greater than 750 people km~2 (approximately 75,000 people per 0.1°]0.1° cell). Different from U.S. cities with an extensive sprawl, Asian cities are concentrated in a relatively small area and have high population density. In this analysis, cells with a population of more than 150,000 were classified as urban. 2.4. Description of meteorological data Upper-air sounding data provided by the U.S. National Climatic Data Center for 1990 have been used to determine the height of the inversion layer and the average winds of each layer in the trajectory models. The data include wind and temperature from rawinsonde and pibal observations for Eastern Asia from the surface to the level of 500 mb, with four observations per day (every six hours). Precipitation data used were fields analyzed by the National Meteorological Center. These data were
Air quality impacts
obtained from the U.S. National Center for Atmospheric Research. As with the upper-air sounding data, precipitation data consisted of accumulation values collected in six-hour intervals throughout the region in 1990. These data were provided on a 1.4695°]1.4875° grid spacing resolution.
1389 3. RESULTS AND DISCUSSION
Modeling runs were made for two months—August (representing summer) and December (representing winter)—to provide estimates of seasonal differences in potential impacts. For average concentrations, the
Fig. 5. Estimated annual average SO concentrations for the years 1990 (a) and 2010 (b). 2
Y.-S. CHANG et al.
Fig. 6. Percentages of total hours exceeding the WHO 1 h average SO concentration of 350 kg m~3 for 2 the years 1990 (a) and 2010 (b).
predicted values for two months are averaged; and for deposition rates, predicted values for two months are added and multiplied by six to arrive at annual total cumulative values. Potential air quality impact ana-
lyses were made in terms of human health and ecological impacts. Figure 5 shows annual average SO concentrations 2 for the years 1990 (Fig. 5a) and 2010 (Fig. 5b) with the
Air quality impacts
Fig. 7. Estimated annual average SO~2 concentrations for the years 1990 (a) and 2010 (b). 4
NFC scenario. In 1990, annual average SO concen2 trations were predicted to exceed the long-term WHO guideline of 40—60 kg m~3 throughout the lower Yangtze River delta, the Nanjing metropolitan area, and the northern industrial regions of Jiangsu. In
2010, SO concentration levels in most areas of 2 Jiangsu Province and Shanghai are projected to exceed the long-term WHO guideline. Xu et al. (1994) conducted a study that compared daily counts of deaths in two residential areas of Beijing with daily
Y.-S. CHANG et al.
Fig. 8. Critical loads map for sulfur deposition for the year 1990 (200 eq ha~1 yr~1+0.32 g S m~2 yr~1).
concentrations of SO . They found a highly signifi2 cant correlation between mortality and SO concen2 trations. A doubling of ambient SO concentration 2 was associated with an 11% increase in all-cause mortality. Concentration levels in 2010 are predicted to be approximately two times higher than those in 1990, which suggests that most residents in the region face serious health problems unless control measures are in place. Compared with the monitoring data, modeling results are predicted to be within the range and to capture the general trends. Monitored annual average SO concentrations in Shanghai, Nanjing, Suzhou, 2 and Xuzhou were 95—99 kg m~3, 50—74 kg m~3, 65—77 kg m~3, and 71—81 kg m~3, respectively, during the period of 1989—1991 (Sinton, 1996). Concentration levels for the year 1990 in Shanghai, Nanjing, Suzhou, and Xuzhou were predicted to be 60—80 kg m~3, 80—100 kg m~3, 60—80 kg m~3, and 60—80 kg m~3, respectively. Predicted values are comparable to monitored ones, which suggests that the current modeling produces reasonable results. Figure 6 shows the percentage of total hours exceeding the WHO short-term (1 h average) guideline of 350 kg m~3, an indicator of acute health impacts. In 1990, the WHO short-term guideline is exceeded in only a few isolated areas close to the largest point sources. By contrast, in 2010, the guideline is exceeded in large areas of northern, central western, and southern Jiangsu Province for more than 5% of the time.
A study of the association between air pollution and mortality in six U.S. cities found that mortality increases by about 30% for each 10 kg m~3 increase in annual average sulfate concentrations (Dockery et al., 1993). The calculated sulfate concentrations are presented in Fig. 7. In most of Jiangsu Province and Shanghai, sulfate concentration levels in 2010 are predicted to increase by more than 10 kg m~3 compared with 1990 levels. Although the dangers in inter-country extrapolation are recognized, these results do suggest a significant health risk associated with increasing sulfate levels in the region. The large quantities of sulfur predicted in the region can also have adverse impacts on the ecosystems through direct and indirect effects. Of particular concern is the potential impact on agriculture in the region. The model results were also used to assess these potential impacts. Elevated SO concentra2 tions can have a direct impact on vegetation. One simplistic, but useful, index for assessing the direct effects is the concept of critical levels. Many crops and trees show declines in yield and productivity when exposed to SO concentrations in excess of some 2 threshold value (or critical level). A critical level of 40 kg m~3 has been employed in the assessment of direct impacts of SO on plants (Hettelingh et al., 2 1996). In referring back to Fig. 5, crops grown in the fertile regions along the lower Yangtze River delta and in the northern regions are exposed to SO con2 centrations in excess of the critical level at present
Air quality impacts
levels of emissions. By the year 2010, under the NFC scenario, essentially all of the agricultural regions in the study domain are projected to be exposed to concentrations in significant excess of this critical level.
Indirect effects of air pollution on ecosystems are also of concern, especially through acidification. A widely used method for assessing the effects of acid deposition on ecosystems in Europe, and recently in Asia, is based on the concept of critical loads
Fig. 9. Estimated exceedance critical loads (CLs) for the years 1990 (a) and 2010 (b) (200 eq ha ~1 yr~1+ 0.32 g S m~2 yr~1).
