- Email: [email protected]
Transport of airborne particulate matters originating from Mentougou, Beijing, China
Available online at www.sciencedirect.com
China Particuology 5 (2007) 408–413
Transport of airborne particulate matters originating from Mentougou, Beijing, China Duoxing Yang a,∗ , Yongwei Han b , Jixi Gao b , Jesse Th´e c a
Institute of Crust Dynamics, CEA, Beijing 100085, China Chinese Academy of Environmental Sciences, Beijing 100012, China c Lakes Environmental Software Inc., 419-3 Phillips St., Waterloo, ON N2L 3X2, Canada b
Received 13 March 2007; accepted 7 July 2007
Abstract In this study, a coupled regional air quality modeling system is applied to investigate the time spatial variations in airborne particulate matters (PM10 ), originating from Mentougou to Beijing municipal area in the period of April 1–7, 2004, and the influences of complex terrain and meteorological conditions upon boundary layer structure and PM10 concentration distributions. An intercomparison of the performance with CALPUFF against the observed data is presented and an examination of scatter plots is provided. The statistics show that the correlation coefficient and STD between the modeled and observed data are 0.86 and 0.03, respectively. Analysis of model results illustrates that the pollutants emitted from Mentougou can be transported to Beijing municipal area along certain transport pathways, and PM10 concentration distributions show heterogeneity characteristics. Contributions of the Mentougou sources to the PM10 concentrations in Beijing municipal area are up to 0.1–15 g/m3 . © 2007 Chinese Society of Particuology and Institute of Process Engineering, Chinese Academy of Sciences. Published by Elsevier B.V. All rights reserved. Keywords: Airborne particulate matters; CALPUFF; PM10 ; SMOKE; Emission inventory
1. Introduction Atmospheric aerosols consist of a mixture of solid and liquid particles suspended in ambient air (Zhang, Wang, Sheng, Kanai, & Ohta, 2004; Zhang, Wang, & Xia, 2002), ranging in size from the smallest superfine nano-meters (nm) particles, to coarse particles, with diameters of several micrometers (m) or more (Zhang et al., 2003). Their importance arises from their ability to interact with visible radiation and to affect cloud properties, as well as their long atmospheric residence times (IPCC, 1995). ‘Particulate matter’, also known as particle pollution or PM, is a complex mixture of extremely small particles and liquid droplets. Particle pollution is made up of a set of components, including acids (such as nitrates and sulfates), organic chemicals, metals, and soil or dust particles (Sokolik & Toon, 1996). The size of particles is directly linked to their potential for causing health problems. PM particles 10 m in size or smaller, referred to as PM10 , are typically formed by “crustal” or earth-based
∗
Corresponding author. Tel.: +86 10 62842631; fax: +86 10 62842631. E-mail address: [email protected] (D. Yang).
material and enter the air through a variety of actions including “entrainment” into the atmosphere as wind blown dust. The even smaller or “fine” PM material, referred to as PM2.5 (2.5 m or smaller), are understood to be more a product of combustion (Streets et al., 2003). PM2.5 is believed to penetrate deeper into the lungs and remain lodged there rather than exhaled, causing adverse impacts on health. In the last decade a variety of transport and deposition models have been applied to address PM transport and deposition over Beijing (Zhou, Levy, Hammitt, & Evans, 2003). This study is another attempt to investigate the behavior of PM10 over Beijing, and to examine the relative role of meteorological fields and removal mechanisms in regulating PM10 behavior with CALPUFF (Bennett et al., 2002; Scire, Strimaitis, & Yamartino, 1999) model on the basis of a newly estimated emission inventory for Mentougou with a resolution of 1 km × 1 km. In this study both temporal and spatial variations in airborne particulate matters (PM10 ) as well as transport processes of PM10 originating from Mentougou to Beijing municipal area in the period of April 1–7, 2004, and the influences of complex terrain and meteorological conditions upon PM10 concentration distributions are discussed.
1672-2515/$ – see inside back cover © 2007 Chinese Society of Particuology and Institute of Process Engineering, Chinese Academy of Sciences. Published by Elsevier B.V. All rights reserved.
