Environ. Sci. Technol. 2002, 36, 3230-3233
Transboundary Atmospheric Lead Pollution Y I G A L E R E L , * ,†,‡ T A M A R A X E L R O D , † ALAIN VERON,§ YITZAK MAHRER,| PETROS KATSAFADOS,⊥ AND U R I D A Y A N ‡,# Institute of Earth Sciences, The Multi-Disciplinary Center for Environmental Research, Faculty of Agriculture, and Department of Geography, The Hebrew University of Jerusalem, Israel, CNRS 132, CEREGE, Universite Aix-Marseille III, 13545 Aix en Provence Cedex 4, France, and University of Athens, Laboratory of Meteorology, Ippocratous 33, Athens 10680, Greece
A high-temporal resolution collection technique was applied to refine aerosol sampling in Jerusalem, Israel. Using stable lead isotopes, lead concentrations, synoptic data, and atmospheric modeling, we demonstrate that lead detected in the atmosphere of Jerusalem is not only anthropogenic lead of local origin but also lead emitted in other countries. Fifty-seven percent of the collected samples contained a nontrivial fraction of foreign atmospheric lead and had 206Pb/207Pb values which deviated from the local petrol-lead value (206Pb/207Pb ) 1.113) by more than two standard deviations (0.016). Foreign 206Pb/207Pb values were recorded in Jerusalem on several occasions. The synoptic conditions on these dates and reported values of the isotopic composition of lead emitted in various countries around Israel suggest that the foreign lead was transported to Jerusalem from Egypt, Turkey, and East Europe. The average concentration of foreign atmospheric lead in Jerusalem was 23 ( 17 ng/m3, similar to the average concentration of local atmospheric lead, 21 ( 18 ng/ m3. Hence, the load of foreign atmospheric lead is similar to the load of local atmospheric lead in Jerusalem.
Introduction Since the introduction of unleaded gasoline there has been a decline in the concentration of lead in the atmosphere of remote locations (1) and some countries, including Israel (2, 3). However, the concentrations of atmospheric lead in other countries remain high (3). Lead has four stable isotopes that in certain cases allow for the identification of lead sources in the environment (2-10). These sources include gasoline combustion, nonferrous smelting, and mining (11, 12). In Israel, because smelting and mining activities are nonsignificant, petrol-lead combustion remains the major source of lead in the atmosphere (10). This is particularly evident in an urban center such as Jerusalem. In a previous study we * Corresponding author phone: 972-2-6586515; fax: 972-25662581; e-mail:
[email protected]. † Institute of Earth Sciences, The Hebrew University of Jerusalem. ‡ The Multi-Disciplinary Center for Environmental Research, The Hebrew University of Jerusalem. § Universite Aix-Marseille III. | Faculty of Agriculture, The Hebrew University of Jerusalem. ⊥ University of Athens. # Department of Geography, The Hebrew University of Jerusalem. 3230
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 15, 2002
determined that during the late 1990s, the 206Pb/207Pb isotopic ratio of Israeli petrol-lead was relatively constant, with a value of 1.113 ( 0.008 (10) (1σ). This value was determined by measuring lead in gasoline and by repeatedly collecting aerosols along major roads in the country. In the same study we were able to demonstrate that some of the lead sampled along the Israeli Mediterranean coast in August 1995 originated from Ukraine (10). Approximately 90% of the air trajectories within the atmospheric mixed layer reach Israel from five major source regions including the following: (1) northwest Europe, (2) Eastern Europe, (3) Saudi Arabia and Jordan, (4) North African coast, and (5) inland North Africa (13). The northwest Europe source prevails throughout the year, while the Eastern Europe source is most frequent in the summer and in the winter (13). The Arabian source contributes material in the fall, usually during well-developed Red Sea trough conditions. The North African sources are mostly active in the spring and are associated with the passage of warm and shallow depressions crossing the southwestern parts of Israel (13, 14). To identify possible sources for nonlocal atmospheric lead in Jerusalem, we used published 206Pb/207Pb values of lead emitted in the countries located in the source regions of the air trajectories that reach Israel (3, 9, 13) and applied atmospheric transport models based on the synoptic conditions prevailing during the sampling period.
