al., 1996; Chatfield et al., 1998; Wotawa and Trainer, 2000; Forster et al., 2001; ... Bowman, 2001; Chatfield et al., 2002] and from the main industrial regions of ...
A Modeling Study of the Export Pathways of Pollution from Europe: Seasonal and Interannual Variations (1987-1997) B. N. Duncan and I. Bey Laboratoire de Modélisation de la Chimie Atmosphérique, École Polytechnique Fédérale de Lausanne (EPFL), Switzerland
Abstract. We present a study of the seasonal and interannual variations of the export pathways of pollution from Europe for eleven years, 1987-1997, using the GEOS-CHEM model of 3-d trace gas/aerosol chemistry and transport. The dominant export pathways in winter are advection to i. the middle and high latitudes of the North Atlantic Ocean, including the European Arctic; ii. Russia and the Russian Arctic; and iii. the middle and low latitudes of the North Atlantic Ocean from western Europe and from northern Africa via the Mediterranean Basin. In summer, export occurs by both advection and convection. Transport by advection occurs predominantly to Russia and the Mediterranean Basin/northern Africa. There are two major regions of convection in summer that loft European pollution into the free troposphere, one centered over Germany and the other over the Ural Mountains in Russia. Another region of lofting, not associated with moist convection, occurs in northwestern Africa by the quasi-permanent Western Saharan Low. Summertime ozone in our model is enhanced by European pollution (~5 ppbv on average) in the middle troposphere near these three regions of lofting. In addition, European pollution causes summertime, surface ozone levels in northern Africa and the Near East, regions with a total population of about 200 million people, to exceed regularly the European Council’s human health standard. European pollution does not contribute as significantly to ozone levels in other populated regions outside of Europe. The two dominant causes of interannual variation in the export of European pollution over our study period are variations in transport, especially associated with the North Atlantic Oscillation (NAO), and changes in anthropogenic emissions. The tropospheric burden of carbon monoxide from European sources varies by as much as ±20% over both western Europe and the North Atlantic Ocean and ±15% over the Arctic during the 11year study period in winter because of interannual variations in transport. When the NAO is in the positive (negative) phase, the carbon monoxide burden from European sources tends to be lower (higher) over the North Atlantic Ocean and higher (lower) over the Arctic. Emissions of carbon monoxide and nitrogen oxides decrease by 36% and 20%, respectively, in Europe during 1
our study period, with a nearly linear, concomitant decrease in the carbon monoxide burden associated with European emissions. We find that it is necessary to consider both of these factors when interpreting trends in observed trace gas concentrations over Europe. Our work should be seen as an exploratory study that could help in the design of measurement networks and field campaigns dedicated to the sampling of European pollution, as only sparse observations are available in these regions affected by it.
1. Introduction Trace gases and aerosols (e.g., ozone (O3) and it precursors, carbon monoxide (CO), persistent organic pollutants (POPs), mercury, desert dust) can undergo long-range transport (LRT) and impact the tropospheric composition thousands of kilometers downwind from their source regions. There is clear evidence of LRT from regions experiencing biomass burning [e.g., Hsu et al., 1996; Chatfield et al., 1998; Wotawa and Trainer, 2000; Forster et al., 2001; Rogers and Bowman, 2001; Chatfield et al., 2002] and from the main industrial regions of North America, Europe, and East Asia [e.g., Jaffe et al., 1997; Stohl and Trickl, 1999; Pochanart et al., 2003; Stohl et al., 2003; Trickl et al., 2003]. Recently, researchers have focused on the intercontinental transport of O3 and its precursors between North America, Europe and East Asia [Berntsen et al., 1999; Jacob et al., 1999; Jaffe et al., 1999; Collins et al., 2000; Fiore et al., 2002a; Prather et al., 2003; Stohl et al., 2003; Trickl et al., 2003]. However, there are still major uncertainties of how LRT i. affects the global budget of tropospheric O3 (the third most important greenhouse gas after carbon dioxide and methane [IPCC, 2001]) and ii. influences regional photochemical episodes. Li et al. [2002a], for example, estimated that 20% of the violations of the European Council O3 standard would not have occurred in the summer of 1997 without North American anthropogenic emissions. Collins et al. [2000] estimated that for summertime O3 over Europe to remain relatively constant in the next few decades, about a 50% decrease in European emissions of nitrogen oxides (NOx) will be necessary to offset the expected increase in the Northern Hemisphere background O3. Prather et al. [2003] reported the predictions of several models on future background O3 through the twenty-first century. They estimated that background O3 near the surface in the Northern Hemisphere will be about 5 ppbv higher by 2030 and possibly greater than 20 ppbv by 2100. There were several aircraft campaigns that explored the continental outflow of North America (e.g., NARE [Fehsenfeld et al., 1996]) and Asia (e.g. PEM-West B [Hoell et al., 1997], TRACEP [Jacob et al., 2003]). The observations of trace gases from these experiments provide a 2
valuable snapshot of the continental outflow’s composition, its transport pathways, and help to constrain global model simulations of pollution export [e.g., Bey et al., 2001a; Liu et al., 2003]. Surprisingly, there have been few studies, either experimental or model-based, devoted to the export of European pollution. Previous model studies and observations indicate that the export of European pollution occurs to Russia, the Arctic, East Asia, and Africa. Stohl et al. [2002] found that export occurs predominantly from Europe to the northeast over Russia year-round, though not as strongly in summer, and accumulates in the Arctic Circle, especially in winter. European pollution was identified in the 1980s as the main contributor to thick layers of haze seen throughout the Arctic, especially in winter when the air over the Arctic is more stagnant than in summer [e.g., Schnell, 1984; Iversen, 1984; Shipman et al., 1992; Lamarque and Hess, 2003 and references therein]. The Arctic Haze phenomenon diminishes in summer due to more frequent pollutant scavenging of the Arctic troposphere by summer rains, changes in the largescale circulation, and less air stagnation in the Arctic [Shipman et al., 1992; Klonecki et al., 2003; Lamarque and Hess, 2003]. Wild and Akimoto [2001] used a chemical tracer model to also find that European O3 and its precursors tend to accumulate in the Arctic. In addition, they found that in North America, O3 produced from European pollution contributes nearly equally as O3 produced from Asian pollution. They attributed this to the transport of European pollution via the Arctic, which is a shorter pathway than mid-latitude transport around the globe. Some European pollution that is exported to Russia travels around the northern edge of the Siberian High to the industrial regions of East Asia. Model studies have shown that CO from Europe is an important component of total CO exported from East Asia in winter and spring [Bey et al., 2001a; Staudt et al., 2001]. Newell and Evans [2000] performed back-trajectory analyses to estimate that up to 40% of the air arriving in East Asia in winter may be polluted by European sources, with a minimum contribution in summer. Stohl et al. [2002] described the southward flow of European pollution in summer over the Mediterranean Basin and into Africa. Lelieveld et al. [2002] reported that 60-80% of CO in the boundary layer over the Mediterranean in summer is attributable to European sources and that the 8-hour mean O3 standard of 55 ppbv is exceeded regularly there because of European pollution. Hamelin et al. [1989] deduced from lead isotopes measurements from Barbados in the western Atlantic that isotopes from Europe were exported to Africa and then transported across the Atlantic Ocean to Barbados in the trade winds. The export of Europe pollution mainly occurs at low altitudes at all times of the year and convection becomes a minor, but non-negligible, export pathway in summer [Wild and Akimoto, 3
2001; Stohl et al., 2002; Traub et al., 2003]. Only a few episodes of vertical lifting of pollution over Europe have been reported in the literature. Using measurements of nonmethane hydrocarbons performed during the European Export of Precursors and Ozone by Long-Range Transport (EXPORT) experiment in August 2000, Purvis et al [2003] observed rapid lifting of pollution due to convection embedded within a cold front. Hov and Flatøy [1997] estimated from model results over two 10-day periods in summer that between 7 and 12% of NOx emissions were lofted by convection from the boundary layer as NOx and 13-15% as peroxyacetylnitrate (PAN) and nitric acid (HNO3). Fischer et al. [2003] reported a deep convective event over the Mediterranean in which pollution was injected into the stratosphere. The Warm Conveyor Belt (WCB) is an important mechanism that lifts pollution from the surface to upper troposphere, especially in East Asia and eastern North America [Bethan et al., 1998; Stohl and Trickl, 1999]. Stohl [2001], however, pointed out that WCBs rarely originate near Europe and are not a mechanism to lift European emissions. The present study is intended to give a comprehensive view of the export of pollution from Europe, using the GEOS-CHEM 3-D model of transport and chemistry in the troposphere [Bey et al., 2001b]. We identify the various export pathways in different seasons and examine the factors that contribute to the interannual variability of export from 1987 to 1997. We use CO as a tracer of anthropogenic pollution because it has a long lifetime relative to the time-scales of transport and can be easily associated to a given geographical source region in the model. As it will be shown in the following sections, only sparse observations are available in regions downwind of Europe. Thus, our work should be seen as an exploratory study that could help for the design of field campaigns dedicated to the sampling of exported European pollution. We describe the GEOS-CHEM model in Section 2 and evaluate it with observations of CO where available. We discuss export pathways in fall, winter, and spring in Section 3, as well as their year-to-year variations. We find that there are two principal causes of the interannual variations in the export of pollution: i. changes in meteorology primarily associated with the North Atlantic Oscillation (NAO); and ii. changes in European fossil fuel emissions. In Section 4, we discuss the interannual variations of the summertime export of European pollution and assess the impact of the transport of European pollution on summertime O3 in northern Africa and the Near East. We present a summary of conclusions in Section 5.