doi:10.1016/j.cpart.2007.07.003
D. Yang et al. / China Particuology 5 (2007) 408–413
409
2. Model configuration and setup The initial phase of CALPUFF which we used for our primary model involves deriving meteorological files using CALMET (Scire, Robe, Fernau, & Yamartino, 2000; Yang, Chen, Liu, & Zhao, 2006), which is a simple diagnostic windfield model. Much of the structure in the wind fields is determined by CALMET using its diagnostic wind field module (Douglas & Kessler, 1988). As shown in Fig. 1, the basic coordinate grid for CALPUFF/CALMET consists of 100 grid cells along the x-axis (east–west) and 100 grid cells along the y-axis (north–south), spaced 1 km apart. The coordinate system was converted to a Lambert Conical Projection grid. The 10 vertical layers incorporated into the CALMET processing had heights of 20, 50, 100, 150, 300, 600, 1000, 1500, 2200 and 3000 m. The MM5 mesoscale model was used in this study to develop high-resolution, three-dimensional meteorological fields (i.e., wind, temperature, pressure, etc.) through FDDA simulations (Stauffer & Seaman, 1994). MM5 generated the meteorological fields, which we used to improve the wind distribution results from CALMET. This is an important step since the MM5 model can resolve wind features caused by topography, such as terrain channeling and gravity-driven slope flows (Levy, Spengler, Hlinka, Sullivan, & Moon, 2002; Zhou et al., 2003). The first-guess atmosphere data for MM5 are extracted from the NCAR/NCEP FNL archives. The NCEP Final Analysis (FNL) data archived at NCAR is for every 6 h at a spatial resolution of 1◦ × 1◦ at standard pressure levels under 100 hPa. This data set includes two-dimensional variables, sea surface temperature, sea level pressure, three-dimensional variables of temperature, geo-potential height, U and V velocity components, and relative humidity (Bromwich et al., 2001). We combined the MM5 prognostic model (Douglas & Kessler, 1988) outputs with mesoscale data assimilation systems for 7 days (April 1, 2004, 00:00:00–April 7, 2004, 23:00:00). The present MM5 model domain has 32 vertical levels, going up to about 13 km AGL, with vertical grid spacing stretched from about 20 m near the ground to 800 m near the top of the domain. This allowed CALMET to interpolate from a higher-to-a-lower-resolution grid (since CALMET uses 10 vertical layers). One-way nesting was used to generate ambient wind fields at multiple grid-cell resolutions (27, 9, 3 km). For each domain, the basic coordinate grid for MM5 consisted of 103 grid cells along the x-axis (east–west) and 103 grid cells along the y-axis (north–south), with the center point of the model domain set to (39.58◦ N, 116.58◦ E) (Yang, Chen, & Zhang, 2006). CALPUFF is the Lagrangian, non-steady-state, gridded puff model to simulate the transport, dispersion, chemical reactions and deposition of the air pollutants in the atmosphere. The plume rise, stack tip downwash and vertical wind shear above the stack top were modeled in our case. PM dry deposition was included with the default data of size distribution (Scire et al., 2000). As to wet deposition, the scavenging coefficient for liquid precipitations was set at 1 × 10−4 s−1 for PM10 (Zhou et al., 2003). In the following figures Mentougou is located at the centerleft side and Beijing municipal area (Beijing, described in
Fig. 1. Source characteristics. (a) Domain configuration for CALPUFF and terrain characteristics (m); (b) emission distributions of PM10 in the study domain (g/(s m2 )). Also shown is the location of the observation site (open circle).
Fig. 1(b)) in the center-right and upper-right side of the maps. The southwest corner of Beijing municipal area is located at longitude 116.4◦ E, latitude 39.4◦ N; the northeast corner is located at longitude 117.0◦ E, latitude 39.8◦ N. While the southwest corner of Mentougou (Mentougou, described in Fig. 1(b)) is located at longitude 116.0◦ E, latitude 39.5◦ N; the northeast corner is located at longitude 116.4◦ E, latitude 39.7◦ N.
410
D. Yang et al. / China Particuology 5 (2007) 408–413
3. Source characteristics For this case study, we evaluated the aggregate impacts of 9 point sources and 96 area sources in Mentougou on a grid approximately 100 km × 100 km (Fig. 1(a)). The emission of PM10 from the point and area sources was estimated. The emission was calculated using the measured emission factors from Beijing Municipal Environmental Monitoring Station. The average emission factor of PM10 from coal burning was 1.7 kg/t coal and the emission of PM10 is 1631.65 t/year. The emission inventories for the case study were processed through the Sparse Matrix Operator Kernel Emissions model (SMOKE). This was done to prepare gridded, temporalized and speciated emissions for use in the CALPUFF. SMOKE is a state-of-the-art emission inventory processing system recently developed by the MCNC Supercomputing Center in North Carolina (Streets & Waldhoff, 2000). By using SMOKE, the authors developed an emission inventory of 1 km × 1 km meant to reflect current PM10 emissions. The location of the observation site and PM10 emission rate are shown in Fig. 1(b). 4. Model validation The modeling system was evaluated comprehensively to ensure reasonable estimates of ambient PM10 . We present a time series of the PM10 concentrations at the observation site. For example, Fig. 2 compares hourly averaged values (g/m3 ) (April 1, 01:00–11:00) for the observation site against the modeled estimates for the grid cell in which that observation site is located. From Fig. 2 one can observe that the model captured the time variation of PM10 in the simulated period reasonably well. Analysis of the time series of surface PM10 indicates that the horizontal distributions of PM10 are reasonably well simulated. We further examined the statistics data, including correlation coefficient and standard deviation (STD). Fig. 3 shows the scatter plots of the modeled concentrations versus the observed values. The straight line in Fig. 2 is computed by linear regression of
Fig. 3. Time series of CALPUFF predicted (open square) and observed PM10 concentrations (open circle) at the monitor site from 01:00 to 11:00 local time, April 1, 2004 (g/m3 ). Table 1 Statistical data of model predicted concentrations against the observed data Normalized correlation
Fraction of factor 2 (%)
Standard deviation error
0.86
78.60
0.03
the modeled concentrations according to the relation: CMOD = 0.8027COBS − 0.0467,
(1)
where CMOD and COBS are the modeled and observed concentrations, respectively (g/m3 ). Table 1 summarizes the statistical values for comparison. It shows that the model performed well in predicting PM10 . Note that the monitor and model are in fact providing PM10 levels at different locations, as the monitor reflects concentrations at a single point location and the modeled estimate reflects the average over a 1 km × 1 km grid cell. From Fig. 3 we also see that the model tends to underestimate PM10 . Besides uncertainties in meteorology and emissions, some plausible cause for the difference is that dry and wet depositions as well as chemical formation of PM10 might not be sufficiently represented in the model. Uncertainties could arise through influence of the model grid resolution on the calculated results. 5. Results and discussion
Fig. 2. Scatter plots of modeled versus observed concentrations (g/m3 ).
