Asian anthropogenic emissions and decadal trends in springtime tropospheric ozone over Japan:

Transcription

1 GEOPHYSICAL RESEARCH LETTERS, VOL. 36, L23802, doi: /2009gl041382, 2009 Asian anthropogenic emissions and decadal trends in springtime tropospheric ozone over Japan: Hiroshi Tanimoto, 1 Toshimasa Ohara, 1 and Itsuhi Uno 2 Received 15 October 2009; accepted 26 October 2009; published 1 December [1] We examine springtime ozone trends at nine remote locations in Japan during the last decade. The observed decadal ozone trends are relatively small at surface sites but are substantially larger at a mountainous site. We use a regional chemistry-transport model to explore the observed changes and how changes in Asian anthropogenic emissions have contributed to the observed increasing trends. The model with yearly-dependent regional emissions successfully reproduces the levels, variability, and interannual variations of ozone at all the surface sites. It predicts increasing trends at the mountainous site, suggesting that increasing Asian anthropogenic emissions account for about half the observed increase. However, the discrepancy between the observation and model results after 2003 (the time of largest emission increase) suggests significant underestimation of the actual growth of the Asian anthropogenic emissions and/or incompleteness in the modeling of pollution export from continental Asia. Citation: Tanimoto, H., T. Ohara, and I. Uno (2009), Asian anthropogenic emissions and decadal trends in springtime tropospheric ozone over Japan: , Geophys. Res. Lett., 36, L23802, doi: /2009gl Introduction [2] Historical records of lower-tropospheric O 3 in Europe have shown an apparent increase in the background O 3 levels during the twentieth century [Volz and Kley, 1988; Staehelin et al., 1994]. There are a number of observational reports of increases in boundary layer O 3 during recent decades in the northern hemisphere, with growth rates ranging from 0.5 to 0.8 ppbv yr 1 [e.g., Jaffe et al., 2003; Lelieveld et al., 2004; Oltmans et al., 2006; Zbinden et al., 2006; Derwent et al., 2007; Parrish et al., 2009]. [3] Chemistry-transport model (CTM) simulations have not yet quantitatively explained the observed increases. Simulated historical O 3 levels for the preindustrial period are significantly higher than observed, perhaps due to overestimates by the models of natural sources such as lightning or calibration errors in early measurements. A global CTM simulation predicted an increase of only 2 ppbv over [Fiore et al., 2002], in contrast to 10 ppbv increase during the mid 1980 s to 2002 observed over California in spring [Jaffe et al., 2003]. The major causes for this discrepancy have not yet been identified, and reconciling the observed trends by models is still challenging. 1 Asian Environment Research Group, National Institute for Environmental Studies, Tsukuba, Japan. 2 Research Institute for Applied Mechanics, Kyushu University, Kasuga, Japan. [4] In East Asia, especially China, emissions of O 3 precursors (nitrogen oxides (NOx = NO + NO 2 ) and volatile organic compounds (VOCs)) have increased greatly in recent decades [Zhang et al., 2007; Ohara et al., 2007]. The growth of NOx emissions in China has been more dramatic since 2000 than during the 1990 s. The resulting impact on the regional air quality downstream is significant and needs to be assessed. [5] We use surface O 3 measurements at Japanese remote stations and a regional CTM to investigate springtime (March, April, and May) O 3 trends during the last decade in terms of anthropogenic source regions in East Asia. We address the question of whether the springtime O 3 over Japan reflects the rapidly changing anthropogenic emissions in continental Asia and quantitatively evaluate the contributions of emission increases to the decadal O 3 trend, particularly after Measurements and Modeling 2.1. Sites and Methods [6] Measurement data used in this work are compiled from two regional monitoring programs operated in the international framework. One is operated by the Ministry of the Environment of Japan as part of the Acid Deposition Monitoring Network in East Asia (EANET) program [Network Center for EANET, 2007], and the other by the Japan Meteorological Agency as part of the Global Atmosphere Watch (GAW) programme of the World Meteorological Organization [Japan Meteorological Agency, 2005].We focus on nine remote sites in different geographical regions where quality-controlled long-term records are available. Of these, six sites represent continental rim sites (i.e., Rishiri Island, Cape Tappi, Sado Island, Oki Island, Okinawa Island, Yonagunijima Island), two represent open ocean sites (i.e., Ogasawara Island, Minamitorishima Island), and one a mountainous site (Mt. Happo at 1850 m) in the Japanese mainland (see Table S1 and Figure S1 for details including site codes that will be referred to hereafter). 