2. GEOS-CHEM Model Description 4
We use the GEOS-CHEM 3-d model of transport and trace gas/aerosol chemistry, first described by Bey et al. [2001b]. We employ version 5.04 of the model (http://wwwas.harvard.edu/chemistry/trop/geos/index.html). The model captures interannual variations in transport as it is driven by assimilated meteorological data from the Goddard Earth Observing System Data Assimilation Office (GEOS DAO) [Schubert et al., 1993]. Two versions of GEOS assimilated meteorological data are required for our study period: GEOS-1 (January, 1987 November, 1995) and GEOS-STRAT (December, 1995 - December, 1997). The meteorological fields are provided on a sigma coordinate system with 20 vertical levels for GEOS-1 to 10 hPa, and 46 levels for GEOS-STRAT to 0.1 hPa. For computational expedience, we degrade the vertical resolution of the GEOS-STRAT fields by merging the vertical levels above the lower stratosphere, retaining a total of 26 levels. The sigma vertical levels of the model are given in Bey et al. [2001b]. All meteorological data are provided as 2º latitude by 2.5º longitude horizontal resolution. The model’s emissions from fossil fuel combustion vary annually to capture important changes during our study period. We use the same 1985 base emissions inventory for CO and NOx as Wang et al. [1998]. Emissions for individual countries are scaled from the 1985 values [Bey et al., 2001b] by annual emissions estimates provided by the European Monitoring and Evaluation Program (EMEP) [EMEP, 1997] for European countries. Table 1 presents the annual emissions from Europe for 1987 and 1997. Between these years, total European emissions of CO and NOx decrease by 36% and 20%, respectively. The emission changes are different for eastern and western Europe. Eastern Europe underwent an economic contraction during the late 1980s and 1990s, especially in the former Soviet Union, decreasing CO and NOx emissions by 43% and 24%, respectively. Emissions of CO and NOx in western Europe decrease by 25% and 17%, respectively, mostly through regulatory controls. Figure 1 shows the annual emission rates of CO and NOx in 1987. The highest emission rates are generally in western Europe, but there are relatively high emission rates over much of the continent. Figure 1 shows also the ratio of the annual emission rates in 1997 to 1987 for CO and NOx. The emission rates decline most everywhere, except the Iberian Peninsula and Poland. 2.1. Model Simulations We first performed an 11-year period simulation from January 1987 to December 1997 with a 4º latitude x5º longitude horizontal resolution. This simulation carries 24 trace gases and includes a comprehensive description of O3-NOx-NMHC chemistry, as well as the radiative and 5
heterogeneous effect of aerosols. We do not show the results of this simulation in this manuscript. We use the monthly mean hydroxyl radical (OH) fields archived from this first experiment to conduct a 2ºx2.5º horizontal resolution, CO-only simulation, first described in Bey et al. [2001b] and Staudt et al. [2001], to which we refer to as SIM-TAGCO. Eighteen CO tracers are transported, tagged by their source region, to resolve the contribution of various sources to total CO over the whole study period. For example, “European” CO refers to CO emitted from fossil fuel combustion in Europe, which we use to identify the export pathways of pollution. The OH fields are used to calculate the loss of CO and production of CO by methane oxidation. CO production from the oxidation of nonmethane hydrocarbons (NMHC) emitted from fossil fuel combustion is accounted for by increasing the direct CO emissions by 20% [Duncan et al., 2003b]. CO concentrations from the full-chemistry run are reproduced closely in the CO-only run when CO from all tracers is summed. We conducted two other CO-only simulations (SIM-FF87 and SIM-MET87) to separate the impact of changing emissions from the impact of variations of transport on export from Europe. The SIM-FF87 simulation is the same as the SIM-TAGCO simulation, except that it uses fixed annual fossil fuel emissions to values for 1987 for the entire 11-year period. The SIM-MET87 simulation is also the same as the SIM-TAGCO simulation, but uses the meteorological fields of 1987 for all eleven years instead of the fields specific to each year. Finally, we performed two additional experiments (SIM-SUM1 and SIM-SUM2) to understand the impact of European emissions on summertime O3. SIM-SUM1 is a standard O3NOx-NMHC chemistry simulation as previously described for May to July 1994, but uses a 2ºx2.5º horizontal resolution. SIM-SUM2 is the same as SIM-SUM1, except that we ‘turn off’ European fossil fuel emissions of trace gases, from 10ºW to 60ºE and 35º-70ºN. In the same region, black and organic carbon aerosols are zeroed. Organic carbon aerosols are predominantly of anthropogenic origin, although terpene oxidation is a relatively minor source over Europe. 2.2. Model Evaluation Extensive evaluations of the model’s output and observations of CO, O3, NOx, etc. are presented in a number of papers, which give confidence in the ability of the model to reproduce the general features of tropospheric chemistry [Bey et al., 2001ab; Li et al., 2001; Palmer et al., 2001; Chandra et al., 2002; Fiore et al., 2002ab; Li et al., 2002ab; Liu et al., 2002; Martin et al., 2002ab; Duncan et al., 2003a; Martin et al., 2003; Palmer et al., 2003]. Auvray et al. [2003] 6
presented an evaluation of model O3 from an O3-NOx-NMHC simulation over Europe, using observations from the EMEP network [EMEP, 2002], vertical profiles from ozonesondes [Logan, 1999], and the Measurement of Ozone and Water Vapor by Airbus In-Service Aircraft (MOZAIC) program [Marenco et al., 1998]. They found that the model reasonably reproduces O3 over Europe in the whole troposphere with the exception that O3 is overestimated at some surface sites in summer. In this manuscript, we focus on a model evaluation using total column and surface CO observed in and downwind of Europe for 1987 to 1997. The geographical location of each measurement station is given in Table 2. We evaluate the model’s performance using the Pearson linear correlation coefficient, R, with the number of monthly-averaged observations, N, common to the time period of the model simulation [Press et al., 1989] (Table 2). It is important to note that part of the correlation is driven simply by the model’s ability to capture the seasonal cycle of observed CO. Nevertheless, we believe R2 gives some quantitative indication of model performance. We estimate also the mean bias of the model at each measurement station, where a positive bias means that the model tends to over-predict the observed concentrations. Unfortunately, there are few long-term observations with which to compare our model during our study period in the regions impacted by European pollution. Figure 2 shows observed and modeled (SIM-TAGCO) column CO at Jungfraujoch, Switzerland [Mahieu et al., 1997] and Zvenigorod, Russia [Yurganov et al., 2002]. There are about five observations each month to create the monthly average at Jungfraujoch and about three in winter and eight in summer at Zvenigorod, though there are no observations for some months. The model reproduces reasonably well the seasonality and year-to-year variations of CO at these sites. The R2’s between model and observations at Jungfraujoch and Zvenigorod are 0.52 and 0.43, respectively. There is no bias at Jungfraujoch, but there is a negative, timedependent bias at Zvenigorod. The model tends to under-predict the CO column at Zvenigorod in the latter half of our study period. Although, Yurganov et al. [2002] reported that systematic and slowly-varying errors may be as high as ± 5-6% in the observations, this under-prediction could also indicate that the downward trend in the model’s European CO emissions is overpredicted. Surface CO is measured routinely around the world as part of the cooperative National Oceanic and Atmospheric Administration (NOAA)/Climate Monitoring and Diagnostics Laboratory (CMDL) flask sampling program [Novelli et al., 1992]. Figure 3 shows monthly7
mean observed and model CO from 1987 to 1997 at eleven NOAA/CMDL sites in and near Europe. The monthly-mean, observed data include observations flagged by NOAA/CMDL as being non-background because we do not screen for non-background air in our model. The model does a reasonable job of capturing the variations of observed CO, including the seasonal minima and maxima. The R2’s at all the stations, except at Sede Boker, Israel, are between 0.60 and 0.75 (Table 2). However, most of the stations in Europe have statistically significant positive mean biases (5.9-19.4%), except at Hegyhatsal, Hungary which has a statistically significant negative mean bias (i.e., -8.6%), and Constanta, Romania and Terceira Island, Azores, which have no mean biases. The positive mean biases at the higher latitude stations are likely associated with the way we account for the production of CO from the oxidation of anthropogenic NMHC in the CO-only simulation as discussed in Section 2.1, which causes total CO to be over-predicted near the fossil fuel source regions, especially in winter when the lifetimes of NMHCs are long. The model reproduces much of the interannual variations of CO at the NOAA/CMDL sites. For example, Mace Head, located on the west coast of Ireland, is subjected alternately to relatively clean air from the Atlantic Ocean and polluted air from the European continent [Simmonds et al., 1997; Cape et al., 2000]. The strong interannual variations of the observations at this site are associated with interannual variations in transport [Allen et al., 1996]. However, some individual high months, such as February 1994 at Hegyhatsal, Hungary, are not captured by the model or are under-predicted as at Mace Head. Part of the discrepancy between the model and observations may be due to the fact that the monthly-average model CO is the average of the entire month, while the number of observations that were used to create the monthly-averaged values typically range anywhere from 3 to 20. We find much better agreement between model CO and observations in adjacent model boxes for three sites: Baltic Sea (BAL), Hegyhatsal, Hungary (HUN), and Sede Boker, Israel (WIS). The statistics shown in Table 2 and the model data plotted in Figure 3 for these three stations represent the model CO in these adjacent boxes. The Baltic Sea site is located in a model box comprised of sea and land with high continental emissions, characteristic of north-central Europe, while the measurements were taken at sea. Comparisons for this site are made, therefore with the model box to the northeast, which lies mostly over the sea. The R2 improves from 0.44 to 0.61 and the mean bias from 30% to 19%. The R2 improves at Hegyhatsal from 0.53 to 0.70 when we sample the model box to the southwest, a box with lower emissions more representative of regional concentrations than the original box that includes Budapest, but the mean bias 8
worsens from 4.3% to –8.6%. The better mean bias in the original box, however results simply from a balance between a large model under-prediction of observed CO in February 1994 by about 200 pbbv and two model over-predictions in the winters of 1995 and 1996 by about 100 ppbv each (not shown). Excluding the one observed outlier in February 1994, the R2 and mean bias become 0.76 and –7.9%, respectively. We find that the R2 and mean bias are better in the model box to the northeast of the box including the site at Sede Boker, Israel. The R2 and mean bias improve from 0.24 to 0.40 and 8.4% to 2.7%, respectively. Despite this improvement, the statistics show that the model does a poor job of reproducing observations at this site. It is not clear why the model comparison with observations in the box to the northeast is better, but it is comprised of all land, while the original box is part land and sea.