The simulated period covers April 1–7, 2004, but model results are discussed for the period from April 2 to 5. Contours of PM10 concentration and wind vectors are shown in Fig. 4. These contour plots demonstrate that the high PM10 concentrations appear in the source regions. Close examination of Fig. 4 confirms that those PM10 episodes were associated with continental outflow. For example, the high PM10 episode was associated with the occurrence of a cold air outbreak. At 14:00 on April 2, PM10 concentrations above 0.05 g/m3 were located southwest of Beijing. Because of enhancement from local emissions of
D. Yang et al. / China Particuology 5 (2007) 408–413
411
Fig. 4. Horizontal distributions of predicted PM10 concentrations (shaded, g/m3 ) and wind vectors (m/s). (a) April 2, 2004, 14:00; (b) April 3, 2004, 14:00; (c) April 4, 2004, 14:00; (d) April 5, 2004, 14:00.
PM10 , the maximum PM10 levels reached 70 g/m3 . By April 3, this polluted air mass moved eastward, and the center was located over the central part of Beijing. The maximum PM10 levels reached 30 g/m3 . By April 4, the air mass had moved northeast over the north part of Beijing. While the maximum PM10 levels decreased to 10.5 g/m3 as the air mass flowed out into the north area, which reflects the decrease in local emissions of PM10 . In the following 24 h the center enclosed by the
0.05 g/m3 contour line moved westward and covered the western part of Mentougou. We see in Fig. 4(a and d) that strong northeast winds swept the plume to the southwest of Beijing, while west winds pushed pollutants within the west out to east, and brought them eastward (shown in Fig. 4(b and c)). PM10 concentration distributions show heterogeneity characteristics. One can see from Fig. 4(a–d) that the meteorological situation is genuinely central to the distribution of PM10 . This is caused by
412
D. Yang et al. / China Particuology 5 (2007) 408–413
meteorological parameters impacting both the deposition processes and the transport phenomena, which govern the evolution of these distributions. As can be observed from Figs. 1 and 4, the study domain is characterized by significant topographic diversity and weather variations. Strongly stratified local meteorological circulation can result in very complex surface and boundary layer meteorology. Calm conditions, surface-based upper air inversions, and mountain-valley breezes are reasonably frequent. Mountain-valley features can result in significant potential for complex air pollution characteristic. As shown in Figs. 2 and 3, the model reproduced well the observed variations of PM10 , and simulated and observed values are in good agreement. Analysis of model results shows that PM10 concentrations exhibit pronounced variations in time and space. A process analysis was performed to illustrate the impacts of the transport and deposition upon PM10 concentrations. The budgets for PM10 show that ∼46% PM10 emitted is transported out of Mentougou and ∼53% deposited by dry and wet removal processes. The wind flow pattern shown in Fig. 4 indicates the typical pollution transport pathways. PM10 was transported to Beijing municipal region along these paths. Contributions of the Mentougou sources to the PM10 concentrations in Beijing municipal area are up to 0.1–15 g/m3 . 6. Summary The authors utilized the CALPUFF modeling system with meteorological fields from the MM5 meteorological model to examine the Mentougou outflow of PM10 over Beijing municipal area during the period of April 1–7, 2004. Comparison between observations and model calculations indicates that the model is capable of reproducing many of the observed features of PM10 impacts in Beijing. A PM10 budget analysis demonstrates that of the total PM10 , 46% was transported from Mentougou to the Beijing municipal area, while about 53% deposited in Mentougou. Analysis of model results illustrates that the pollutants emitted from Mentougou can be transported to Beijing municipal area along certain transport pathways, and PM10 concentration distributions show heterogeneity characteristics. It also shows that transport is the dominant term in regulating the PM10 spatial distribution, while dry and wet removal processes are the primary factors that control PM10 quantities. Acknowledgements This work was partly sponsored by Institute of Crustal Dynamics, CEA (project number: J2207812), National Science and Technology supported project (Grant no. 2006BAC01B0203-10) and Beijing Science and Technology Commission (project number: D0605046040191-204). The authors greatly appreciated the referees for their detailed comments on this manuscript. References Bennett, M. J., Yansura, M. E., Hornyik, I. G., Nall, J. M., Caniparoli, D. G., & Ashmore, C. G. (2002). Evaluation of the CALPUFF long-range transport