3 [7] Measurement details were given in technical reports on the programs [Acid Deposition and Oxidants Research Center, 2001; World Meteorological Organization, 2007]. The mixing ratios of O 3 were determined using photometric instruments based on absorption at nm emitted by a low-pressure mercury lamp. The instrument was periodically referenced to a working standard placed at each site. The standards referenced in both EANET and GAW programs routinely went through intercomparison with a reference photometer, which was referenced to the Standard Reference Photometer (SRP) #2 at the National Institute of Copyright 2009 by the American Geophysical Union /09/2009GL Auxiliary materials are available in the HTML. doi: / 2009GL L of5

2 Figure 1. (left) Decadal changes in NOx and VOC emissions from East Asia including China and Japan, estimated by the updated version of the Regional Emission inventory in Asia (REAS). (right) Sectoral distributions for China s emissions are shown. The emissions prescribed for the 2007 mode runs were adopted from 2006 (as indicated by shaded bars). Standards and Technology (NIST). The data presented here are thus traceable to the NIST-SRP standard. The data were scrutinized in the post-analysis, including Quality Assurance/Quality Control (QA/QC) activity, and were then made available to the public Regional CTM and Emissions Inventory [8] The three-dimensional regional CTM used in this study is based on the Models-3 Community Multi-scale Air Quality (CMAQ) modeling system [Byun and Schere, 2006]. Our previous application and evaluation of CMAQ are found elsewhere [Tanimoto et al., 2005; Uno et al., 2007]. The horizontal and vertical resolutions are km 2 and 14 layers up to 23 km, respectively. The SAPRC-99 scheme is applied for gas-phase chemistry. The monthly averaged lateral boundary conditions are obtained from a global CTM [Sudo et al., 2002]. Stratospheric O 3 influx is thus taken into account, but no interannual variation is assumed. We use a yearly-dependent emissions inventory from the Regional Emission inventory in Asia (REAS), which was updated to 2006 by using the latest statistics of energy consumption and industrial activities, and employ the same emission factors and removal efficiencies as for year Detailed description of REAS and comparison with other global and regional inventories including EDGAR and David Streets are found elsewhere [Ohara et al., 2007]. Biomass burning sources are supplied climatologically from Streets et al. [2003]. [9] Recent emissions of O 3 precursors from East Asia have rapidly increased during the last decade (Figure 1). Chinese emissions of NOx and VOCs have contributed approximately 80% to the East Asian emissions, with NOx emissions showing accelerating growth after 2000, mainly due to the growth in power plants and transport sectors. In contrast, the Japanese emissions have been relatively small (10% of the Chinese emissions for NOx and VOCs) and almost constant. Two types of multi-year sensitivity simulations to the East Asian anthropogenic emissions are performed: (1) with the emissions inventory for each year (i.e., 1998 to 2007; for 2007, the 2006 inventory is tentatively used due to unavailability of statistical data) and (2) with a fixed emissions inventory from the year Observed and Modeled Ozone Trends in Spring [10] Springtime O 3 trends observed at nine Japanese sites from 1998 to 2007 are exhibited in Figure 2. There are distinct differences in the O 3 levels between the sites that depend on the distance from the continent and latitude. The O 3 levels at the continental rim sites (i.e., RIS, TPI, SDO, OKI, ONW, YON) are substantially