3. Winter The main features of wintertime circulation in the Northern Hemisphere are seen in Figure 4a, which shows the January-average (1987-1997) sea-level pressures (SLPs) and winds at 800 mb, respectively. On average, the pressure gradient between the Icelandic Low and the Azores High, both located in the North Atlantic Ocean, drives the flow over Europe. The average position of the Icelandic Low stretches from about southern Greenland to the Barents Sea and the Azores High is part of an area of high pressure ridge that stretches across southern Europe to the Siberian High in Asia. The SLP standard deviation from 1987 to 1997 in Figure 4b reveals that there is a high degree of year-to-year variation over the North Atlantic Ocean and Europe (i.e., 5-10 mb), especially between Iceland and Norway (i.e., 10-15 mb), the highest variation in the Northern Hemisphere. Much of this variability is associated with the climate phenomenon, the North Atlantic Oscillation (NAO), which accounts for more than one-third of the total variance in SLPs in winter [Hurrell et al., 2003]. In the extra-tropics of the Northern Hemisphere, the NAO is one of the dominant influences on climate variability from eastern North America to Siberia from fall to spring, but especially during winter [Hurrell et al., 2002], explaining about 30% of the variance in mean surface temperature in the Northern Hemisphere during winter [Hurrell, 1996]. As we will show, the NAO strongly influences the strengths and locations of the export pathways from Europe. In this section, we also discuss briefly the export pathways in spring and fall. The Azores High grows and expands through spring while the Icelandic Low weakens and is displaced to the 9
west of Iceland, and vice-versa in fall. As a result, the pressure gradients between the Icelandic Low and Azores High are generally less in spring and fall than in winter, although the export pathways are similar in the three seasons. 3.1. North Atlantic Oscillation The NAO is described as a latitudinal redistribution of atmospheric mass over the North Atlantic Ocean. Walker [1924] and Walker and Bliss [1932] developed the concept of the NAO Index (NAOI) to describe the phase of the NAO at any given time. We calculate the NAOI from the average SLPs at Ponta Delgada, Azores and Stykkisholmur/Reykjavik, Iceland, though there are a number of other stations that can be used to calculate the NAOI [Jones et al., 2003]. SLP data were taken from the website of the Climatic Research Unit, University of East Anglia (http://www.met.rdg.ac.uk/cag/NAO/index.html). We follow the expanded definition of Hurrell et al. [1995]: NAOI = (SLPA – SLPAmean)/SDA – (SLPI-SLPImean)/SDI
(1)
where SLPA and SLPI are monthly-average SLPs for a given year at the Azores and Iceland, respectively; SLPAmean and SLPImean are the monthly-average SLPs from 1864-1997 at the Azores and Iceland, respectively; and SDA and SDI are the standard deviations of the SLPs at the Azores and Iceland, respectively, from 1864-1997. Figure 5 shows the observed and model NAOI for January from 1987 to 1997. There is a good agreement between the model and observed NAOI, which is to be expected since the model uses assimilated meteorological fields. The January-average NAOI is negative in 1987, 1992, and 1996-97, positive in 1988-1991, and 1993-1995. In winter during our study period, the observed NAOIs for the months of December through March are positive in 22 months, neutral in 12 (i.e., –1 < NAOI < +1), and negative in 9. The NAOI can be highly variable within a season [Hurrell et al., 2003]. Figure 4c-f shows the January-average SLPs and winds at 800 mb in 1994 (NAO+), 1995 (NAO+), 1996 (NAO–), and 1997 (NAO–). When the NAO is positive (i.e., NAOI > +1), the pressure of the Azores High is higher and the pressure of the Icelandic Low is lower than when the NAO is negative (i.e., NAOI < –1). The Icelandic Lows and Azores Highs are relatively strong in January 1994 and 1995 (NAO+) and, subsequently, there are strong westerlies over Europe. Similar meteorological situations are seen in other Januarys of our study period when the NAOIs are positive, 1988-1991 and 1993 (Figure 5). The Icelandic Low and Azores High 10
are relatively weak in 1997 (NAO–) as compared to the other three years in Figure 4c-e. The center of the Icelandic Low is shifted south of Greenland and a ridge of high pressure associated with the Siberian High reaches into western Europe. Consequently, the winds are weak over most of Europe. A second area of low pressure lies in the European Arctic. The meteorological situations of January 1992 (NAO–) and 1987 (NAO–) (not shown) are similar to January 1997, except that the second area of low pressure over the Barents Sea is not strong in 1987. The pressure gradient between the Icelandic Low and the ridge of high pressure over Europe results in southerly flow in the North Atlantic. This flow moves into the Arctic Circle where it falls under the influence of the pressure gradient between the high pressure ridge over Europe and the second area of low pressure over the Barents Sea. Consequently, there is northwesterly flow over northern Europe, westerly flow over eastern Europe, and weak winds over western Europe (Figure 4f). Unfortunately, the station-based NAOI presented here is fixed in space and does not give all the necessary information on the spatial structure and intensity of the NAO [Hurrell et al., 2003; Jones et al., 2003]. For example, the NAOI in January 1996 is negative, which, according to the definition of the negative phase of the NAO, should be characterized by a relatively weak Icelandic Low and Azores High as in January 1997. This is not the case as both pressure centers are relatively strong (Figure 4e). The negative NAOI for January 1996 occurs because of the unusual position of the Icelandic Low with its center of pressure lying approximately between the two stations. That is, the NAOI, in this case, is not a measure of the strength of the pressure gradient between the Icelandic Low and Azores High. Nevertheless, the flow shares features common to other Januarys in our study period when the NAOIs are negative. For instance, there is southerly flow between the Icelandic Low and high pressure ridge over Europe, but the flow is shifted over Europe instead of the North Atlantic because the position of the Icelandic Low is shifted from south of Greenland to south of Iceland. 3.2. Export Pathways We used the NASA Langley trajectory model [Pierce and Fairlie, 1993; Pierce et al., 1994] to compute five-day isentropic forward-trajectories to identify the major export pathways of pollution from Europe, using the model’s 2° latitude by 2.5° longitude horizontal resolution wind fields described in Section 2. One trajectory was initialized for each day beginning at 00 UTC at potential temperatures of 280 K in January 1995 and 1996, 285 K in April and October 1995, and 295 K in July 1995, which represent the lower troposphere. Figures 6-9 shows 5-day 11
forward trajectories from Moscow, Russia, Madrid, Spain, Brussels, Belgium, and Milan, Italy, respectively, for January, April, July, and October 1995, and January 1996. The trajectories remain below 600 mb or about 4 km (not shown). The trajectories indicate that the strengths of the export pathways vary with the phase of the NAO. Three major export pathways from Europe are seen when the phase of the NAO is positive, as, for example, in January 1995 (NAO+) (Figures 6-9). The first pathway is controlled by the prevailing westerlies, which dominate, for example, the flow from Moscow in eastern Europe. Stohl et al. [2002] identified this pathway as the dominant export pathway from Europe in winter. Several trajectories indicate flow from Europe to the northeast around the northern edge of the Siberian High to the Russian Arctic and then to the southeast over China. Export in the westerlies is evident also from Brussels in western Europe and Milan in southern Europe. Many trajectories from Moscow, Brussels, and Milan cross into the Arctic Circle. The second pathway is northward and northwestward around the Icelandic Low into the Atlantic Ocean and is evident in the trajectories from Madrid and Brussels in western Europe. This pathway directly transports western European pollution to the European Arctic. The third pathway is southward flow around the Azores High into the Mediterranean Basin, northern Africa, and the Atlantic Ocean from western Europe and is seen from Madrid, Brussels, and Milan. The second and third pathways are routes for European pollution to cross the Atlantic Ocean to North America in the lower troposphere [Li et al., 2002a]. In January 1996, the phase of the NAO is negative, though, as discussed in Section 3.1, the meteorological flow for the month is not a typical example of the phase. The trajectories from Moscow show that westerly flow around the northern edge of the Siberian High is the predominant export pathway from eastern Europe as in January 1995 (Figure 6). However, the major export pathway from western Europe is to the north and northwest in the strong flow between the Icelandic Low and western extension of the Siberian High to the Arctic and the North Atlantic Ocean (Figures 7-9). The export to the Mediterranean Basin and Africa seen in 1995 is not evident in 1996. 3.3. Long-Range Transport We used “European” CO (i.e., CO emitted from fossil fuel combustion in Europe) to identify the major long-range transport pathways of pollution from Europe, as a 5-day trajectory indicates only the initial transport pathway. We find in the GEOS-CHEM model that the export of pollution from Europe is limited to low altitudes in winter during our study period, which is in 12
agreement with the findings of Wild and Akimoto [2001] and Stohl et al. [2002]. In our model, there is a strong gradient in “European” CO from the boundary layer to an altitude of four kilometers in winter. The concentration of “European” CO, in general, is less than 50 ppbv at three kilometers and 20 ppbv at six kilometers above Europe, while the surface concentration is about 300 ppbv and higher. In fact, its concentration is sometimes higher in the East Asian boundary layer, greater than 70 ppbv, than above the European boundary layer. Figure 10 shows the monthly-average, tropospheric column of “European” CO from the SIMTAGCO run in January of 1995 (NAO+), 1996 (NAO–), and 1997 (NAO–). The maximum CO columns (>0.6 x1018 molec cm-2) are over the major pollution source regions in western Europe (Figure 1) in 1996 and 1997 when the winds are relatively weak over Europe (Figure 4ef). However, in 1995, the maximum column is displaced to the east of the main emission region by the strong westerlies (Figure 4d). The total CO at Mace Head is lower in both the observations and model in the winters of 1995 and 1997 than 1996 (Figure 3). The increased model CO in 1996 is due to European sources. Observations at other stations in the North Atlantic (i.e., Svalbard, Ocean Station M, and Storhofdi, Iceland) do not show evidence of enhanced CO in 1996 relative to 1995, but the model does at Ocean Station M and possibly at Storhofdi, Iceland. Consistent with the trajectory analyses in Figures 7-9, the Mediterranean Basin is more polluted in 1995 than 1996 and the North Atlantic Ocean is more polluted in 1996 than 1995. A common feature to all three years is accumulation of CO at latitudes greater than 70°N, however the European Arctic is more polluted in 1996 than in the other two years because of the southerly flow over western Europe (Figure 5). In January 1997, the CO column indicates that European pollution from western and eastern Europe enters the Arctic through Russia, while some of the pollution from western Europe enters the Arctic through Scandinavia and the North Atlantic in 1995 and 1996 according to the trajectories in Figures 7-9. 3.4. Impacts of Change in Fossil Fuel Emissions and Transport Variations In this section, we use the SIM-TAGCO, SIM-FF87, and SIM-MET87 runs to understand the relative contributions of the year-to-year changes in anthropogenic emissions and transport to variations in the regional tropospheric burdens and surface concentrations of “European” CO. The global burden of “European” CO decreases by about 36% from 39 Tg in January 1987 to 25 Tg in January 1997, approximately the same percent decrease as in anthropogenic European emissions. Declines in total model CO are seen at several of the NOAA/CMDL stations, such as 13