screening technique by comparison to refined CALPUFF results for several power plants in both the eastern and western United States. In Proceedings of the air & waste management association’s 95th annual conference. Paper #43454. Bromwich, D. H., Cassano, J. J., Kelin, T., Heinemann, G., Hines, K. M., Steffen, K., et al. (2001). Mesoscale modeling of katabatic winds over greenland with the polar MM5. Monthly Weather Review, 129, 2290– 2309. Douglas, S. G., & Kessler, R. C. (1988). User’s guide to the diagnostic wind model (Version 1.0). San Rafael, CA: Systems Applications, Inc. Intergovernmental Panel on Climate Change (IPCC). (1995). Climate change 1994: Radiative forcing of climate change and an evaluation of the IPCC IS92 emission scenario (p. 339). New York: Cambridge University Press. Levy, J. I., Spengler, J. D., Hlinka, D., Sullivan, D., & Moon, D. (2002). Using CALPUFF to evaluate the impacts of power plant emissions in Illinois: Model sensitivity and implications. Atmospheric Environment, 36, 1063–1075. Scire, J. S., Robe, F. R., Fernau, M. E., & Yamartino, R. J. (2000). A user’s guide for the CALMET meteorological model (Version 5). Concord, MA: Earth Tech, Inc. Scire, J. S., Strimaitis, D. G., & Yamartino, R. J. (1999). A user’s guide for the CALPUFF dispersion model (Version 5.0). Concord, MA: Earth Tech, Inc. Sokolik, I. N., & Toon, O. B. (1996). Direct radiative forcing by anthropogenic airborne mineral aerosols. Nature, 381(20), 681–683. Stauffer, D. R., & Seaman, N. L. (1994). Multiscale four dimensional data assimilation. Journal of Applied Meteorology, 33, 416–434. Streets, D. G., Bond, T. C., Carmichael, G. R., Fernandes, S. D., Fu, Q., He, D., et al. (2003). An inventory of gaseous and primary aerosol emissions in Asia in the year 2000. Journal of Geophysical Research, 108(D21), 8809. doi:10.1029/2002JD003093 Streets, D. G., & Waldhoff, S. T. (2000). Present and future emissions of air pollutants in China: SO2 , NOx and CO. Atmospheric Environment, 34, 363–374. Yang, D. X., Chen, G. C., Liu, C., & Zhao, X. H. (2006). The Lagrangian air quality modeling system and simulation verification of its meteorological parameters. Journal of Southwest Agricultural University (Natural Science), 28(5), 776–781 (in Chinese). Yang, D. X., Chen, G. C., & Zhang, R. J. (2006). Estimated public health exposure to H2 S emissions from a sour gas well blowout in Kaixian county, China. Aerosol and Air Quality Research, 6(4), 430–443. Zhang, M., Uno, I., Carmichael, G. R., Akimoto, H., Wang, Z., Tang, Y., et al. (2003). Large-scale structure of trace gas and aerosol distributions over the western Pacific ocean during the transport and chemical evolution over the Pacific (TRACE-P) experiment. Journal of Geophysical Research, 108(D21), 8820–8823. Zhang, R. J., Wang, M. X., Sheng, L. F., Kanai, Y., & Ohta, A. (2004). Seasonal characterization of dust days, mass concentration and dry deposition of atmospheric aerosals over Qingdao, China. China Particuology, 2(5), 196–199. Zhang, R. J., Wang, M. X., & Xia, X. G. (2002). Chemical composition of aerosols in winter/spring in Beijing. Journal of Environmental Sciences, 14(1), 7–11. Zhou, Y., Levy, J. I., Hammitt, J. K., & Evans, J. S. (2003). Estimating population exposure to power plant emissions using CALPUFF: A case study in Beijing, China. Atmospheric Environment, 37(6), 815–826.
Glossary CALPUFF: an advanced non-steady-state meteorological and air quality modeling system developed by ASG scientists. The main components of the modeling system are CALMET (a diagnostic three-dimensional meteorological model), CALPUFF (an air quality dispersion model), and CALPOST (a post-processing package).
D. Yang et al. / China Particuology 5 (2007) 408–413 CALMET: one of the main components of the CALPUFF modeling system, a diagnostic three-dimensional meteorological model. MM5: fifth-generation PSU/NCAR mesoscale model. FDDA: four-dimensional data assimilation. NCEP: National Centers for Environmental Prediction.
NCAR: National Center for Atmospheric Research. SMOKE: Sparse Matrix Operator Kernel Emissions model system. AGL: above ground level. LCC: Lambert Conformal Conic.