higher than at the open ocean sites (i.e., OGS and MNM). The six continental rim sites exhibit the O 3 level of 50 ppbv for the year Parrish et al. [2009] reported springtime O 3 near 40 ppbv for the marine boundary layer sites on the eastern Pacific. These observations suggest significant O 3 gradient in the marine boundary layer across the Pacific. The highest O 3 levels are observed at the mountain site, HPO, with median O 3 levels exceeding 60 ppbv. At the eight surface sites (excluding HPO), the O 3 trends observed from 1998 to 2007 are relatively small or statistically insignificant. This is likely due to a combined effect of large interannual variability driven by meteorology in the boundary layer and the relatively short time record (i.e., 10 years). In contrast, particularly after 2000, a distinct increase of O 3 is observed at HPO. The trend in the median from 1998 to 2007 is statistically significant (R 2 = 0.79) associated with an increase rate of 1.25 ± 0.53 ppbv yr 1. [11] The model incorporating year-dependent emissions inventory simulates the levels and the range of variability for springtime O 3 in the boundary layer quite well. Substantial interannual variations are also accounted for in the 2of5

3 Figure 2. Trends of springtime tropospheric ozone from 1998 to Plotted are the ozone mixing ratios observed at Japanese stations (black), and overlaid with a regional chemistry-transport model simulations with yearly-dependent anthropogenic emissions in Asia (gray). Observed trends are shown by median (black lines) and interquartile range (vertical bars), and the modeled trends by median (gray lines) and interquartile range (shaded). RIS, Rishiri Island; TPI, Cape Tappi; SDO, Sado Island; OKI, Oki Island; ONW, Okinawa Island; YON, Yonagunijima Island; OGS, Ogasawara Island; MNM, Minamitorishima Island; HPO, Mt. Happo. See Table S1 for detailed information for individual sites. The coefficient of determination (r 2 ), slopes (s, ppbv/year) and intercepts for the year 2000 (i, ppbv) with their 95% confidence limits obtained by t-test are given for observations (obs) and model (mdl) results. simulation (R 2 = 0.45). The good agreement over the wide latitudinal coverage demonstrates the capability and usefulness of the model in our study. For the high altitude HPO site, the model is able to produce an increase in O 3 (statistically significant (R 2 = 0.67) rate of 0.52 ± 0.30 ppbv yr 1 ) but not the observed growth rate. [12] It is interesting to note, however, that O 3 levels in the lower troposphere (1.5 3 km) over Tsukuba (Japan) showed an increase in the 1970 s and 80 s, remained relatively constant in the early 1990 s and actually started to decline after 2000 (to 2004), apparently inconsistent with the recent increase in the Asian emissions, as noted by Oltmans et al. [2006]. Zbinden et al. [2006] reported from the MOZAIC aircraft measurements a positive trend of 1.0% yr 1 (i.e., 0.6 ppbv yr 1 ) for mid-tropospheric (2 8 km) springtime O 3 over Japan since 1994, followed by a slight decline after These measurements were obtained close to major cities in Japan (i.e., Tsukuba is 50 km away from Tokyo; MOZAIC flights near Tokyo, Nagoya, and Osaka), suggesting that they were significantly influenced by Japanese emissions, which have remained relatively constant over the last decade. This leads to the suggestion that the observed O 3 increase at HPO is indicative of the capability of this remote high-altitude site to effectively detect East Asian emission changes. 4. Decadal Trend at Mt. Happo [13] Tanimoto [2009] noted that the observed O 3 increases at HPO are larger at higher percentiles in the probability distribution, suggesting impacts from regional O 3 pollution driven by increasing anthropogenic emissions in East Asia. Following this implication, we examine the modeled O 3 trends at HPO from 50th to 95th percentiles, deconvolute impacts from emissions and meteorology, and discuss their contributions quantitatively, as shown in Figure 3. [14] For the period, the observed growth rates at HPO are as large as 1 to 2 ppbv yr 1 (12 to 22 ppbv decade 1 ) at the 50th 95th percentiles. The modeled growth rates for the period at HPO driven by growing East Asian emissions are 0.5 to 1 ppbv yr 1 (5.2 to 9.6 ppbv decade 1 ) at the 50th 95th percentiles, with greater increases associated with the 75th and 95th percentiles. The model shows larger O 3 increase after 2003, in agreement with the observation. However, the modeled 3of5

