Docstoc

tracer_final

Document Sample
tracer_final Powered By Docstoc
					             A Study of Pollutant Transport Using an On-Line MM5 Tracer Model




                       S.-H. Chen1,@ T. Kindap2, J. Dudhia3, and J. S. Kain4




1
    Dept. of Land, Air, and Water Resources, University of California, Davis, CA 95616, USA
2
    Eurasia Institute of Earth Sciences, Istanbul Technical University, Maslak, 34469 Istanbul,

    Turkey
3
    Mesoscale and Microscale Meteorology Division/NCAR, Boulder, CO, USA
4
    National Severe Storms Laboratory, Norman, OK, USA




                             Submitted to Atmospheric Environment

                                           July 11, 2011




_________________________________________________
@
    Corresponding author's address: Dr. Shu-Hua Chen, Department of Land, Air, and Water

Resources, University of California, One Shields Avenue, Davis, CA 95616; E-mail:

shachen@ucdavis.edu.




                                                 1
                                           Abstract



       Istanbul is one of the most polluted cities in Turkey. The pollutants can be emitted

locally, but the long-range transport from other high polluted cities is also possible. Istanbul

recently experienced two high-pollution episodes, 7-8 and 10-12 January 2002. These two

episodes are examined to document the potential utility of MM5T, a new on-line tracer model

based on the fifth-generation Penn State/NCAR Mesoscale model (MM5).

       The simulated meteorological results from MM5T were comparable with

observations. The low-level northerly and northwesterly prevailing winds associated with a

high pressure system over Central Europe and a low over Western Russian, combined with a

frontal inversion layer in the boundary layer, were favorable for transporting pollutants from

several upstream sources of pollution, including Warsaw, Silesia, and Krakow in Poland,

Bucharest in Romania, and Plovdiv in Bulgaria. Numerical tracer simulations showed that

pollutants transported from these cities could have contributed to these two high pollution

events in Istanbul.    The significance of these results and potential uses of MM5T are

discussed.

Keywords: pollutants, long-range transport, on-line tracer model, MM5T




                                               2
1. Introduction

         Air quality has become an important issue to society since it strongly affects human

health and visibility, and may play a role in climate changes.                     Since local emission of

pollutants can be augmented significantly by long-distance transport, it is important to take

into account the transport effect when considering new policies aimed at improving air

quality.

         Numerous studies have examined the transport of airborne pollutants (e.g., Kubilay et

al. 2000; Toon et al. 1988; Westphal et al. 1988; Bergametti et al. 1989; Nickovic and

Dobricic, 1996; Kallos et al. 1998; Rodriguez et al. 2001; Kindap et al. 2006), but the problem

of pollutant transport from Europe to Northern and Western Turkey has received relatively

little attention. Yet, such transport may have contributed to two high pollution events that

occurred in Istanbul, Turkey on 7-8 and 10-12 January 2002. During these events, the

atmospheric circulation over central Europe was dominated by a cold-core surface anti-

cyclone, a climatologically favored feature during the cold season (Kallos et al. 1998). Air

trajectories associated with this anti-cyclone were capable of transporting high concentrations

of PM10 (particles with diameter < 10 μm ) over long distances to Istanbul. Therefore, we

hypothesize that PM10 transported from other highly polluted cities in Europe was partially

responsible for these two pollution events in Istanbul.

         Kindap et al. (2006) studied both air pollution episodes, 7-8 and 10-12 January 2002,

in Istanbul using an off-line1 air pollution model. They used meteorological data from the

fifth-generation NCAR / Penn State Mesoscale Model (MM5; Grell et al. 1994), which has

been often used to drive air pollution models (Seigneur et al. 2003; Lee et al. 2004; Hogrefe et

al. 2004; Jimenez et al. 2005). Their sensitivity simulations of different emission rates over

different countries showed that the contribution of long-range pollutant transport from

1
 The term off-line refers to a procedure in which an air pollution model utilizes output from a meteorological
model, but the two are run separately.


                                                        3
Romania was the highest compared with those from other countries and could reach a

maximum of 25% of PM10 concentrations in Istanbul.               However, because the tracer

techniques was not applied by their study, to estimate the percentage (i.e., 25%) it was

necessary to assume a linear relationship between the emission rate and the pollutant amount

that propagated to Istanbul.

        The off-line approach is commonly used in pollution transport and air quality studies.

Inconsistency between the air pollution model and the meteorological model is one of primary

errors for off-line air quality studies. Potential inconsistencies include the use of different

parameterization schemes (e.g., boundary layer turbulence mixing and deep convection)

between the two models and the interpolation of meteorological data in time and space

(Seaman 2000).     In particular, in most of the applications the meteorological fields are

updated quite infrequently (typically on the order of hours) due to the limitation of storage;

the temporal interpolation between update times can introduce significant errors in transport

computations, especially over the areas where boundary layer height and/or wind vary

significantly with time (e.g., mountainous regions and unstable lower atmosphere). Such

errors can make it very difficult to arrive at sound policy decisions regarding the mitigation of

air pollution.