413
China Particuology 5 (2007) 408–413
Transport of airborne particulate matters originating from Mentougou, Beijing, China Duoxing Yang a,∗ , Yongwei Han b , Jixi Gao b , Jesse Th´e c a
Institute of Crust Dynamics, CEA, Beijing 100085, China Chinese Academy of Environmental Sciences, Beijing 100012, China c Lakes Environmental Software Inc., 419-3 Phillips St., Waterloo, ON N2L 3X2, Canada b
Received 13 March 2007; accepted 7 July 2007
Abstract In this study, a coupled regional air quality modeling system is applied to investigate the time spatial variations in airborne particulate matters (PM10 ), originating from Mentougou to Beijing municipal area in the period of April 1–7, 2004, and the influences of complex terrain and meteorological conditions upon boundary layer structure and PM10 concentration distributions. An intercomparison of the performance with CALPUFF against the observed data is presented and an examination of scatter plots is provided. The statistics show that the correlation coefficient and STD between the modeled and observed data are 0.86 and 0.03, respectively. Analysis of model results illustrates that the pollutants emitted from Mentougou can be transported to Beijing municipal area along certain transport pathways, and PM10 concentration distributions show heterogeneity characteristics. Contributions of the Mentougou sources to the PM10 concentrations in Beijing municipal area are up to 0.1–15 g/m3 . © 2007 Chinese Society of Particuology and Institute of Process Engineering, Chinese Academy of Sciences. Published by Elsevier B.V. All rights reserved. Keywords: Airborne particulate matters; CALPUFF; PM10 ; SMOKE; Emission inventory
1. Introduction Atmospheric aerosols consist of a mixture of solid and liquid particles suspended in ambient air (Zhang, Wang, Sheng, Kanai, & Ohta, 2004; Zhang, Wang, & Xia, 2002), ranging in size from the smallest superfine nano-meters (nm) particles, to coarse particles, with diameters of several micrometers (m) or more (Zhang et al., 2003). Their importance arises from their ability to interact with visible radiation and to affect cloud properties, as well as their long atmospheric residence times (IPCC, 1995). ‘Particulate matter’, also known as particle pollution or PM, is a complex mixture of extremely small particles and liquid droplets. Particle pollution is made up of a set of components, including acids (such as nitrates and sulfates), organic chemicals, metals, and soil or dust particles (Sokolik & Toon, 1996). The size of particles is directly linked to their potential for causing health problems. PM particles 10 m in size or smaller, referred to as PM10 , are typically formed by “crustal” or earth-based
∗
Corresponding author. Tel.: +86 10 62842631; fax: +86 10 62842631. E-mail address: [email protected] (D. Yang).
material and enter the air through a variety of actions including “entrainment” into the atmosphere as wind blown dust. The even smaller or “fine” PM material, referred to as PM2.5 (2.5 m or smaller), are understood to be more a product of combustion (Streets et al., 2003). PM2.5 is believed to penetrate deeper into the lungs and remain lodged there rather than exhaled, causing adverse impacts on health. In the last decade a variety of transport and deposition models have been applied to address PM transport and deposition over Beijing (Zhou, Levy, Hammitt, & Evans, 2003). This study is another attempt to investigate the behavior of PM10 over Beijing, and to examine the relative role of meteorological fields and removal mechanisms in regulating PM10 behavior with CALPUFF (Bennett et al., 2002; Scire, Strimaitis, & Yamartino, 1999) model on the basis of a newly estimated emission inventory for Mentougou with a resolution of 1 km × 1 km. In this study both temporal and spatial variations in airborne particulate matters (PM10 ) as well as transport processes of PM10 originating from Mentougou to Beijing municipal area in the period of April 1–7, 2004, and the influences of complex terrain and meteorological conditions upon PM10 concentration distributions are discussed.
1672-2515/$ – see inside back cover © 2007 Chinese Society of Particuology and Institute of Process Engineering, Chinese Academy of Sciences. Published by Elsevier B.V. All rights reserved.