Y.-S. CHANG et al.
(Hettelingh et al., 1996). The critical load is defined as the highest deposition of a compound that will not cause harmful effects on an ecosystem. The higher the critical load, the less sensitive is the ecosystem to acid deposition. The critical load map for sulfur deposition for ecological impact analysis extracted from the RAINS-Asia database is shown in Fig. 8 (Hettelingh et al., 1996). The upper and lower portions of the province are relatively more sensitive to acid deposition than the middle portion. Sulfur deposition was also calculated by the models. The total sulfur deposition can be compared with the critical loads to determine which regions are receiving sulfur deposition in excess of the critical loads (the amount of sulfur in excess of the critical load is defined as the exceedance) and therefore which regions are at risk of damage. Predicted exceedances of the critical loads for sulfur deposition are shown in Fig. 9. The white denotes no exceedance (and no risk), and the dark shading denotes large exceedances (and high risk). In 1990, areas at risk are sparsely located, primarily around the largest sources. In 2010, the area at risk is projected to cover most of the lower Yangtze River delta and the northern regions. However, the central regions of eastern Jiangsu Province, due to their high critical loads, remain risk free of indirect effects due to acid deposition.
Potential air quality impacts on human health and agriculture in Jiangsu Province and Shanghai were assessed for 1990 and for 2010 under an NFC scenario. The SO emissions in this important agricul2 tural and industrial region are already high and are projected to double by the year 2010. By using two air quality models, ambient concentrations of SO and 2 sulfate and sulfur deposition were calculated for 1990 and 2010. The results were interpreted in terms of human health and ecological risk assessments. These estimates were found to be generally consistent with the limited observational data available, especially for the long-term average. From the standpoint of human health risk, these model calculations confirm that large regions are already being exposed to SO concentrations in 2 excess of the WHO long-term exposure guideline, with isolated regions having concentrations in excess of the short-term exposure guidelines. The situation in 2010 is markedly different, with essentially the entire domain exceeding the long-term guideline and significant regions exposed to concentrations well in excess of the short-term guideline. Adverse risks related to increased levels of ambient sulfate aerosol were also identified as another growing health concern in the region. These results also suggest that the continued growth of sulfur emissions may have profound
impacts on the agricultural productivity of the region. The lower Yangtze River delta and the northern regions are projected to be at risk to both direct and indirect effects of air pollution and acid deposition. The central regions of Jiangsu Province, while projected not to be at risk related to acid deposition, are identified to be at risk due to high levels of ambient SO . 2 These preliminary results indicate the nature of the potential impacts and the challenges that this region faces over the next few decades. The situation will be exacerbated by the attendant increase in emissions of nitrogen oxides (NO ), which will pose additional x environmental concerns through increases in ambient ozone levels and acid deposition (via nitric acid). Many factors will determine the actual path that this dynamic region will follow in the future, and no model can hope to forecast the reality of this region’s development. However, the current approach provides a useful tool for preliminary analysis of the potential air quality impacts of this growth and provides useful information for emission control planning and strategies. The clear message for planners in China is that emissions should be limited. Although improved energy efficiency and advanced energy technology can contribute to this objective, it is likely that flue-gas desulfurization (FGD) will be the main strategy. This recommendation is in line with the new version of the Atmospheric Pollution Control Law passed by the National People’s Congress in August 1995, which calls for the creation of ‘‘sulfur dioxide control regions’’ where emission reduction measures would be required of power plants and other large coal users through a permitting system and emissions fees. By the year 2000, a total of 10 GW of coal-fired generating capacity is planned to be fitted with FGD systems by the Ministry of Electric Power.
Arndt, R. and Carmichael, G. (1995) Long range transport and deposition of sulfur in Asia. ¼ater, Air and Soil Pollution 85, 2283—2288. Calori, G. and Carmichael, G. R. (1996) A trajectory model for sulfur in Asian megacities: Model concepts and preliminary application Atmospheric Environment, submitted. Chicago ¹ribune (1995) ‘‘China — Province Lost 70,000 to Cancer in 1994,’’ 24 September. Dockery, D. W., Pope III, C. A., Xu, X., Spengler, J. D., Ware, J. H., Fay, M. E., Ferris, Jr. B. G. and Speizer, F. E. (1993) An association between air pollution and mortality in six U.S. cities. ¹he New England Journal of Medicine 329, 1753—1759. Earthwatch, Global Environmental Monitoring System (1992) ºrban Air Pollution in Megacities of the ¼orld. World Health Organization and the United Nations Environment Programme. Hettelingh, J. P., Sverdrup, H. and Zhao, D. (1996) Deriving critical loads for Asia. ¼ater, Air and Soil Pollution 85, 2565—2570.
Air quality impacts Sinton, J. E. (ed.) (1996) China Energy Databook. Report LBL-32822-Rev. 3, Lawrence Berkeley Laboratory, Berkeley, CA. Streets, D., Chung, C., Krummel, J. and Su, H. (1995) Remote sensing of global change: growth in China’s Jiangsu Province. International Journal of Sustainable Development and ¼orld Ecology. 2, 257—266. Tobler, W., Deichmann, U., Gottsegen, J. and Maloy, K. (1995) ¹he Global Demography Project. Technical Report TR-95-6, National Center for Geographic Information
and Analysis, Department of Geography, University of California at Santa Barbara. U.S. Environmental Protection Agency (1995) Supplement C to the Guideline on Air Quality Models (Revised), EPA-450/2-78-027R-C. Xu, X. P., Gao, J., Dockery, D. W. and Chen, Y. D. (1994) Air pollution and daily mortality in residential areas of Beijing, China. Archives of Environmental Health 49, 216—222.