        To ameliorate this problem, an on-line tracer model has been developed to investigate

long-range pollutant transport more accurately and precisely. In this study, the capabilities of

the on-line model are demonstrated by re-examining the two high pollution episodes that

occurred in Istanbul in January 2002. The meteorological output from the model is used to

document the model’s skill in reproducing the weather conditions that favored the

accumulation and transport of pollutants in the atmospheric boundary layer, while the

predicted concentrations of various pollutants are used to demonstrate the efficacy of the new

coupled tracer calculations. Further, these results are used to estimate the potential impact of



                                               4
pollutant transport compared to local generation of boundary-layer pollutants during these two

high pollution episodes in Istanbul.

       This paper is organized as follows.            The numerical model development and

experiment design are described in section 2. The results are discussed in section 3, and the

concluding remarks are given in section 4.



2. Model Development and Experiment Design

2.1 Development of an on-line tracer model

       Based on the MM5 version 3.6, an on-line tracer model (MM5T) was developed to

study the long-range transport of pollutants. In addition to the governing equations used in

the MM5 model, a continuity equation describing the amount of tracers present (C) was

introduced in MM5T and is written as:

        C    
            V  C  C pbl  C cov  SC  E C ,                           (1)
        t
      
where V is the three-dimensional wind vectors. The transport effects due to advection
    
(  V  C ), boundary layer mixing ( Cpbl), sub-grid cumulus convective mixing ( C cov ), and

sedimentation ( S C ) are taken into account.        E C is defined as the emission rate.   The

sedimentation is calculated using the formula:

           ρ              d (ρCVT )
       SC  o *10  3 * g           ,                                       (2)
            ρ                 dσ

where ρ o , ρ, σ, g, and VT are the base state density, full density, terrain-following vertical

coordinate, gravitational acceleration, and terminal velocity, respectively, and the terminal

velocity is simply calculated using VT  0.01 C0.2 .         In this simple version, chemistry

reactions are excluded and wet deposition is ignored as well (the latter is not an important

process for both events studied because very little precipitation was produced over the



                                                 5
relevant pollutant paths). Therefore, the only source term is the emission ( E C ) from the

surface and the only sink term is the deposition to the surface due to sedimentation in the

lowest model layer.

       The Medium Range Forecast (MRF) boundary layer scheme (Hong and Pan 1996),

which includes local and non-local mixing, and the Kain-Fritsch cumulus convection scheme

(Kain 2004), which includes deep and shallow convection, were chosen and modified to

account for the tracer mixing effects. With this configuration, tracer evolution is treated

essentially the same as that of any other scalar in the model, except that it has unique source

and sink terms.    This on-line approach can avoid temporal interpolation errors that can

inherently limit the accuracy of more commonly used off-line calculations of pollutant

transport and diffusion.

       In MM5T, tracers were designed as a 4D array which makes it easier to control the

number of tracers. The same type of pollutants emitted from different cities on different dates

can be tracked individually using different tracers.     Therefore, the source and date of

pollutants from upstream can be identified more accurately and precisely. This detailed

information can be important when policy decisions are made to improve air quality.



2.2. Experiment design

       Figure 1 shows the time series of observed PM10 concentration at Umraniye,

Uskudar, Besiktas, and Sarachane stations in Istanbul from January 5 to 12. Two high

pollutant episodes were consistently recorded in these stations during 7-8 (e.g., 48-72 h) and

10-12 (120-168 h) January 2002. While pollutants can be emitted locally, the long-range

transport is a possible contributor to high pollution events when weather conditions are

favorable. The new developed MM5T was used to investigate the possibility that pollutants

were transported from other cities upstream of Istanbul, Turkey.



                                              6
        For MM5T simulations, a single domain with grid-spacing of 30 km covering the

entire European continent and nearby seas was configured. It had 227 x 176 x 38 grid points

in the east-west, north-south, and vertical directions, respectively. In addition to the MRF

boundary layer parameterization and the Kain-Fritsch convection scheme, the RRTM (Rapid

Radiative Transfer Model) radiation scheme (Mlawer et al. 1997), and simple ice

microphysics scheme (Dudhia 1989) were chosen. A time step of 90 seconds was used.

        Two tracer simulations were conducted, experiment 1 (EXP1) and experiment 2

(EXP2), corresponding to the two peak events of PM10 concentration in Istanbul, 7-8 and 10-

12 January 2002. Specifically, the model was integrated from 00 UTC 5 to 00 UTC 8 January

for EXP1 and from 00 UTC 8 to 0 UTC 11 January for EXP2. The National Centers for

Environmental Prediction (NCEP) Global Data Assimilation System (GDAS) data (with 2.5

latitude/longitude resolution) were used for MM5T boundary and initial conditions.