doi:10.1016/j.cpart.2007.07.003
D. Yang et al. / China Particuology 5 (2007) 408–413
409
2. Model configuration and setup The initial phase of CALPUFF which we used for our primary model involves deriving meteorological files using CALMET (Scire, Robe, Fernau, & Yamartino, 2000; Yang, Chen, Liu, & Zhao, 2006), which is a simple diagnostic windfield model. Much of the structure in the wind fields is determined by CALMET using its diagnostic wind field module (Douglas & Kessler, 1988). As shown in Fig. 1, the basic coordinate grid for CALPUFF/CALMET consists of 100 grid cells along the x-axis (east–west) and 100 grid cells along the y-axis (north–south), spaced 1 km apart. The coordinate system was converted to a Lambert Conical Projection grid. The 10 vertical layers incorporated into the CALMET processing had heights of 20, 50, 100, 150, 300, 600, 1000, 1500, 2200 and 3000 m. The MM5 mesoscale model was used in this study to develop high-resolution, three-dimensional meteorological fields (i.e., wind, temperature, pressure, etc.) through FDDA simulations (Stauffer & Seaman, 1994). MM5 generated the meteorological fields, which we used to improve the wind distribution results from CALMET. This is an important step since the MM5 model can resolve wind features caused by topography, such as terrain channeling and gravity-driven slope flows (Levy, Spengler, Hlinka, Sullivan, & Moon, 2002; Zhou et al., 2003). The first-guess atmosphere data for MM5 are extracted from the NCAR/NCEP FNL archives. The NCEP Final Analysis (FNL) data archived at NCAR is for every 6 h at a spatial resolution of 1◦ × 1◦ at standard pressure levels under 100 hPa. This data set includes two-dimensional variables, sea surface temperature, sea level pressure, three-dimensional variables of temperature, geo-potential height, U and V velocity components, and relative humidity (Bromwich et al., 2001). We combined the MM5 prognostic model (Douglas & Kessler, 1988) outputs with mesoscale data assimilation systems for 7 days (April 1, 2004, 00:00:00–April 7, 2004, 23:00:00). The present MM5 model domain has 32 vertical levels, going up to about 13 km AGL, with vertical grid spacing stretched from about 20 m near the ground to 800 m near the top of the domain. This allowed CALMET to interpolate from a higher-to-a-lower-resolution grid (since CALMET uses 10 vertical layers). One-way nesting was used to generate ambient wind fields at multiple grid-cell resolutions (27, 9, 3 km). For each domain, the basic coordinate grid for MM5 consisted of 103 grid cells along the x-axis (east–west) and 103 grid cells along the y-axis (north–south), with the center point of the model domain set to (39.58◦ N, 116.58◦ E) (Yang, Chen, & Zhang, 2006). CALPUFF is the Lagrangian, non-steady-state, gridded puff model to simulate the transport, dispersion, chemical reactions and deposition of the air pollutants in the atmosphere. The plume rise, stack tip downwash and vertical wind shear above the stack top were modeled in our case. PM dry deposition was included with the default data of size distribution (Scire et al., 2000). As to wet deposition, the scavenging coefficient for liquid precipitations was set at 1 × 10−4 s−1 for PM10 (Zhou et al., 2003). In the following figures Mentougou is located at the centerleft side and Beijing municipal area (Beijing, described in
Fig. 1. Source characteristics. (a) Domain configuration for CALPUFF and terrain characteristics (m); (b) emission distributions of PM10 in the study domain (g/(s m2 )). Also shown is the location of the observation site (open circle).
Fig. 1(b)) in the center-right and upper-right side of the maps. The southwest corner of Beijing municipal area is located at longitude 116.4◦ E, latitude 39.4◦ N; the northeast corner is located at longitude 117.0◦ E, latitude 39.8◦ N. While the southwest corner of Mentougou (Mentougou, described in Fig. 1(b)) is located at longitude 116.0◦ E, latitude 39.5◦ N; the northeast corner is located at longitude 116.4◦ E, latitude 39.7◦ N.
410
D. Yang et al. / China Particuology 5 (2007) 408–413
3. Source characteristics For this case study, we evaluated the aggregate impacts of 9 point sources and 96 area sources in Mentougou on a grid approximately 100 km × 100 km (Fig. 1(a)). The emission of PM10 from the point and area sources was estimated. The emission was calculated using the measured emission factors from Beijing Municipal Environmental Monitoring Station. The average emission factor of PM10 from coal burning was 1.7 kg/t coal and the emission of PM10 is 1631.65 t/year. The emission inventories for the case study were processed through the Sparse Matrix Operator Kernel Emissions model (SMOKE). This was done to prepare gridded, temporalized and speciated emissions for use in the CALPUFF. SMOKE is a state-of-the-art emission inventory processing system recently developed by the MCNC Supercomputing Center in North Carolina (Streets & Waldhoff, 2000). By using SMOKE, the authors developed an emission inventory of 1 km × 1 km meant to reflect current PM10 emissions. The location of the observation site and PM10 emission rate are shown in Fig. 1(b). 4. Model validation The modeling system was evaluated comprehensively to ensure reasonable estimates of ambient PM10 . We present a time series of the PM10 concentrations at the observation site. For example, Fig. 2 compares hourly averaged values (g/m3 ) (April 1, 01:00–11:00) for the observation site against the modeled estimates for the grid cell in which that observation site is located. From Fig. 2 one can observe that the model captured the time variation of PM10 in the simulated period reasonably well. Analysis of the time series of surface PM10 indicates that the horizontal distributions of PM10 are reasonably well simulated. We further examined the statistics data, including correlation coefficient and standard deviation (STD). Fig. 3 shows the scatter plots of the modeled concentrations versus the observed values. The straight line in Fig. 2 is computed by linear regression of
Fig. 3. Time series of CALPUFF predicted (open square) and observed PM10 concentrations (open circle) at the monitor site from 01:00 to 11:00 local time, April 1, 2004 (g/m3 ). Table 1 Statistical data of model predicted concentrations against the observed data Normalized correlation
Fraction of factor 2 (%)
Standard deviation error
0.86
78.60
0.03
the modeled concentrations according to the relation: CMOD = 0.8027COBS − 0.0467,
(1)
where CMOD and COBS are the modeled and observed concentrations, respectively (g/m3 ). Table 1 summarizes the statistical values for comparison. It shows that the model performed well in predicting PM10 . Note that the monitor and model are in fact providing PM10 levels at different locations, as the monitor reflects concentrations at a single point location and the modeled estimate reflects the average over a 1 km × 1 km grid cell. From Fig. 3 we also see that the model tends to underestimate PM10 . Besides uncertainties in meteorology and emissions, some plausible cause for the difference is that dry and wet depositions as well as chemical formation of PM10 might not be sufficiently represented in the model. Uncertainties could arise through influence of the model grid resolution on the calculated results. 5. Results and discussion
Fig. 2. Scatter plots of modeled versus observed concentrations (g/m3 ).