Svalbard, Storhofdi, Iceland, and Constanta, Romania (Figure 3). The decreases in observed CO at these stations are not obvious as none of the stations have records that are longer than about six years. However, there is a decrease in observed CO at stations with longer records that are also part of the NOAA/CMDL network (not shown) in the mid/high-latitudes of the Northern Hemisphere in the 1990s, which is attributed to the reduction in European emissions [Novelli et al., 2003]. Figure 11 shows the monthly-average, tropospheric burdens of “European” CO for four regions in the three CO-only simulations from 1987 to 1997. The regions are: western Europe (38°-70°N; 10°W-20°E), eastern Europe (38°-70°N; 20-60°E), North Atlantic (38°-70°N; 45°10°W), and Arctic (70°-90°N; 180°W-180°E). The impact of emissions changes on the burdens is apparent, especially in winter, for all four regions, by comparing the burdens in the SIMTAGCO and SIM-FF87 runs. The winter burden in the Arctic region is nearly equal to the combined burdens of the eastern and western Europe regions, illustrating that European pollution accumulates in the Arctic each winter. The impact of transport variations, including those variations due to the NAO, are clear in the western and eastern Europe regions (i.e., the source regions) by comparing, for instance, the burdens in the SIM-TAGCO and SIM-MET87 runs in the winters of 1989 and 1990. However, variations in transport have less impact on the burdens in the North Atlantic and Arctic regions, indicating that pollution from Europe accumulates in these regions despite the variations in transport. We find that the NAOI, especially if it is greater than +1 or less than –1 (i.e., clearly in the positive or negative phase), can be used as a general indicator of the predominant export pathways of European pollution in our model in fall, winter and spring. Figure 12 shows the anomalies of the monthly-mean, tropospheric “European” CO burden for the same four regions as in Figure 11 versus the NAOI in winter (i.e., December-March, 1987-1997; black dots), in spring (i.e., April-May; gray dots), and in fall (i.e., October-November; gray dots). We use the SIM-FF87 run for our calculations as to remove the influence of the substantial changes in fossil fuel emission rates. The wintertime anomalies for western Europe (R2 = 0.42) and the North Atlantic (R2 = 0.19) tend to be positive (negative) when the NAOIs are negative (positive). One can see, however, that the correlation is better when we do not take into account the neutral wintertime NAOIs (i.e., between –1 and +1). The same is true for the Arctic (R2 = 0.14), except the correlation is positive. There is no obvious correlation for eastern Europe (R2 = 0.08) in winter even when the NAOIs are less (greater) than –1 (+1). As seen in Figure 4b, the standard 14
deviation of the SLPs during our study period is smaller over eastern Europe than western Europe, indicating that the NAO has less impact on transport in eastern Europe. The export pathways for spring and fall are similar to those in winter, although the correlation between the variation in “European” CO burden and the NAOI is not as strong, in part because about twice as many NAOIs in spring and fall lie between +1 and –1 than in winter. This result is not surprising as the flow in spring and fall is dominated by the transition between the summer and winter flow patterns, causing the pressure gradients between the Icelandic Low and Azores High to be typically lower in spring and fall than winter. Our findings suggest that the year-to-year variations in transport need to be considered when one calculates a trend from observations of tropospheric pollutants over Europe. Figure 13 shows the January-average trend in the model “European” CO (ppbv/year) at the surface from 1987 to 1997 in the SIM-TAGCO and SIM-MET87 runs. The trend is the fit of the data in each model grid box to the linear model, y = A + Bx, by minimizing the Chi-square error statistic. The trend in SIM-TAGCO, the standard run, is negative over much of Europe (Figure 13, top panel), but is positive over western Europe and Russia despite the substantial reduction in emissions that occurred there (Table 1; Figure 1). The trend in SIM-MET87, the run with meteorology fixed to 1987 values, is negative everywhere, indicating that interannual variations in transport mask the effect of the considerable change in CO emissions on tropospheric concentrations (Figure 13, bottom panel). As annual-average surface concentrations over Europe are weighted more to the higher wintertime concentrations, the trend will be influenced by the variations in the phase of the NAO.
4. Summer The Azores High grows and expands through spring while the Icelandic Low weakens and is displaced to the west of Iceland, and by summer, the high dominates much of the flow over Europe [Wallén, 1970]. Figure 14a shows the 1987-1997 July-average SLPs and winds at 800 mb. On average, the Azores High is part of an area of high pressure that extends over western Europe where the relative strength of the flow is weak. The interannual variations in pressure during our study period are low, especially for western and southern Europe, as seen in the standard deviation of the SLPs from 1987 to 1997 (Figure14b). This implies that advective export pathways of pollution vary less from year-to-year than in winter. Figure 14c-e show the
15
July-average SLPs and winds at 800 mb in 1987, 1995, and 1996. The Azores High is particularly strong in July 1996 relative to 1987 and 1995. 4.1. Advection Figures 15, 6-9 and 16 show 5-day forward trajectories in July of 1987, 1995 and 1996, respectively, from the same four cities as discussed in Section 3.2. We initialized parcels at a potential temperature of 295 K, which typically corresponds to the boundary layer in summer. The trajectories indicate that the predominant flow from Moscow in July 1987 and 1996 is to the west as in winter, but there is also southward flow toward the Mediterranean basin in 1987 and 1995. This result is similar to the findings of Stohl et al. [2002]. The southward flow occurs also from Milan, Brussels, and Madrid in 1987 and 1996, but not in 1995 when strong high pressure is in control over Europe resulting in weak winds in the lower troposphere (Figure 14ce). Some of the pollution on the Iberian Peninsula moves directly off the coast to the Atlantic Ocean in the flow around the relatively strong Azores High in 1996. The short trajectories in 1987 and 1995 from Milan and Madrid indicate stagnant conditions which are typical of summer in those regions [Wallén, 1970]. Figure 17 shows the “European” CO column in July for 1995 and 1996. The column in summer is about half of the column in winter (Figure 10) as the sink of CO by OH reaches a maximum in summer. The column’s spatial distributions are similar for the two years. The maximum column is near the source regions and LRT occurs to Siberia and East Asia, to the North Atlantic Ocean, and northern Africa. The spatial distribution, however, of the CO column varies little from year-to-year in summer during our study period when the decrease in anthropogenic emissions is removed (i.e., SIM-FF87 run), except over the North Atlantic Ocean and Scandinavia where it can vary by a factor of two due to variations in the strength and position of the Azores High. 4.2. Vertical Lifting In summer, studies have shown that convection is a minor, but non-negligible export pathway over Europe [Wild and Akimoto, 2001; Stohl et al., 2002; Traub et al., 2003]. Purvis et al [2003] reported, for example, an episode of export of hydrocarbons from the boundary layer during the EXPORT campaign in August 2000. They suggested that convection embedded within a frontal system is the dominant mechanism for pollution to be lifted to the free troposphere rather than slantwise ascent. The trajectories in July of 1987, 1995, and 1996 show little vertical evolution 16
(not shown) remaining generally at pressures of 800 hPa or greater. However, the trajectory model does not represent sub-grid-scale convective transport or the effect of latent heating on vertical (diabatic) motion. We find in our model that convection is an important pathway for the export of European pollution to the middle and upper troposphere in summer. Figure 18 shows the monthly distribution of total precipitation in the model, which we use as an indicator of the locations of convection for July in 1987 and 1994. There are two areas of maximum precipitation (i.e., 75100 mm) centered over Germany and western Russia in 1987, though relatively high precipitation rates (i.e., >50 mm) occur from France to Russia. A similar distribution of rainfall occurs in 1994 though not as high over Germany as in 1987. Vertical lifting over Germany is associated typically with thunderstorms and lofting over Russia, near the Ural Mountains, occurs because of frequent weak low-pressure systems in the area [Lydolph, 1977]. Rainfall reaches a maximum in August around the Ural Mountains as summer showers are frequent [Lydolph, 1977]. These two regions of high precipitation are evident in the July-average rainfall for the entire 11-year study period, 1987-1997, which indicates that they are regular features (Figure 18). The average maximum monthly rainfall is 50-75 mm over Germany and Russia. The standard deviation of precipitation (Figure 18) is 10-25 mm over Germany and 25-50 mm over Russia, which indicates that the year-to-year variation in total rainfall is greater over Russia than Germany. Trajectories initialized at a potential temperature of 320 K (~500 mb) over the maximum areas of rainfall in Germany and Russia in July (not shown) indicate that the flow is dominated by the prevailing westerlies at that altitude. Figure 19a shows July-average “European” CO in 1994 as a vertical slice along 10°E latitude through Germany, Italy, and between the Algerian/Libyan border. “European” CO between 50°60°N is lofted into the upper troposphere over Germany increasing concentrations there by greater than 20 ppbv. O3 and O3-precursors, such as NOx, from Europe are lofted also to the free troposphere. Upper tropospheric measurements of NOx taken in the summer of 1995 indicate an area of high concentrations over the Ural Mountains [Brunner et al., 1998], although it is difficult to differentiate between lightning and anthropogenic sources. Figure 19b shows the monthly-average difference in July 1994 at 500 mb in O3 between SIM-SUM1 and SIM-SUM2, ∆O3, which is due only to the effects of trace gas emissions and aerosols from Europe, as the transport (i.e., advection and convection) is the same in both simulations. The ∆O3 is greater than 4 ppbv over Germany and over a broad area of Russia, and is greater than 2 ppbv over the 17
North Atlantic Ocean, Siberia and northern Africa. In addition to the areas of lofting over Germany and Russia, a third area of lofting is evident over northern Africa in Figure 19ab, although it is not associated with moist convection as discussed in more detail in the next section. 4.3. Africa and Near East In this section, we focus on the transport of European pollution in summer to northern Africa and the Near East, though transport occurs year-round in the model (Figure 3). Most of the land area of northern Africa and the Near East is sparsely populated, but there are about 200 million inhabitants, concentrated in the Nile River Valley, Turkey, and along the southern and eastern coasts of the Mediterranean Sea. As we will demonstrate in this section, European pollution makes a substantial contribution to total surface O3 to these regions in summer. The Mediterranean Basin is under the influence of the westerlies in winter and the quasipermanent Azores High in summer [Davis et al., 1997]. The southward flow of air and, subsequently pollution, from Europe occurs to the Mediterranean Basin in the lower troposphere between the Alps and the Pyrenees, the Alps and the Balkans, and the Balkans and Turkish Highlands [Leroux, 2001]. The flow is driven by the pressure difference created by the Azores High and the area of low pressure associated with the Southwest Asian monsoon [Griffiths, 1972] (Figure 14a). Lelieveld et al. [2002] reported that 60-80% of CO in the boundary layer over the Mediterranean in summer is attributable to European sources and that the 8-hour mean O3 standard of 55 ppbv is exceeded regularly there because of European pollution. Little lofting of pollution occurs in the Mediterranean Basin in summer as depressions are rare [Wallén, 1970]. The flow of pollution continues from the Mediterranean Basin into Africa, mainly across Libya and Egypt as the Atlas Mountains, extending from Morocco to Tunisia, hinder the flow [Leroux, 2001]. Some of the flow from Europe goes into the Near East to the southeast of the Zagros Mountains in eastern Iran, moving over Arabia and into the northern Indian Ocean [Leroux, 2001]. Once in northern Africa, the flow of European pollution is controlled predominantly by the Intertropical Front (ITF), which separates the Atlantic Monsoon (southwesterly, humid oceanic air) from the Harmattan Trade (northeasterly, dry continental air) in summer [Asnani, 1993; Jonquières et al., 1998; Leroux, 2001]. The humid, oceanic monsoon (0-1.5 km) flows northward from the Atlantic Ocean over hundreds of kilometers into northern Africa under the dry, continental air [Asnani, 1993; Jonquières et al., 1998; Leroux, 2001]. The leading edge of 18