Experimental Methods Aerosol Collection and Analysis. To improve the temporal resolution of our aerosol sampling we designed a fully automated collector and sampled aerosols in 3-h intervals. The collector was situated on top of the Earth Sciences building within the Hebrew University campus in western Jerusalem. Well-established clean techniques were used in order to prevent sample contamination (15). The aerosols were collected on precleaned 0.5-µm pore-size Teflon filters that were then eluted with 10 mL of 0.5 M HNO3 for 24 h. Lead concentrations were measured with a Perkin-Elmer Elan 6000 ICP-MS (1σ ) 5%), and 206Pb/207Pb ratios were determined with a VG sector TIMS. The TIMS instrumental precision and accuracy (1σ ) 0.1%) were continuously monitored by using SRM-981 standards. Numerical Modeling. The numerical model is consisted of a regional atmospheric modeling system (RAMS (16)) and a transport and diffusion model (TDM). The RAMS provides the necessary meteorological fields for the TDM. The RAMS is a state-of-the-art, well-documented meso-scale atmospheric model, a multipurpose 3-D versatile numerical prediction model designed to simulate weather systems. It is constructed around a full set of equations in a terrainfollowing coordinate system, which governs atmospheric motions. The equations are supplemented with optional parametrizations for turbulence, radiation, thermodynamics, clouds, soil type, and vegetation. The RAMS is equipped with a multiple grid-nesting scheme that allows a two-way interaction between computational grids of different spatial resolution. In the current simulation, the RAMS was executed in hierarchical, two-level nested grids to allow zooming in from synoptic scale phenomena to the meso-scale dynamics. The telescoping from large-scale environment, low-resolution grid cells to small-scale atmospheric systems with finemeshed high-resolution grid cells enables small-scale atmospheric features of the target area to be taken into account while simultaneously providing the impact of much larger meteorological systems. 10.1021/es020530q CCC: $22.00
2002 American Chemical Society Published on Web 06/25/2002
FIGURE 2. Daily changes in 206Pb/207Pb and lead concentration values in aerosols collected in Jerusalem on May 1st-2nd, 1997.
FIGURE 1. 206Pb/207Pb values of aerosol samples collected between March 1996 and May 1997 at the Hebrew University, Givat Ram Campus, Jerusalem.
clearly demonstrates the transboundary invasion of contaminant lead into Jerusalem.
Results and Discussion
Lead isotopic values of two foreign sources in addition to local petrol-lead were measured within 2 days of May (1st2nd) 1997, during an advance of tropical air in the region. The isotopic compositions and the concentrations of lead in the atmosphere of Jerusalem on May 1st and 2nd, 1997 are plotted in Figure 2. Between 9:04 and 11:50 AM, May 1st, the sampled lead had a 206Pb/207Pb value of 1.118, very similar to local petrol-lead (1.113 ( 0.016). The concentration of lead at that time period was 5.0 ng/m3. The next sample, collected between 12:04 and 14:48 of the same day, had a dramatically higher 206Pb/207Pb value of 1.151, although the concentration of lead increased only slightly to 7.7 ng/m3. The 206Pb/207Pb values of lead in Jerusalem went back toward local values in the next sample collected between 14:48 and 17:40 which had a 206Pb/207Pb value of 1.139 while the concentration of lead increased to 11 ng/m3. The last sample collected on that day between 17:40 and 20:22 had a 206Pb/ 207Pb value of 1.123 (within the local petrol-lead value of 1.113 ( 0.016) and lead concentration of 20 ng/m3 (Figure 2). The 206Pb/207Pb value of 1.151 is very similar to the 206Pb/ 207 Pb value measured in Cairo in the 1990s (1.153 ( 0.001), where the concentrations of lead have been 2900 ( 200 ng/ m3, 2 orders of magnitude higher than in Jerusalem (3). The concentrations of lead in the atmosphere of Cairo are so high that it is very likely that the reported isotopic composition truly represents local emissions, and hence, any foreign contribution should be insignificant.