The simulated period covers April 1–7, 2004, but model results are discussed for the period from April 2 to 5. Contours of PM10 concentration and wind vectors are shown in Fig. 4. These contour plots demonstrate that the high PM10 concentrations appear in the source regions. Close examination of Fig. 4 confirms that those PM10 episodes were associated with continental outflow. For example, the high PM10 episode was associated with the occurrence of a cold air outbreak. At 14:00 on April 2, PM10 concentrations above 0.05 g/m3 were located southwest of Beijing. Because of enhancement from local emissions of
D. Yang et al. / China Particuology 5 (2007) 408–413
411
Fig. 4. Horizontal distributions of predicted PM10 concentrations (shaded, g/m3 ) and wind vectors (m/s). (a) April 2, 2004, 14:00; (b) April 3, 2004, 14:00; (c) April 4, 2004, 14:00; (d) April 5, 2004, 14:00.
PM10 , the maximum PM10 levels reached 70 g/m3 . By April 3, this polluted air mass moved eastward, and the center was located over the central part of Beijing. The maximum PM10 levels reached 30 g/m3 . By April 4, the air mass had moved northeast over the north part of Beijing. While the maximum PM10 levels decreased to 10.5 g/m3 as the air mass flowed out into the north area, which reflects the decrease in local emissions of PM10 . In the following 24 h the center enclosed by the
0.05 g/m3 contour line moved westward and covered the western part of Mentougou. We see in Fig. 4(a and d) that strong northeast winds swept the plume to the southwest of Beijing, while west winds pushed pollutants within the west out to east, and brought them eastward (shown in Fig. 4(b and c)). PM10 concentration distributions show heterogeneity characteristics. One can see from Fig. 4(a–d) that the meteorological situation is genuinely central to the distribution of PM10 . This is caused by
412
D. Yang et al. / China Particuology 5 (2007) 408–413
meteorological parameters impacting both the deposition processes and the transport phenomena, which govern the evolution of these distributions. As can be observed from Figs. 1 and 4, the study domain is characterized by significant topographic diversity and weather variations. Strongly stratified local meteorological circulation can result in very complex surface and boundary layer meteorology. Calm conditions, surface-based upper air inversions, and mountain-valley breezes are reasonably frequent. Mountain-valley features can result in significant potential for complex air pollution characteristic. As shown in Figs. 2 and 3, the model reproduced well the observed variations of PM10 , and simulated and observed values are in good agreement. Analysis of model results shows that PM10 concentrations exhibit pronounced variations in time and space. A process analysis was performed to illustrate the impacts of the transport and deposition upon PM10 concentrations. The budgets for PM10 show that ∼46% PM10 emitted is transported out of Mentougou and ∼53% deposited by dry and wet removal processes. The wind flow pattern shown in Fig. 4 indicates the typical pollution transport pathways. PM10 was transported to Beijing municipal region along these paths. Contributions of the Mentougou sources to the PM10 concentrations in Beijing municipal area are up to 0.1–15 g/m3 . 6. Summary The authors utilized the CALPUFF modeling system with meteorological fields from the MM5 meteorological model to examine the Mentougou outflow of PM10 over Beijing municipal area during the period of April 1–7, 2004. Comparison between observations and model calculations indicates that the model is capable of reproducing many of the observed features of PM10 impacts in Beijing. A PM10 budget analysis demonstrates that of the total PM10 , 46% was transported from Mentougou to the Beijing municipal area, while about 53% deposited in Mentougou. Analysis of model results illustrates that the pollutants emitted from Mentougou can be transported to Beijing municipal area along certain transport pathways, and PM10 concentration distributions show heterogeneity characteristics. It also shows that transport is the dominant term in regulating the PM10 spatial distribution, while dry and wet removal processes are the primary factors that control PM10 quantities. Acknowledgements This work was partly sponsored by Institute of Crustal Dynamics, CEA (project number: J2207812), National Science and Technology supported project (Grant no. 2006BAC01B0203-10) and Beijing Science and Technology Commission (project number: D0605046040191-204). The authors greatly appreciated the referees for their detailed comments on this manuscript. References Bennett, M. J., Yansura, M. E., Hornyik, I. G., Nall, J. M., Caniparoli, D. G., & Ashmore, C. G. (2002). Evaluation of the CALPUFF long-range transport