the ITF reaches nearly to the Tropic of Cancer by August. The European pollution moves to the west in the Harmattan Trade over northern Africa and, eventually, over the North Atlantic Ocean. This flow of pollution is evident in Figure 17 as a “tongue” of “European” CO column reaching from Africa over the Atlantic Ocean. Some of the European pollution moves into the African Easterly Jet (AEJ), which flows westward at approximately 45 km/h above the oceanic air in July and is located generally between 500 and 700 mb at about 15°N, south of the surface line of the ITF [Jonquières et al., 1998; Leroux, 2001]. Almost all the “European” CO over Africa lies below 600 mb (Figure 19a). Little rainfall occurs near the ITF as the hot dry continental air (i.e., less dense) is uplifted by the moist cool oceanic air. Convection is suppressed typically by shearing, however squall lines can move from east to west in the region covered by the Atlantic monsoon [Asnani, 1993; Leroux, 2001]. Lifting of European pollution over northern Africa, albeit weaker than moist convective lifting, occurs via the quasi-permanent Western Saharan Low, which is situated between the Atlas and Ahaggar Mountains near the surface location of the ITF (Figure 15). The lifting of pollution by the low is evident in ∆O3 at 500 mb (Figure 19ab), which is about 5 ppbv over much of northwestern Africa. Observations are sparse in northern Africa and the Near East during our study period. There are NOAA/CMDL stations in the Sede Boker, Israel (WIS), Tenerife, Canary Islands (IZO), which is downwind of the Sahara in the Atlantic Ocean, and Assekrem, Algeria (ASK), which is located in the Ahaggar Mountains in the south of the country (Table 2; Figure 3). The transport of European pollution to northern Africa and the Near East occurs year-round (Figure 3), but the CO concentration maximizes in winter in these regions partly because of the long lifetime of CO (i.e., >3 months) at that time and, subsequent, seasonal growth of its burden in the Northern Hemisphere. The trajectory analyses presented in Figures 7-9 reveals that the southward flow of pollution occurs in January 1995 (NAO+), but not in January 1996 (NAO–). There is a secondary CO maximum in summer that is evident in the model and due to “European” sources. However, this secondary maximum is less obvious in the short record of observations. For instance, it is seen in the model in the summers of 1996 and 1997 at Sede Boker, Israel and Assekrem, Algeria, but only seen in the observations in 1996 at the two stations. There is no evidence of the summertime maximum at the station at Tenerife, Canary Islands in either the observations or model, possibly because of the slow transport times of pollution from Europe to the site and the short lifetime of CO in summer (i.e., 3-4 weeks). 19
Figure 19c shows the July-average European contribution to surface O3 (ppbv) (i.e., ∆O3 between the SIM-SUM1and SIM-SUM2 runs) in 1994. The ∆O3 is greater than 2 ppbv over a broad area from the North Atlantic Ocean across Europe and northern Africa to Asia, greater than 20 ppbv over much of Europe, and greater than 30 ppbv over the Netherlands/Belgium region and the central Mediterranean. An interesting feature is that ∆O3 from the surface to about 700 mb is 5-20 ppbv, or 5-40% of the total O3, over much of northern Africa and the Near East. The total model O3 (not shown) is between 50 and 60 ppbv there. This finding indicates that the European Council human health standard for O3, 55 ppbv averaged over 8 hours, is exceeded regularly during July 1994 in the model over much of northern Africa and the Near East as a result of O3 generated from European pollution. The southward flow of pollution in the lower troposphere from Europe to these regions is typical in summer (Figure 14a), which implies that European pollution causes exceedences of the European O3 standard each summer there.
5. Conclusions We presented simulations of the GEOS-CHEM model of 3-d trace gas/aerosol chemistry and transport for 11 years, 1987-1997, with the purpose to identify the seasonal and interannual variations of the export pathways of pollution from Europe. We conducted sensitivity simulations to examine the factors (i.e., year-to-year changes in fossil fuel emissions and meteorology) that control the interannual variations of export pathways and to assess the impact of European pollution on summertime, surface O3 in downwind regions. The model is driven by assimilated meteorology and uses an emission inventory of trace gases specific to the study period. Its ability to capture most of the variations of observed CO in regions affected by the pollution lends confidence for its use as a tool to study the export pathways of the pollutants and their impacts on tropospheric chemistry. The export of pollution from Europe in fall, winter, and spring from 1987 to 1997 occurred by advection by three primary pathways: i. to the northeast in the prevailing westerlies to Russia, the Russian Arctic and Siberia; ii. to the northwest to the North Atlantic Ocean and European Arctic; and iii. to the south to the North Atlantic Ocean via the Iberian Peninsula and via the Mediterranean Basin and northern Africa. As concluded in other studies, we also found that pollution export is limited generally to the lower troposphere in winter. In summer, the export of pollution from Europe was controlled largely by the strength and position of the Azores High. The winds in summer were generally lighter than in winter as 20
Europe was under the influence of a ridge of high pressure associated with the Azores High. The dominant flow of pollution was to the west over Russia and to the south over the Mediterranean Basin and northern Africa. Convection was also an important mechanism for the export of pollution in summer, unlike in winter. There were two maximum areas of convection. The first was centered over Germany and Poland and the second over Russia near the Ural Mountains. The surface pollution was lofted to the upper troposphere, a region of strong westerlies. The lofting (not associated with moist convection) of European pollution into the middle troposphere occurred over northwestern Africa by the quasi-permanent West Saharan Low. In fact, model O3 was enhanced by as much as five ppbv on average in the middle troposphere in the regions of these three regions of lofting (i.e., Germany, Russia, and northwestern Africa). The two dominant causes of interannual variation in the export of European pollution over our study period are variations in transport, especially associated with the NAO and convection, and changes in European anthropogenic emissions. During our study period, 1987-1997, the NAO was in the positive phase about half of the winter months (i.e., December through March) and in the negative phase about 20%. The remaining months were neutral. In general, the export of pollution from the European source regions was more frequent when the NAO was in the positive phase as the winds were strong over Europe. Consequently, the highest levels of pollution were shifted to the northwest of the main emissions regions. Conversely, the pollution concentrations were highest over the source regions when the phase of the NAO was negative as the winds were relatively weak over the main European emission regions. The tropospheric burden of CO emitted from European sources varied by as much as ±20% over both western Europe and the North Atlantic Ocean and ±15% over the Arctic during the 11-year study period because of interannual variations in transport. When the NAO was in the positive (negative) phase, the burden of CO from Europe tended to be lower (higher) over the North Atlantic Ocean and higher (lower) over the Arctic. There were no obvious tendencies for the regional burdens when the NAO was in the neutral phase. Hoerling et al. [2001] argued that the trend in the NAO over the last 50 years is to the positive phase because there has been a progressive warming of the equatorial oceans, which has resulted in a change of atmospheric circulation that has strengthened the mid-latitude westerlies. There is a high degree of uncertainty in the prediction of the NAO, but there could be an intensification of the NAO to the positive phase as global warming takes place [IPCC, 2001 and references therein; Gillett et al., 2003].
21
On average, emissions of CO and NOx decreased by 36% and 20%, respectively, over our study period, however there were larger declines in eastern Europe, especially Russia, than western Europe. The emissions in Spain, Portugal, and Poland increased. There was a nearly linear decrease in the “European” CO burden in winter in the model with the decrease in CO emissions. “European” CO at the surface declined by more than five ppbv/year over almost all of Europe and more than 25 ppbv/year over parts of eastern Europe in a model sensitivity simulation where the effects of year-to-year variations in meteorology were removed. However, in a model simulation that included both the year-to-year variations in emissions and transport, the trend in “European” CO was actually positive for our study period over parts of western Europe. Therefore, we conclude that it is necessary to consider both the effects of changes in emissions and year-to-year variations in transport in order to interpret trends in observed CO over Europe. We demonstrated that O3 generated from European pollution enhanced regularly summer O3 levels not only in Europe, but also in northern Africa and the Near East, regions with a population of about 200 million people. In July 1994, model O3 formed from European pollution increased local levels in northern Africa and the Near East by 5-20 ppbv on average, which caused the European Council’s human health standard for O3 to be regularly exceeded there. The southward, surface flow of air, and subsequently of pollution, from Europe was favored in the summer during our study period because of the pressure gradient between the Azores High and the Southwest Asian monsoon. On average in July 1994 in the model, European pollution did not contribute significantly to O3 levels in populated regions outside of Europe other than northern Africa and the Near East. While there are a number of long-term observational stations for CO and O3 in Europe, only sparse observations are available to evaluate our model in the regions affected by European pollution export. Thus, our work aids in the design of measurement networks and field campaigns to sample exported European pollution. Acknowledgements. We gratefully acknowledge L. Yurganov who provided us with CO column data from Zvenigorod, Russia. The data from the Jungfraujoch have been derived by the group of the University of Liège, with funding primarily provided by the Office of Scientific, Technical and Cultural affairs (OSTC), Brussels. We thank R. B. Pierce and T. D. Fairlie for use of the NASA Langley trajectory model and Inna A. Megrestkaia for processing NOAA/CMDL CO data.