Approximately 70 aerosol samples were collected between March 1996 and May 1997 (Figure 1). The average concentration of atmospheric lead in the urban sampling station (The Hebrew University of Jerusalem, Givat Ram Campus) has been 22 ( 17 ng per cubic meter (20), much lower than concentrations of atmospheric lead reported previously in Jerusalem (21), reflecting the decrease in petrol-lead emission in Israel since 1988 (22). In the late 1980s, allowable lead content in gasoline was lowered significantly, and unleaded petrol was put to use in the early 1990s. Hence, during the course of the current study approximately 50% of the gasoline in Israel was unleaded (22). Fifty-seven percent of the 70 samples we collected had 206Pb/207Pb values that deviated from the local petrol-lead value (206Pb/207Pb ) 1.113) by more than two standard deviations (0.016), which indicates a significant fraction of foreign atmospheric lead (Figure 1). Here we illustrate our methodology and the capability of stable lead isotopes to distinguish urban pollution sources with a case study that
Based on the synoptic chart at sea-level-pressure on May 1st, 1997 06 UTC (The Israeli Meteorological Service), it is apparent that weak south westerly flow at shallow atmospheric layers over southeastern Mediterranean was induced by a high-pressure system located over Jordan. At this time, several hours prior to the arrival of the air mass from Egypt to Israel, this weak surface flow (3.5 m s-1) flowed within a reduced mixed layer (256 m) and lead to a minimal ventilation rate of 896 m2 s-1. Such a poor ventilation condition favored accumulation of pollutants in the air-mass on its way to Israel. Barometric low from Saharan origin caused the synoptic conditions leading to an advance of continental tropical air over the eastern Mediterranean. The eastward movement of such cyclones has been occasionally blocked by a highpressure system formed above Saudi Arabia and Jordan. Under these conditions a pressure gradient should build between these two systems inducing southerly winds which could transport mineral dust and occasionally polluted air masses from North Africa into Israel.
The first grid was applied to an area of 1400 × 1400 kilometers (meso-R scale (17)) with a 20-km grid resolution to derive synoptic phenomena. The second grid was applied to areas of 250 × 350 km with a horizontal grid resolution of 5 km. The last grid also accounts for the sea/land breeze circulation along the coast of Israel and Egypt. The simulations were performed for 96 h (29/4/97-2/5/97) starting at 00:00. The simulations were initialized and updated every 6 h with ECMWF (European Center for Medium-Range Weather Forecasts) data fields. The topographic data was obtained from the GTOPO30 project which is a global digital elevation model (DEM) with a horizontal grid spacing of 30 arc s (approximately 1 km). The transport and diffusion model (TDM) is based on the work of Hurley and Physick (18) and Physick and Abbs (19). It is a Lagrangian 3-D model that simulates the motion of atmospheric pollutants under the influence of atmospheric flow. The TDM was applied to a high-resolution grid with a vertical grid resolution of 50 m and up to a height of 2 km. The TDM was initiated and driven by the meteorological fields produced by the RAMS and interpolated in time and space to the location of the pollutant elements.
VOL. 36, NO. 15, 2002 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
3231
FIGURE 3. Simulation of pollution migration from Cairo area to Jerusalem on May 1st, 1997 based on the RAMS and the TDM models for 96 h (29/4/97-2/5/97). The numerical model was applied to the same time frame by following the horizontal distribution of particles (aerosols) at a height of 100 m (AGL) emitted from the Cairo metropolitan area for selected hours from April 29th until May 2nd, 1997. Each emitted particle represents an amount of pollutant emitted per minute. At 24:00 LST on April 30th, 48 h after emission had begun, particles were only transported a short distance and were mainly mixed with freshly released pollution from the source area as a result of the light wind on these days. During the next 12 h (Figure 3, first frame) the polluted air mass was diffused while moving northward. On May 1st, 1997 at 12:00 LST (Figure 3, second frame) the polluted air mass which originated at Cairo, pushed by westerly prevailing winds, reached the area around Jerusalem more or less at the time when we actually recorded the shift in isotopic composition from local values toward Cairo-like isotopic composition (Figure 2). With the continuation of inland penetration of the particles over the mountain range, polluted air mass advanced eastward. The air coming from the west should have introduced local petrol-lead as well; however, this did not happen until after 14:48, as long as 3232
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 15, 2002