4 increase at HPO cannot be explained by the assigned anthropogenic emissions from East Asia. Figure 3. Decadal changes in springtime ozone at Mt. Happo (HPO) at (top) 50th, (middle) 75th, and (bottom) 95th percentiles in the probability distributions of ozone. MyyEyy and MyyE00 indicate model runs with yearlydependent and constant emissions inventory (year 2000), respectively, with yearly-dependent meteorological fields. MyyEyy MyyE00 denotes contributions from anthropogenic emissions only. All the data are normalized to those in growth rates are substantially lower than the observation, with noticeable discrepancy after The model simulations with constant emissions reveal that transport causes interannual variability of 2 ppbv at the 50th 95th percentiles; however, changes in emissions are the major cause of the observed variation in O 3 at HPO after [15] The differences between the two sets of multi-year model simulation results indicate that the anthropogenic emission and atmospheric transport variations contributed almost equally to the observed O 3 growth during , but that the anthropogenic emission contribution became dominant after 2003, amounting to ppbv yr 1 ( ppbv decade 1 ), showing HPO to be an ideal location for detecting rapid changes in anthropogenic emissions in East Asia. However, about half of the observed O 3 5. Discussion: Uncertainties in Emissions and Transport [16] The observation-vs-model discrepancy described above suggests existence of additional sources and/or processes which are not well implemented in our model. [17] Since a regional model is used in the present study, long-range transport from outside Asia is a possibility. Considering that the intercontinental O 3 transport from Europe to East Asia mainly occurs in the boundary layer over the Eurasian continent [Wild and Akimoto, 2001], and that European and North American NOx emissions have declined over the recent decades, intercontinental transport to East Asia is unlikely to have any significant influence on the decadal O 3 variation at HPO. [18] Ordóñez et al. [2007] suggested that changes in the lowermost stratospheric O 3 could have influenced the variability of O 3 levels in the lower free troposphere over Europe during the 1990 s, particularly in winter and spring. However, Terao et al. [2008] showed that although tropospheric O 3 can be highly influenced by lower stratospheric O 3 over Canada and Europe, the influence is much smaller over East Asia, and suggested that pollution transport is a dominant contributor. Previous measurements at HPO suggest that the variations of springtime O 3 at the site are mainly controlled by long-range transport of photochemically produced O 3 in the troposphere [Kajii et al., 1998]. Thus, stratospheric influence is unlikely one of the main causes of the rapidly increasing O 3 observed at HPO during the last decade. [19] Fu et al. [2007] found that bottom-up VOC emissions from China by Streets et al. [2003] were underestimated by a factor of five for biomass burning sources. Emissions of reactive VOCs from agricultural burning in the North China Plain are largest in June, but also substantial as early as March over southern China. Fu et al. s [2007] updated emissions inventory for Asian VOCs has resulted in enhancements of springtime monthly mean afternoon surface O 3 by 5 20 and 1 5 ppbv over central China and Japan, respectively. Although this factor would improve the agreement between the observed and modeled O 3 at HPO, recent estimates by Song et al. [2009] showed no increasing trends in NOx and VOC emissions from the open burning sources in China. [20] Anthropogenic NOx emissions estimates still have uncertainty. The increase of Chinese NOx emissions was found to be not quantitatively consistent with the satellitederived NO 2 column over China [Richter et al., 2005]. The growth rate of tropospheric NO 2 column was found to be larger than the estimated changes in NOx emissions since 2000, implying that the actual NOx emissions are likely larger than the emission estimates [Akimoto et al., 2006]. Zhang et al. [2007] reported that the growth in their NOx emissions inventory, which is basically the same as REAS, was as much as 30% lower than the satellite NO 2 column from 1996 to Stavrakou et al. [2008] derived topdown anthropogenic NOx emissions using satellite NO 2 measurements, proposing a +15% increase in the summertime surface O 3 over China for the period. This 4of5