22
References Allen, D. J., P. Kasibhatla, A. M. Thompson, R. B. Rood, B. G. Doddridge, K. E. Pickering, R. D. Hudson, and S.-J. Lin, Transport-induced interannual variability of carbon monoxide determined using a chemistry and transport model, J. Geophys. Res., 101, 28,655-28,669, 1996. Asnani, G.C. Tropical Meteorology, Vol. 1, Nobel Printers Pvt. Ltd., Pune, India, 1993. Auvray et al., A modeling study of the background ozone over Europe: origin and interannual variability, in preparation, 2003 Berntsen, T.K., S. Karlsdottir, and D.A. Jaffe, Influence of Asian emissions on the composition of air reaching the North Western United States, Geophys. Res. Lett., 26, 2171-2174, 1999. Bethan, S., G. Vaughan, C. Gerbig, A. Volz-Thomas, H. Richer, and D.A. Tiddeman, Chemical air mass differences near fronts, J. Geophys. Res., 103, 13,413-13,434, 1998. Bey, I., D.J. Jacob, J.A. Logan, and R.M. Yantosca, Asian chemical outflow to the Pacific: Origins, pathways and budgets, J. Geophys. Res., 106, 23,097-23,113, 2001a. Bey, I., D.J. Jacob, R.M. Yantosca, J.A. Logan, B.D. Field, A.M. Fiore, Q. Li, H. Liu, L. J. Mickley, and M. Schultz, Global modeling of tropospheric chemistry with assimilated meteorology: Model description and evaluation, J. Geophys. Res., 106, 23,073-23,095, 2001b. Brunner, D., J. Staehelin, and D. Jeker, Large-scale nitrogen oxide plumes in the tropopause region and implications for ozone, Science, 282, 1305-1309, 1998. Cape, J.N., J. Methven, and L.E. Hudson, The use of trajectory cluster analysis to interpret trace gas measurements at Mace Head, Ireland, Atmos. Environ., 34, 3651-3663, 2000. Chandra, S., J.R. Ziemke, P.K. Bhartia, and R.V. Martin, Tropical tropospheric ozone: Implications for dynamics and biomass burning, J. Geophys. Res., 10.1029/2001JD000447, 2002. Chatfield, R.B., J.A. Vastano, L. Li, G.W. Sachse, and V.S. Connors, The Great African plume from biomass burning: Generalizations from a three-dimensional study of TRACE A carbon monoxide, J. Geophys. Res., 103, 28,059-28,077, 1998. Chatfield, R.B., Y. Guo, G.W. Sasche, D.R. Blake, and N.J. Blake, The subtropical global plume in the Pacific Exploratory Mission-Tropics A (PEM-Tropics A), PEM-Tropics B, and the Global Atmospheric Sampling Program (GASP): How tropical emissions affect the remote Pacific, J. Geophys. Res., 107, 10.1029/2001JD000497, 2002. Collins, W.J., D.S. Stevenson, C.E. Johnson, and R.G. Derwent, The European regional ozone distribution and its links with the global scale for the years 1992 and 2015, Atmos. Environ., 34, 255-267, 2000. Davis, R.E., B.P. Hayden, D.A. Gay, W.L. Phillips, and G.V. Jones, The North Atlantic Subtropical Anticyclone, J. Climate, 10, 728-744, 1997. Duncan, B.N., I. Bey, M. Chin, L.J. Mickley, T.D. Fairlie, R.V. Martin, and H. Matsueda, Indonesian Wildfires of 1997: Impact on Tropospheric Chemistry, J. Geophys. Res., 108, doi:10.1029/2002JD003195, 2003a. Duncan, B. N. et al., Interannual variability of carbon monoxide (1988-1997): 1. Model formulation, evaluation, and sensitivity, in preperation, 2003b. EMEP, Transboundary Air Pollution in Europe, Part 1: Emissions, dispersion, and trends of acidifying and eutrophying agents, EMEP/MSC-W Report 1/97, Norwegian Meteorological Institute, P.O. Box 43-Blinden, N0313 Oslo 3, Norway.
23
EMEP, Transboundary acidification, eutrophication and ground level ozone in Europe, EMEP Report 1 and 2/2002, Norwegian Meteorological Institute, P.O. Box 43-Blinden, N-0313 Oslo 3, Norway. Fehsenfeld, F.C., P. Daum, W.R. Leaitch, M. Trainer, D.D. Parrish, and D. Hübler, Transport and processing of O3 and O3 precursors over the North Atlantic: An overview of the 1993 North Atlantic Regional Experiment (NARE) summer intensive, J. Geophys, 101, 28,877-28,891, 1996. Fiore, A.M., D.J. Jacob, B.D. Field, D.G. Streets, S.D. Fernandes, and C. Jang, Linking ozone pollution and climate change: The case for controlling methane, Geophys. Res. Lett., 29(19), 1919, doi:10.1029/2002GL015601, 2002a. Fiore, A.M., D.J. Jacob, I. Bey, R.M. Yantosca, B.D. Field, A.C. Fusco, and J.G. Wilkinson, Background ozone over the United States in summer: Origin,trend, and contribution to pollution episodes, J. Geophys. Res., 107 (D15), doi:10.1029/2001JD000982, 2002b. Fischer, F., M. de Reus, M. Traub, J. Williams, J. Lelieveld, J. de Gouw, C. Warneke, H. Schlager, A. Minikin, R. Scheele, and P. Siegmund, Deep convective injection of boundary layer air into the lowermost stratosphere at midlatitudes, Atmos. Chem. Phys., 3, 739-745, 2003. Forster, C., et al., Transport of boreal forest fire emissions from Canada to Europe, J. Geophys. Res., 106, 22,88722,906, 2001. Gillett et al., Climate change and the North Atlantic Oscillation, in The North Atlantic Oscillation: Climatic Significance and Environmental Impact, Eds. J.W. Hurrell et al., 193-210, 2003. Griffiths, J.F., Climates of Africa, World Survey of Climatology, Vol. 10, Elsevier Scientific Publishing Company, New York, 1972. Hamelin, B., F.E. Grousset, P.E. Biscaye, and A. Zindler, Lead Isotopes in Trade Wind Aerosols at Barbados: The Influence of European Emissions Over the North Atlantic, J. Geophys. Res., 94, 16,243-16,250, 1989. Hoell, J.M., D.D. Davis, S.C. Liu, R.E. Newell, H. Akimoto, R.J. McNeal, and R.J. Bendura, The Pacific Exploratory Mission – West Phase B: February-March, 1994, J. Geophys. Res. 102, 28,223-28,239, 1997. Hoerling, M.P., J.W. Hurrell, and T. Xu, Tropical origins for recent North Atlantic climate change, Science, 292, 9092, 6 April 2001. Hov, Ø., and F. Flatøy, Convective redistribution of ozone and oxides of nitrogen in the troposphere over Europe in summer and fall, J. Atmos. Chem., 28, 319-337, 1997. Hsu, N.C., J.R. Herman, P.K. Bhartia, C.J. Seftor, O. Torres, A.M. Thompson, J.F. Gleason, T.F. Eck, and B.N. Holben, Detection of biomass burning smoke from TOMS measurements, Geophys. Res. Lett., 23, 745-748, 1996. Hurrell, J.W., Decadal Trends in the North Atlantic Oscillation: Regional Temperatures and Precipitation, Science, 269, 676-679, 1995. Hurrell, J., Influence of variations in extratropical wintertime teleconnections on Northern Hemisphere temperature, Geophys. Res. Lett., 23, 665-668, 1996. Hurrell, J.W., and C.K. Folland, A change in the summer atmospheric circulation over the North Atlantic, Exchanges: Selected Research Papers, Climate Variability & Predictability, World Climate Research Programme, No. 25, September 2002. Hurrell, J.W., Y. Kushnir, G. Ottersen, and M. Visbeck, An overview of the North Atlantic Oscillation, in The North Atlantic Oscillation: Climatic Significance and Environmental Impact, Eds. J.W. Hurrell et al., 1-35, 2003. IPCC (Intergovernmental Panel on Climate Change), Climate Change 2001: The Scientific Basis, Eds. J.T. Houghton et al., Cambridge University Press, 2001.
24
Iversen, T., On the atmospheric transport of pollution to the Arctic, Geophys. Res. Lett., 11, 457-460, 1984. Jacob, D.J., J.A. Logan, and P.P. Murti, Effect of rising Asian emissions on surface ozone in the United States, Geophys. Res. Lett., 26, 2175-2178, 1999. Jacob, D.J., J.H. Crawford, M.M. Kleb, V.E. Connors, R.J. Bendura, J.L. Raper, G.W. Sachse, J.C. Gille, and L. Emmons, The Transport and Chemical Evolution over the Pacific (TRACE-P) aircraft mission: design, execution, and overview of first results, J. Geophys. Res., in press, 2003. Jaffe, D., A. Mahura, J. Kelley, J. Atkins, P. C. Novelli, and J. Merrill, Impact of Asian emissions on the remote North Pacific atmosphere: Interpretation of CO data from Shemya, Guam, Midway and Mauna Loa, J. Geophys. Res., 102, 28,627-28,635, 1997. Jaffe, D. et al., Transport of Asian air pollution to North America, Geophys. Res. Lett., 26, 711-714, 1999. Jones, P.D., T.J. Olborn, and K.R. Briffa, Pressure-based measures of the North Atlantic Oscillation (NAO): A comparison and an assessment of changes in the strenght of the NAO and in its influence on surface climate parameters, in The North Atlantic Oscillation: Climatic Significance and Environmental Impact, Eds. J.W. Hurrell et al., 51-62, 2003. Jonquières, I., A. Marenco, and A. Maalej, Study of ozone formation and transatlantic transport from biomass burning emissions over West Africa during the airborne Tropospheric Ozone Campaigns TROPOZ I and TROPOZ II, J. Geophys. Res., 103, 19,059-19,073, 1998. Klonecki, A., P. Hess, L. Emmons, L. Smith, and J. Orlando, Seasonal changes in the transport of pollutants into the Arctic troposphere-model study, J. Geophys. Res., 108 8367, doi:10.1029/2002JD002199, 2003. Lamarque, J.-F., and P.G. Hess, Model analysis of the temporal and geographical origin of the CO distribution during the TOPSE campaign, J. Geophys. Res., 108, doi:10.1029/2002JD002077, 2003. Lelieveld, J., et al., Global air pollution crossroads over the Mediterranean, Science, 298, 794-799, 2002. Leroux, M., The Meteorology and Climate of Tropical Africa, Springer/Praxis Publishing, 2001. Li, Q., D.J. Jacob, I. Bey, R.M. Yantosca, H. Lie, R.V. Martin, A.M. Fiore, B.D. Field, B.N. Duncan, and V. Thouret, A tropospheric ozone maximum over the Middle East, Geophys. Res. Lett., 28, 3235-3238, 2001. Li, Q., D.J. Jacob, I. Bey, P.I. Palmer, B.N. Duncan, B.D. Field, R.V. Martin, A.M. Fiore, R.M. Yantosca, D.D. Parrish, P.G. Simmonds, and S.J. Oltmans, Transatlantic transport of pollution and its effect on surface ozone in Europe and North America, J. Geophys. Res., 10.1029/2001JD001422, 2002a. Li, Q., D.J. Jacob, T.D. Fairlie, H. Liu, R.M. Yantosca, and R.V. Martin, Stratospheric versus pollution influences on ozone at Bermuda: Reconciling past analysis, J. Geophys. Res., 10.1029/2002JD00213, 2002b. Liu, H., D.J. Jacob, L.Y. Chan, S.J. Oltmans, I. Bey, R.M. Yantosca, J.M. Harris, B.N. Duncan, and R.V. Martin, Sources of tropospheric ozone along the Asian Pacific Rim: An analysis of ozonesonde observations, J. Geophys. Res., 10.1029/2001JD002005, 2002. Liu, H., D.J. Jacob, I. Bey, R.M. Yantosca, B.N. Duncan, and G.W. Sachse, Transport pathways for Asian combustion outflow over the Pacific: Interannual and seasonal variations, J. Geophys. Res., submitted, 2003. Logan, J.A., An Analysis of Ozonesonde Data for the Troposphere: Recommendations for Testing 3-D Models, and Development of a Gridded Climatology for Tropospheric Ozone, J. Geophys. Res., 104, 16115-16149, 1999. Lydolph, P.E., Climates of the Soviet Union, World Survey of Climatology, Vol. 7, Elsevier Scientific Publishing Company, New York, 1977.