FIGURE 4. Daily changes in wind direction during May 2nd, 1997 based on the RAMS model. Jerusalem was above the boundary layer and did not receive pollutants from the coastal plain. This is indicated by the synoptic chart on May 1st, 1997 (The Israeli Meteorological Service). On May 2nd, 1997 the first sample was collected between 8:16 and 11:02 AM, and it had a 206Pb/207Pb value of 1.095 and relatively low lead concentration (3.5 ng/m3; Figure 2). The second sample collected between 12:30 and 15:16 had a higher 206Pb/207Pb value (1.105), closer to the local petrollead value, but maintained low lead concentration (3.1 ng/ m3; Figure 2). The third sample collected between 15:16 and 18:02 had a 206Pb/207Pb value of 1.119, within the range of local petrol-lead values (1.113 ( 0.016), and a similar lead concentration (3.9 ng/m3; Figures 1, 2). The 206Pb/207Pb value of 1.095, measured in the morning of this day, is very similar to the value of lead (206Pb/207Pb ) 1.096-1.100) emitted in the United Arab Emirates (and possibly Jordan), east of Israel (3). Indeed, until noon May 2nd the wind in Jerusalem was blowing from the east (Figure 4). The first frame illustrates the wind-field in the dense grid on May 2nd, 1997 at 09:00 UTC. It is apparent that the area around Jerusalem at this
time is dominated by a northeasterly wind. Four hours later (second frame) with the intensification of inland penetrating Mediterranean Sea breeze the southwesterly wind takes over. At this time we detected the change in 206Pb/207Pb values from foreign to local petrol-lead (Figure 2). The synoptic chart at sea-level-pressure on May 2nd, 1997 06 UTC (The Israeli Meteorological Service) indicates that warm depression located over the Nile Delta, Egypt impelling a southeasterly flow from south Jordan and Saudi Arabia toward Israel. The eastward movement of this depression which occurred during the following hours lead to an abrupt veering of the surface wind to a westerly component typified by an air mass originated along the Israeli coast. The ventilation rate during the early hours of May 2nd was excessively low (i.e., approximately 450 m2 s-1). Intensification of the westerly flow that occurred later was accompanied by a pronounced inflation of the mixed layer resulting to a ventilation rate exceeding 7200 m2 s-1. On both days the introduction of local lead coming from the coastal plain took place while the height of the atmospheric boundary layer was increasing which improved vertical mixing of the lower atmosphere and caused a dilution in lead concentrations. The increase in lead concentrations registered in the evening of May 1st, 1997 was accompanied by a decrease in the height of the atmospheric boundary layer (20), reducing the volume of the boundary layer and inducing an increase in the concentrations of lead in the lower atmosphere. Significant fluctuations in 206Pb/207Pb values and lead concentrations were recorded as well in Jerusalem in May, June, July, and November 1996 (20) (Figure 1). The synoptic conditions on these dates and reported values of the isotopic composition of lead emitted in various countries around Israel (3, 9) suggest that the foreign lead was transported to Jerusalem from Egypt, Turkey, and East Europe. The average concentration of foreign atmospheric lead in Jerusalem (206Pb/207Pb > 1.129 or < 1.097) was 23 ( 17 ng/m3, similar to the average concentration of local atmospheric lead (206Pb/ 207Pb ) 1.113 ( 0.016), 21 ( 18 ng/m3. Furthermore, it is interesting to note that the average 206Pb/207Pb value of all of our samples (1.133 ( 0.012, 1σ) is similar to the value measured by others (3) in Israel during the 1990s (1.1281.136 in Haifa and 1.128 in Jerusalem at the same site we collected our samples). Our average 206Pb/207Pb value (1.133 ( 0.012) integrates values of both local and foreign lead, as probably is the case for the other samples which were collected for several days (3). Because of the reduction of lead emissions in Israel, the isotopic composition of lead collected during conventional, long sampling periods reflects mixing of local atmospheric lead with foreign lead. Based on our isotopic and concentration data we argue that foreign lead has been a significant fraction of the atmospheric load of lead in Jerusalem in the past few years. Significance of Results. Shifts in the isotopic composition of atmospheric lead in major urban centers in Europe (Paris and Oslo) have been reported recently (2, 23). These shifts were attributed mostly to changes in local sources of lead (industrial and wood combustion versus petrol-lead). Nevertheless, it is also possible that these shifts reflect an increase
in the relative contribution of petrol-lead emitted in East and Central European countries where local emissions are much higher and the isotopic composition of lead is consistent with the observed changes in Paris and Oslo (2, 9, 23). More work is needed in order to test this possibility. The underlying features regarding the transport of lead pollution are generally well appreciated, and it is widely known that the consequences of atmospheric pollution are not restricted by national boundaries, nevertheless, we are among the first to demonstrate that transboundary migration of atmospheric lead affects not only remote sites but also urban centers where the introduction of unleaded gasoline has lowered local lead emissions. These results call for regional if not global cooperation in emission control of atmospheric pollutants, such as continued use of leaded petrol in the Middle East. No country will be able to eliminate atmospheric pollutants in its urban centers unless it invests in pollution control of other countries as well.