screening technique by comparison to refined CALPUFF results for several power plants in both the eastern and western United States. In Proceedings of the air & waste management association’s 95th annual conference. Paper #43454. Bromwich, D. H., Cassano, J. J., Kelin, T., Heinemann, G., Hines, K. M., Steffen, K., et al. (2001). Mesoscale modeling of katabatic winds over greenland with the polar MM5. Monthly Weather Review, 129, 2290– 2309. Douglas, S. G., & Kessler, R. C. (1988). User’s guide to the diagnostic wind model (Version 1.0). San Rafael, CA: Systems Applications, Inc. Intergovernmental Panel on Climate Change (IPCC). (1995). Climate change 1994: Radiative forcing of climate change and an evaluation of the IPCC IS92 emission scenario (p. 339). New York: Cambridge University Press. Levy, J. I., Spengler, J. D., Hlinka, D., Sullivan, D., & Moon, D. (2002). Using CALPUFF to evaluate the impacts of power plant emissions in Illinois: Model sensitivity and implications. Atmospheric Environment, 36, 1063–1075. Scire, J. S., Robe, F. R., Fernau, M. E., & Yamartino, R. J. (2000). A user’s guide for the CALMET meteorological model (Version 5). Concord, MA: Earth Tech, Inc. Scire, J. S., Strimaitis, D. G., & Yamartino, R. J. (1999). A user’s guide for the CALPUFF dispersion model (Version 5.0). Concord, MA: Earth Tech, Inc. Sokolik, I. N., & Toon, O. B. (1996). Direct radiative forcing by anthropogenic airborne mineral aerosols. Nature, 381(20), 681–683. Stauffer, D. R., & Seaman, N. L. (1994). Multiscale four dimensional data assimilation. Journal of Applied Meteorology, 33, 416–434. Streets, D. G., Bond, T. C., Carmichael, G. R., Fernandes, S. D., Fu, Q., He, D., et al. (2003). An inventory of gaseous and primary aerosol emissions in Asia in the year 2000. Journal of Geophysical Research, 108(D21), 8809. doi:10.1029/2002JD003093 Streets, D. G., & Waldhoff, S. T. (2000). Present and future emissions of air pollutants in China: SO2 , NOx and CO. Atmospheric Environment, 34, 363–374. Yang, D. X., Chen, G. C., Liu, C., & Zhao, X. H. (2006). The Lagrangian air quality modeling system and simulation verification of its meteorological parameters. Journal of Southwest Agricultural University (Natural Science), 28(5), 776–781 (in Chinese). Yang, D. X., Chen, G. C., & Zhang, R. J. (2006). Estimated public health exposure to H2 S emissions from a sour gas well blowout in Kaixian county, China. Aerosol and Air Quality Research, 6(4), 430–443. Zhang, M., Uno, I., Carmichael, G. R., Akimoto, H., Wang, Z., Tang, Y., et al. (2003). Large-scale structure of trace gas and aerosol distributions over the western Pacific ocean during the transport and chemical evolution over the Pacific (TRACE-P) experiment. Journal of Geophysical Research, 108(D21), 8820–8823. Zhang, R. J., Wang, M. X., Sheng, L. F., Kanai, Y., & Ohta, A. (2004). Seasonal characterization of dust days, mass concentration and dry deposition of atmospheric aerosals over Qingdao, China. China Particuology, 2(5), 196–199. Zhang, R. J., Wang, M. X., & Xia, X. G. (2002). Chemical composition of aerosols in winter/spring in Beijing. Journal of Environmental Sciences, 14(1), 7–11. Zhou, Y., Levy, J. I., Hammitt, J. K., & Evans, J. S. (2003). Estimating population exposure to power plant emissions using CALPUFF: A case study in Beijing, China. Atmospheric Environment, 37(6), 815–826.
Glossary CALPUFF: an advanced non-steady-state meteorological and air quality modeling system developed by ASG scientists. The main components of the modeling system are CALMET (a diagnostic three-dimensional meteorological model), CALPUFF (an air quality dispersion model), and CALPOST (a post-processing package).
D. Yang et al. / China Particuology 5 (2007) 408–413 CALMET: one of the main components of the CALPUFF modeling system, a diagnostic three-dimensional meteorological model. MM5: fifth-generation PSU/NCAR mesoscale model. FDDA: four-dimensional data assimilation. NCEP: National Centers for Environmental Prediction.
NCAR: National Center for Atmospheric Research. SMOKE: Sparse Matrix Operator Kernel Emissions model system. AGL: above ground level. LCC: Lambert Conformal Conic.
413