5 is substantially larger than the modeled (9%) and closer to the observed (20%) at HPO. [21] Zhang et al. [2009] reported rapid growth of NOx emissions in Inner Mongolia due to newly constructed power plants. These sources were not implemented in REAS. Changes in locations of emission sources, particularly large point sources at high-altitude regions, can be effective in producing O 3 in the free troposphere. [22] Finer spatial resolution in the model improved convective transport of O 3 from the boundary layer to the free troposphere over continental source regions (M. Lin et al., Quantifying pollution inflow and outflow over East Asia with global and regional models, paper presented at TFMM - TF HTAP Joint Workshop: On Regional-Global and Air Quality-Climate Linkages, L Institut National de l Environnement Industriel et des Risques, La Plaine Saint Denis, Paris, France, 2009). Combined with better treatment of orographic feature, this improvement would likely enhance the long-range transport of pollution from Asian continental source regions to downwind. [23] Findings of this study point to limitations in our current CTMs to quantitatively predict tropospheric O 3 trends, particularly in response to changes in emissions over the continental Asia during the last decade. Modeling effort with improved emissions inventories and finer spatial resolution would be of great importance to better reconcile the difference between the observed tropospheric O 3 trends over western Pacific Ocean during the past decades and model simulations. [24] Acknowledgments. Ambient data used were provided by ADORC and WDCGG. We thank J. Kurokawa for help in model runs. This work was supported by the Global Environment Research Fund (S-7) by the Ministry of the Environment, Japan. References Acid Deposition and Oxidants Research Center (2001), Measurement and analysis, in Quality Assurance/Quality Control (QA/QC) Program for the Air Concentration Monitoring in East Asia, U.N. Environ. Programme, Niigata, Japan. Akimoto, H., T. Ohara, J. Kurokawa, and N. Horii (2006), Verification of energy consumption in China during by using satellite observational data, Atmos. Environ., 40, , doi: /j.atmosenv Byun, D. W., and K. L. Schere (2006), Review of the governing equations, computational algorithms, and other components of the Models-3 Community Multiscale Air Quality (CMAQ) modeling system, Appl. Mech. Rev., 59, 51 77, doi: / Derwent, R. G., P. G. Simmonds, A. J. Manning, and T. G. Spain (2007), Trends over a 20-year period from 1987 to 2007 in surface ozone at the atmospheric research station, Mace Head, Ireland, Atmos. Environ., 41, , doi: /j.atmosenv Fiore, A. M., D. J. Jacob, I. Bey, R. M. Yantosca, B. D. Field, A. C. Fusco, and J. G. Wilkinson (2002), Background ozone over the United States in summer: Origin, trend, and contribution to pollution episodes, J. Geophys. Res., 107(D15), 4275, doi: /2001jd Fu, T.-M., D. J. Jacob, P. I. Palmer, K. Chance, Y. X. Wang, B. Barletta, D. R. Blake, J. C. Stanton, and M. J. Pilling (2007), Space-based formaldehyde measurements as constraints on volatile organic compound emissions in East and South Asia and implications for ozone, J. Geophys. Res., 112, D06312, doi: /2006jd Jaffe, D., H. Price, A. Goldstein, and J. Harris (2003), Increasing background ozone during spring on the west coast of North America, Geophys. Res. Lett., 30(12), 1613, doi: /2003gl Japan Meteorological Agency (2005), Annual report on atmospheric and marine environment monitoring [CD-ROM], Rep. 7, Jpn.Meteorol. Agency, Tsukuba, Japan. Kajii, Y., K. Someno, H. Tanimoto, J. Hirokawa, H. Akimoto, T. Katsuno, and J. Kawara (1998), Evidence for the seasonal variation of photochemical activity of tropospheric ozone: Continuous observation of ozone and CO at Happo, Japan, Geophys. Res. Lett., 25, , doi: /98gl Lelieveld, J., J. van Aardenne, H. Fischer, M. de Reus, J. Williams, and P. Winkler (2004), Increasing ozone over the Atlantic Ocean, Science, 304, , doi: /science Network Center for EANET (2007), Acid Deposition Monitoring Network in East Asia (EANET), Dana Rep. 2006, pp , Network Cent. for EANET, Niigata, Japan. Ohara, T., H. Akimoto, J. Kurokawa, N. Horii, K. Yamaji, X. Yan, and T. Hayasaka (2007), An Asian emission inventory of anthropogenic emission sources for the period , Atmos. Chem. Phys., 7, Oltmans, S. J., et al. (2006), Long-term changes in tropospheric ozone, Atmos. Environ., 40, , doi: /j.atmosenv Ordóñez, C., D. Brunner, J. Staehelin, P. Hadjinicolaou, J. A. Pyle, M. Jonas, H. Wernli, and A. S. H. Prévôt (2007), Strong influence of lowermost stratospheric ozone on lower tropospheric background ozone changes over Europe, Geophys. Res. Lett., 34, L07805, doi: / 2006GL Parrish, D. D., D. B. Millet, and A. H. Goldstein (2009), Increasing ozone in marine boundary layer inflow at the west coasts of North America and Europe, Atmos. Chem. Phys., 9, Richter, A., P. Burrows, H. Nues, C. Grainer, and U. Niemeijer (2005), Increase in tropospheric nitrogen dioxide over China observed from space, Nature, 437, , doi: /nature Song, Y., B. Liu, W. Miao, D. Chang, and Y. Zhang (2009), Spatiotemporal variation in nonagricultural open fire emissions in China from 2000 to 2007, Global Biogeochem. Cycles, 23, GB2008, doi: /2008gb Staehelin, J., J. Thudium, R. Buehler, A. Volz-Thomas, and W. Graber (1994), Trends in surface ozone concentrations at Arosa (Switzerland), Atmos. Environ., 28, 75 87, doi: / (94) Stavrakou, T., J.-F. Müller, K. F. Boersma, I. De Smedt, and R. J. van der A (2008), Assessing the distribution and growth rates of NO x emission sources by inverting a 10-year record of NO 2 satellite columns, Geophys. Res. Lett., 35, L10801, doi: /2008gl Streets, D. G., K. F. Yarber, J.-H. Woo, and G. R. Carmichael (2003), Biomass burning in Asia: Annual and seasonal estimates and atmospheric emissions, Global Biogeochem. Cycles, 17(4), 1099, doi: / 2003GB Sudo, K., M. Takahashi, J. Kurokawa, and H. Akimoto (2002), CHASER: A global chemical model of the troposphere 1. Model description, J. Geophys. Res., 107(D17), 4339, doi: /2001jd Tanimoto, H. (2009), Increase in springtime tropospheric ozone at a mountainous site in Japan for the period , Atmos. Environ., 43, , doi: /j.atmosenv Tanimoto, H., Y. Sawa, H. Matsueda, I. Uno, T. Ohara, K. Yamaji, J. Kurokawa, and S. Yonemura (2005), Significant latitudinal gradient in the surface ozone spring maximum over East Asia, Geophys. Res. Lett., 32, L21805, doi: /2005gl Terao, Y., J. A. Logan, A. R. Douglass, and R. S. Stolarski (2008), Contribution of stratospheric ozone to the interannual variability of tropospheric ozone in the northern extratropics, J. Geophys. Res., 113, D18309, doi: /2008jd Uno, I., et al. (2007), Systematic analysis of interannual and seasonal variations of model-simulated tropospheric NO 2 in Asia and comparison with GOME-satellite data, Atmos. Chem. Phys., 7, Volz, A., and D. Kley (1988), Evaluation of the Montsouris series of ozone measurements made in the nineteenth century, Nature, 332, , doi: /332240a0. Wild, O., and H. Akimoto (2001), Intercontinental transport of ozone and its precursors in a three-dimensional global CTM, J. Geophys. Res., 106, 27,729 27,744, doi: /2000jd World Meteorological Organization (2007), WMO Global Atmosphere Watch (GAW) strategic plan: , WMO/GAW Rep. 172, 104 pp., World Meteorol. Organ., Geneva, Switzerland. Zbinden, R. M., J.-P. Cammas, V. Thouret, P. Nédélec, F. Karcher, and P. Simon (2006), Mid-latitude tropospheric ozone columns from the MOZAIC program: Climatology and interannual variability, Atmos. Chem. Phys., 6, Zhang, Q., et al. (2007), NO x emission trends for China, : The view from the ground and the view from space, J. Geophys. Res., 112, D22306, doi: /2007jd Zhang, Q., D. G. Streets, and K. He (2009), Satellite observations of recent power plant construction in Inner Mongolia, China, Geophys. Res. Lett., 36, L15809, doi: /2009gl T. Ohara and H. Tanimoto, Asian Environment Research Group, National Institute for Environmental Studies, 16-2 Onogawa, Tsukuba, Ibaraki , Japan. (tanimoto@nies.go.jp) I. Uno, Research Institute for Applied Mechanics, Kyushu University, Kasuga Park 6-1, Kasuga, Fukuoka , Japan. 5of5