25
Mahieu, E., R. Zander, L. Delbouille, P. Demoulin, G. Roland, and C. Servais, Observed trends in total vertical column abundances of atmospheric gases from IR solar spectra recorded at the Jungfraujoch, J. Atmos. Chem., 28, 227-243, 1997. Marenco, A., J.C. Medale, and S. Prieur, Study of tropospheric ozone in the tropical belt (Africa, America) from STRATOZ and TROPOZ campaigns, Atmos. Environ., 24A, 2823-2834, 1990. Martin, R.V., K. Chance, D.J. Jacob, T.P. Kurosu, R.J.D. Spurr, E. Buscuela, J.F. Gleason, P.I. Palmer, I. Bey, A.M. Fiore, Q. Li, R.M. Yantosca, and R.B.A. Koelemeijer, An improved retrieval of tropospheric nitrogen dioxide from GOME, J. Geophys. Res., 10.1029/2001JD001027, 2002a. Martin, R.V., D.J. Jacob, J.A. Logan, I. Bey, R.M. Yantosca, A.C. Staudt, Q. Li, A.M. Fiore, B.N. Duncan, H. Liu, P. Ginoux, and V. Thouret, Global model analysis of TOMS and in-situ observations of tropical tropospheric ozone, J. Geophys. Res., 107(D18), 4351, 10.1029/2001JD001480, 2002b. Martin, R.V., D.J. Jacob, R.M. Yantosca, M. Chin, and P. Ginoux, Global and regional decreases in tropospheric oxidants from photochemical effects of aerosols, J. Geophys. Res., 108(D3), 4097, doi:10.1029/2002JD002622, 2003. Newell, R.E., and M.J. Evans, Seasonal changes in pollutant transport to the North Pacific: the relative importance of Asian and European sources, Geophys. Res. Lett., 27, 2509-2512, 2000. Novelli, P. C., L. P. Steele, and P. P. Tans, Mixing ratios of carbon monoxide in the troposphere, J. Geophys. Res., 102, 12855-12861, 1992. Novelli, P.C., K.A. Masarie, P.M. Lang, B.D. Hall, R.C. Myers, and J.W. Elkins, Re-analysis of tropospheric CO trends: Effects of the 1997, 1998 wildfires, J. Geophys. Res., 108, 10.1029/2002JD003031, 2003. Palmer, P.I., D.J. Jacob, K. Chance, R.V. Martin, R.J.D. Spurr, T. P. Kurosu, I. Bey, R. Yantosca, A. Fiore, and Q. Li, Air mass factor formulation for spectroscopic measurements from satellites: Application to formaldehyde retrievals from the Global Ozone Monitoring Experiment, J. Geophys. Res., 106, 14,539-14,550, 2001. Palmer, P.I., D.J. Jacob, A.M. Fiore, R.V. Martin, K. Chance, and T.P. Kurosu, Mapping isoprene emissions over North America using formaldehyde column observations from space, J. Geophys. Res., 108, doi:10.1029/2002JD002153, 2003. Pierce, R.B., and T.D. Fairlie, Chaotic advection in the stratosphere: Implications for the dispersal of chemically perturbed air from the polar vortex. J. Geophys. Res., 98, 18,589-18,595, 1993. Pierce, R.B., T.D. Fairlie, W.L. Grose, R. Swinbank, and A. O'Neill, Mixing processes within the polar night jet, J. Atmos. Sci., 51, 20, 2957-2972, 1994. Pochanart, P., H. Akimoto, Y. Kajii, V.M. Potemkin, and T.V. Khodzher, Regional background ozone and carbon monoxide variations in remote Siberia/East Asia, J. Geophys. Res., 108, doi:10.1029/2001JD001412, 2003. Prather, M., et al., Fresh air in the 21st century?, Geophys. Res. Lett., 30, doi:10.1029/2002GL016285, 2003. Press, W. H., B. P. Flannery, S. A. Teukolsky, and W. T. Vetterling, Numerical Recipes: The Art of Scientific Computing, pp. 484-488, Cambridge Univ. Press, New York, 1989. Purvis R.M., et al, 2003-06-20 Rapid lifting of nonmethane hydrocarbons in a cold front over central Europe, J. Geophys. Res, 108, 4224, doi:10.1029/2002JD002521, 2003. Rogers, C.M., and K. Bowman, Transport of smoke from the Central American fires of 1998, J. Geophys. Res., 106, 28,357-28,368, 2001. Schnell, Arctic haze and the Arctic gas and aerosol sampling program (AGASP), Geophys. Res. Lett., 11, 361-364, 1984.
26
Schubert, S. D., R. B. Rood, and J. Pfaendtner, An assimilated data set for earth science applications, Bull. Amer. Meteorol. Soc., 74, 2331-2342, 1993. Shipman, M.C., A.S. Bachmeier, D.R. Cahoon, Jr., and E.V. Browell, Meteorological overview of the Arctic Boundary Layer Expedition (ABLE 3A) Flight Series, J. Geophys. Res., 97, 16,395-16,419, 1992. Simmonds, P.G., S. Seuring, G. Nickless, and R.G. Derwent, Segregation and interpretation of ozone and carbon monoxide measurements by air mass origin at the TOR station Mace Head, Ireland from 1987 to 1995, J. Atmos. Chem., 28, 45-49, 1997. Staudt, A.C., D.J. Jacob, J.A. Logan, D. Bachiochi, T.N. Krishnamurti, and G.W. Sachse, Continental sources, transoceanic transport, and interhemispheric exchange of carbon monoxide over the Pacific, J. Geophys. Res., 106, 32,571-32,589, 2001. Stohl, A., and T. Trickl, A textbook example of long-range transport: Simultaneous observation of ozone maxima of stratospheric and North American origin in the free troposphere over Europe, J. Geophys. Res., 104, 30,44530,462, 1999. Stohl, A., A 1-year Lagrangian “climatology” of airstreams in the Northern Hemisphere troposphere and lowermost stratosphere, J. Geophys. Res., 106, 7263-7279, 2001. Stohl, A., S. Eckhardt, C. Forster, P. James, and N. Spichtinger, On the pathways and timescales of intercontinental air pollution transport, J. Geophys. Res., 107, 10.1029/2001JD001396, 2002. Stohl, A., C. Forster, S. Eckhardt, N. Spichtinger, H. Huntrieser, J. Heland, H. Schlager, S. Wilhelm, F. Arnold, and O., Cooper, A backward modeling study of intercontinental pollution transport using aircraft measurements, J. Geophys. Res., 108, doi:10.1029/2002JD002862, 2003. Traub, M., H. Fischer, M. de Reus, R. Kormann, J. Heland, H. Ziereis, H. Schlager, R. Holzinger, J. Williams, C. Warneke, J. de Gouw, and J. Lelieveld, Chemical characteristics assigned to trajectory clusters during the MINOS campaign, Atmos. Chem. Phys. Discuss., 3, 107-134, 2003. Trickl, T., O.R. Cooper, H. Eisele, P. James, R. Mücke, and A. Stohl, Intercontinental transport and its influences on the ozone concentrations over central Europe: Three case studies, J. Geophys. Res., 108, doi:10.1029/2002JD002735, 2003. Wallén, C.C., Climates of Northern and Western Europe, World Survey of Climatology, Vol. 5, Elsevier Scientific Publishing Company, New York, 1970. Walker, G.T., Correlations in seasonal variations of weather, IX. Mem. Ind. Meteor. Dept., 24, 75-332, 1924. Walker, G.T., and E. W. Bliss, World weather V. Mem. Roy. Meteor. Soc, 4, 53-84, 1932. Wang, Y., D. J. Jacob, and J. A. Logan, Global simulation of tropospheric O3-NOx-hydrocarbon chemistry: 1. Model formulation, J. Geophys. Res., 103, 10,713-10,725, 1998. Wild, O., and H. Akimoto, Intercontinental transport of ozone and its precursors in a three-dimensional global CTM, J. Geophys. Res., 106, 27,729-27,744, 2001. Wotawa, G., and M. Trainer, The influence of Canadian forest fires on pollutant concentrations in the United States, Science, 288 (5464), 324-328, 2000. Yurganov, L.N., E.I. Grechko, and A.V. Dzhola, Long-Term Measurements Of Carbon Monoxide Over Russia Using A Spectrometer Of Medium Resolution, Recent Research Developments in Geophysics, 4, 249-265, 2002.