Literature Cited (1) Bourton, C. F.; Gorlach, U.; Candelone, J.-P.; Bolshov, M. A.; Delmas, R. J. Nature 1991, 353, 153. (2) Veron, A.; Flament, P.; Laure-Bertho, M.; Alleman, L.; Flegal, R.; Hamelin, B. Atmos. Environ. 1999, 33, 3377. (3) Bollhofer, A.; Rosman, K. J. R. Geochim. Cosmochim. Acta 2001, 65, 1727. (4) Flegal, A. R.; Nriagu, J. O.; Niemeyer, S.; Coals, K. H. Nature 1989, 339, 455. (5) Shirahata, H.; Elias, R. W.; Patterson, C. C. Geochim. Cosochim. Acta 1980, 49, 149. (6) Sturges, W. T.; Barrie, L. A. Nature 1987, 329, 144. (7) Sturges, W. T.; Barrie, L. A. Atmos. Eniron. 1989, 23, 2513. (8) Church, T. M.; Veron, A.; Patterson, C. C.; Settle, D.; Erel, Y.; Maring, H. R.; Flegal, A. R. Global Biogeochemical Cycles 1990, 4, 431. (9) Hopper, J. F.; Ross, H. B.; Sturges, W. T.; Barrie, L. A. Tellus 1991, 43b, 45. (10) Erel, Y.; Veron, A.; Halicz, L. Geochim. Cosmochim. Acta 1997, 61, 4495. (11) Nriagu, J. O.; Pacyna, J. M. Nature 1988, 333, 134. (12) Pacyna, J. M.; Scholtz, M. T.; Li, Y. F. Environ. Rev. 1995, 3, 145. (13) Dayan, U. J. Climate Appl. Meteor. 1986, 25, 591. (14) Alpert, P.; Ziv, B. J. Geophys. Res. 1989, 94, D15, 18495. (15) Erel, Y.; Morgan, J. J.; Patterson, C. C. Geochim. Cosochim. Acta 1991, 55, 707. (16) Pielke, R. A.; Cotton, W. R.; Tremback, C. J.; Lyons, W. A.; Grasso, L. D.; Nicholls, M.; Moran, M. D.; Wesley, D. A.; Lee, T. J.; Copeland, J. H. Meteor. Atmos. Phys. 1992, 49, 69. (17) Orlanski, I. Bull. Am. Meteor. Soc. 1975, 56, 527. (18) Hurley, P.; Physick, W. A. Atmos. Environ. 1991, 25A, 1313. (19) Pysick, W. L.; Abbs, D. J. Monthly Weather Rev. 1991, 119, 1014. (20) Axelrod, T. M.Sc. Dissertation, The Hebrew University of Jerusalem, 2000. (21) Malenky, B.; Grieken, R. V.; Dack, L. V.; Luria, M. Atmos. Environ. 1983, 17, 819. (22) Teutsch, N. Ph.D. Dissertation, The Hebrew University of Jerusalem, 1999. (23) Aberg, G.; Pacyna, J. M.; Stray, H.; Skjelkvale, B. L. Atmos. Environ. 1999, 33, 3335.
Received for review January 16, 2002. Revised manuscript received May 6, 2002. Accepted May 16, 2002. ES020530Q
VOL. 36, NO. 15, 2002 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
3233