27
Figure List Figure 1. Average emission rate (x1012 molec cm-2 s-1) of a) CO and c) NOx from fossil fuel combustion in 1987. Ratio of emission rates (unitless) for 1997 to 1987 of b) CO and d) NOx. The maximum and minimum values of the scales are capped for clarity. Figure 2. Monthly-averaged observed (black) and model (red) CO columns (x1018 molec cm-2) from the SIM-TAGCO run at Jungfraujoch, Switzerland (top) and Zvenigorod, Russia (bottom). Station locations are given in Table 2. A break in the line connecting the observed data indicates missing data. Figure 3. Time series of monthly-average model surface CO from the SIM-TAGCO run with observations from measurement stations that are part of the NOAA/CMDL network. The model’s total and “European” CO are represented by fine and heavy solid lines, respectively, and the observations by filled circles. Station locations are given in Table 2. Figure 4. a) Mean prevailing winds (m/s) at 800 mb with mean SLPs (mb) and b) standard deviation of SLPs (mb) for January from 1987 to 1997. Mean prevailing winds (m/s) at 800 mb with mean SLPs (mb) for January in c) 1994, d) 1995, e) 1996, and f) 1997. The strength of the wind is indicated by the size of the arrow. The wind speed associated with the longest arrow is shown in the lower left-hand side of the top panel. Figure 5. January-average NAOI from 1987 to 1997 as calculated from observed (solid line) and model (dashed line) SLPs at Ponta Delgada, Azores and Reykjavik, Iceland. Figure 6. Forward trajectories (5 days) from Moscow, Russia, initiated at potential temperatures of 280 K, 285 K, 295 K, 285 K, and 280 K in January, April, July, and October 1995, and January 1996, respectively. The first trajectory begins on the first of the month and the last one begins on the last day of the month. Figure 7. Same as Figure 6, except for Madrid, Spain. Figure 8. Same as Figure 6, except for Brussels, Belgium. Figure 9. Same as Figure 6, except for Milan, Italy. Figure 10. January-average “European” CO columns (x1018 molec cm-2) in 1995, 1996 and 1997. Figure 11. Monthly-average “European” CO burden (Tg) for four regions for three model runs: SIM-TAGCO (fine solid line), SIM-FF87 (heavy solid line), and SIM-MET87 (dashed line). Figure 12. Monthly anomaly (%) of the tropospheric burden of model “European” CO versus the monthly NAOI for December through March (black dots) and October-November/April-May (gray dots). Figure 13. Average annual change in model “European” CO (ppbv/year) at the surface in January from 1987 to 1997 in the SIM-TAGCO (top) and SIM-FF87 (bottom) runs. Decreases 28
are shown as gray lines, and increases and no change as black lines. Spacing of the contours is 50, -25, -10, -5, -1, 0, 1, 5, 10, 25, and 50 ppbv/year. Figure 14. a) Mean prevailing winds (m/s) at 800 mb with mean SLPs (mb) and b) standard deviation of SLPs (mb) for July from 1987 to 1997. Mean prevailing winds (m/s) at 800 mb with mean SLPs (mb) for July in c) 1987, d) 1995, and e) 1996. The strength of the wind is indicated by the size of the arrow. The wind speed associated with the longest arrow is shown in the lower left-hand side of the top panel. Figure 15. Forward trajectories (5 days) from Moscow, Russia, Madrid, Spain, Brussels, Belgium, and Milan, Italy, initiated at a potential temperature of 295 K in July 1987. The first trajectory begins on the first of the month and the last one begins on the last day of the month. Figure 16. Same as Figure 15, except for July 1996. Figure 17. Same as Figure 10, except for July of 1995 and 1996. Figure 18. Monthly precipitation (mm) in the GEOS-CHEM model for July in 1987 and 1994, and July-average rainfall for 1987-1997 with standard deviation (mm). Figure 19. a) Average vertical cross-section of “European” CO (ppbv) in July 1994 from the SIM-TAGCO run along 15°E latitude. b) ∆O3 (ppbv) in July 1994 between the SIM-SUM1and SIM-SUM2 runs at 500 mb. c) Same as b), except at the surface.
29
Table 1. Annual Fossil Fuels Emissions
CO (Tg) Emissions ∆Emissions Ratio NOx (Tg) Emissions
Year
Europe1
Western Europe2
Eastern Europe3
87 97 97-87 97/87
174 112 -62 0.64
71 53 18 0.75
103 59 44 0.57
87 32.5 17.3 15.1 97 25.9 14.4 11.5 ∆Emissions 97-87 -6.6 2.9 3.6 Ratio 97/87 0.80 0.83 0.76 1 36°N -48°N; 17°W-72°E and 48°N -88°N; 17°W-180°. 2 22°E.
Table 2. Station Information for Observed Column CO (x1018 molec cm-2) and Observed Surface NOAA/CMDL CO (ppbv) with Statistical Comparison to Model Output (SIM-TAGCO) Station Location
Latitude deg
Longitude deg
Altitude m
Na
R2b
∆c
∆d %
Std. Err.e
Column Sites Jungfraujoch, Switzerland 46.5ºN 8.0ºE > 3600 107 0.52 -0.007 0.7 0.012 Zvenigorod, Russia > 200 116 0.43 -0.070 -1.4* 0.031 55.7°N 37.6°E Surface Sites Ny-Alesund, Svalbard (ZEP) 78.9ºN 11.9ºE 475 46 0.69 7.0 6.4* 2.9 Ocean Station M (STM) 66.0ºN 2.0ºE 5 46 0.60 11.3 8.6* 3.2 Storhofdi, Iceland (ICE) 63.3ºN 20.3ºW 127 59 0.75 6.9 5.9* 2.0 Baltic Sea, Poland (BAL)f 55.4ºN 17.1ºE 28 62 0.61 34.0 19.4* 5.5 Mace Head, Ireland (MHD) 53.3ºN 9.9ºW 25 79 0.65 11.6 9.5* 2.6 Hegyhatsal, Hungary (HUN)f 47.0ºN 16.7ºE 344 57 0.70 -25.4 -8.6* 5.4 Constanta, Romania (BSC) 44.1ºN 28.7ºE 3 32 0.72 -2.1 -0.9 6.1 Terceira Is., Azores (AZR) 38.8ºN 27.4ºW 40 34 0.63 -3.3 -2.0 2.3 Sede Boker, Israel (WIS)f 31.1ºN 34.9ºE 400 25 0.40 2.6 2.7 3.8 Tenerife, Canary Is. (IZO) 28.3ºN 16.5ºW 2360 72 0.67 2.1 3.2 1.4 Assekrem, Algeria (ASK) 23.2ºN 5.4ºE 2730 26 0.62 4.3 5.4 2.1 a N is the number of measurement points coincident with model output; bR is the linear correlation coefficient; cMean bias is the difference between the model and observed CO (x1018 molec cm-2 for column and ppbv for surface measurements); dMean bias (%) relative to the mean observed CO; eStandard error (x1018 molec cm-2 for column and ppbv for surface measurements) of mean bias (∆); fStatistics for adjacent model box. See Section 2.2 for details; *Station has a statistically significant bias; the range of approximately two standard errors about the mean bias contains both positive and negative values.
30
Figure 1. Average emission rate (x1012 molec cm-2 s-1) of a) CO and c) NOx from fossil fuel combustion in 1987. Ratio of emission rates (unitless) for 1997 to 1987 of b) CO and d) NOx. The maximum and minimum values of the scales are capped for clarity. 31
Figure 2. Monthly-averaged observed (black) and model (red) CO columns (x1018 molec cm-2) from the SIM-TAGCO run at Jungfraujoch, Switzerland (top) and Zvenigorod, Russia (bottom). Station locations are given in Table 2. A break in the line connecting the observed data indicates missing data.
32
Figure 3. Time series of model surface CO from the SIM-TAGCO run with observations from measurement stations that are part of the NOAA/CMDL network. The model’s total and “European” CO are represented by fine and heavy solid lines, respectively, and the observations by filled circles. Station locations are given in Table 2. 33
Figure 4. a) Mean prevailing winds (m/s) at 800 mb with mean SLPs (mb) and b) standard deviation of SLPs (mb) for January from 1987 to 1997. Mean prevailing winds (m/s) at 800 mb with mean SLPs (mb) for January in c) 1994, d) 1995, e) 1996, and f) 1997. The strength of the wind is indicated by the size of the arrow. The wind speed associated with the longest arrow is shown in the lower left-hand side of the top panel. 34
Figure 5. January-average NAOI from 1987 to 1997 as calculated from observed (solid line) and model (dashed line) SLPs at Ponta Delgada, Azores and Reykjavik, Iceland.
35
Figure 6. Forward trajectories (5 days) from Moscow, Russia initiated at potential temperatures of 280 K, 285 K, 295 K, 285 K, and 280 K in January, April, July, and October 1995, and January 1996, respectively. The first trajectory begins on the first of the month and the last one begins on the last day of the month.
36
Figure 7. Same as Figure 6, except for Madrid, Spain.
37
Figure 8. Same as Figure 6, except for Brussels, Belgium.
38
Figure 9. Same as Figure 6, except for Milan, Italy.
39
Figure 10. January-average “European” CO columns (x1018 molec cm-2) in 1995, 1996 and 1997.
40
Figure 11. Monthly-average “European” CO burdens (Tg) for four regions for three model runs: SIM-TAGCO (fine solid line), SIM-FF87 (heavy solid line), and SIM-MET87 (dashed line).
41
Figure 12. Monthly anomaly (%) of the tropospheric burden of model “European” CO versus the monthly NAOI for December through March (black dots) and October-November/April-May (gray dots).
42
Figure 13. Trends in model “European” CO (ppbv/year) at the surface in January from 1987 to 1997 in SIM-TAGCO (top) and SIM-MET87 (bottom) runs. Downward trends are shown as gray lines, upward trends as black lines, and no trend as a black line. Spacing of the contours is 50, -25, -10, -5, -1, 0, 1, 5, 10, 25, and 50 ppbv/year.
43
Figure 14. Same as Figure 4, except for July.
44
Figure 15. Forward trajectories (5 days) from Moscow, Russia, Madrid, Spain, Brussels, Belgium, and Milan, Italy, initiated at a potential temperature of 295 K in July 1987. The first trajectory begins on the first of the month and the last one begins on the last day of the month.
45
Figure 16. Same as Figure 15, except for July 1996.
46
Figure 17. Same as Figure 10, except for July of 1995 and 1996.
47
Figure 18. Monthly precipitation (mm) in the GEOS-CHEM model for July in 1987 and 1994, and July-average rainfall for 1987-1997 with standard deviation (mm).
48
Figure 19. a) Average vertical cross-section of “European” CO (ppbv) in July 1994 from the SIM-TAGCO run along 15°E latitude. b) ∆O3 (ppbv) in July 1994 between the SIM-SUM1and SIM-SUM2 runs at 500 mb. c) Same as b), except at the surface.
49