        The emission source term, E C , was represented in an idealized form based on the

emission module developed by Kindap et al. (2006). This module can be used to downscale

in time from annual EMEP (the co-operative programme for monitoring and evaluation of

long range transmission of air pollutants in Europe) emission data to hourly rates. The

maximum emission rate from their module occurs at about 5 p.m. local time. This emission

cycle was mimicked in our tracer simulations and tracers were released from the surface into

the lowest model layer over a daily cycle as shown in Fig. 2, with a maximum about 3.3 x 10-3

unit s-1 at 5 p.m. local time.

        Since this study was designed only to demonstrate the potential utility of using MM5T

to track pollutants from their source, simulated emission sources were limited to a few

potential locations in Europe. In particular, Warsaw, Silesia, and Krakow in Poland; Kiev in

Ukraine; Moscow in Russia; Sofia and Plovdiv in Bulgaria; and Bucharest in Romania were

selected. These cities were chosen because they were potentially significant local sources of



                                              7
anthropogenic pollutants and they were positioned upstream of Istanbul on the dates in

question. The area of tracer emission in each city had a radius of 100 km, and the emission

rate is independent of the radius. To simplify this demonstration, the same emission function

was used for all source locations, while in reality the emission rate and amount can be very

different from one city to another. It is feasible to incorporate an emission module with more

realistic emission data into the MM5T model, but we leave it for future work.

       To further understand the transport characteristics, tracers from the same city but

different days (i.e., 00 UTC to 00 UTC next day) were tracked separately (i.e., saved in

different tracers). Note that tracers released from Silesia and Krakow were not distinguished

because of their similar pollutant characteristics.



3. Results and Discussions

       After 24 h of integration in EXP1, a synoptic-scale circulation pattern that favors low-

level pollutant transport from the selected locations into Turkey was quite evident (Fig. 3).

Specifically, a surface high-pressure center was positioned over central Europe while a

surface low was located over western Russia. The pressure gradient between these two

centers induced strong boundary-layer flow into Turkey from the north-northwest, creating a

mechanism for the fast transport of PM10 from upstream cities to Istanbul. The model

simulated wind fields were quite realistic in the vicinity of Istanbul during the first two days

of the simulation, though the consistency is not as good on the third day when observed winds

were relatively calm (Fig. 4). The model also simulated the low-level temperatures skillfully,

including the slow variation of temperature at Gokceada (sky was overcast) and the diurnal

oscillation at Uzunkopru.      The thermal profile above the surface was consistent with

observations as well. A surface cold front passed through the area during the first 12 h of

EXP1, leaving in its wake an exceptionally strong stable layer at the top of the boundary layer



                                                 8
(Fig. 5).     This stable layer suppressed vertical mixing, apparently allowing pollutants to

remain concentrated near the surface. Note that the upper level northeasterly wind in Fig. 5 is

due to the northeast-southwest orientation of the trough.

            Meteorological results from EXP2 are comparable with those from EXP1.

Therefore, we conclude that the MM5T is capable of producing sufficiently realistic

meteorological data to compute robust estimates of tracer transport. Kindap et al. (2006)

reached similar conclusions regarding their MM5 simulations, even though they used a

coarser resolution (grid spacing of 50 km).

       The snapshot of simulated tracer (pollutant) transport from various source cities to

Istanbul for EXP1 is plotted for the three most significant amounts of tracers in Fig. 6,. For

the first simulation period (00 UTC 5 January to 00 UTC 8 January), although pollution

released from Bucharest on the first day reaches Istanbul, low-level trajectories appear to

carry most of the Bucharest plume to the west of the city (dashed lines in Fig. 6a) and it

makes a relatively small contribution in Istanbul (white bars in Fig. 7a). However, pollution

released in Bucharest on the second day begins to reach Istanbul at 18 UTC 6 January (after

about 18 h transport time) and gives the maximum impact on the city at 00 UTC 7 January,

when the peak of the pollutant plume passes through the city (dashed lines in Fig. 6b and gray

bars in Fig. 7a). Pollution released from Plovdiv, Bulgaria on the first day moves southward

and does not pass over Istanbul (solid lines in Figs. 6a and white bars in Fig. 7b), while the

edges of the plumes released in Plovdiv on the second and third days skirt Istanbul and

contribute relatively small concentrations of pollutants to the city (solid lines in Figs. 6b and

gray bars in Fig. 7b). Since both Bucharest and Plovdiv are close to Istanbul, pollutants

released from both cities on the third day can reach Istanbul before the end of simulations.

The maximum local contribution from Plovdiv is only about 10% of that from Bucharest.




                                               9
        Silesia and Krakow in Poland are relatively far away from Istanbul, yet pollution

released from both cities on the first day reaches Istanbul after about 36 h of transport (by 12

UTC 6 January) and the influence on the city lasts for one and half days (shading in Fig. 6b

and white bars in Fig. 7c). Silesia’s and Krakow’s pollutants from the second day reach the

city as well, arriving at around 12 UTC 7 (i.e., 36 h transport) with a higher concentration

than the first day’s plume (shading in Fig. 6b and gray bars in Fig. 7c). While it takes more

time to transport to Istanbul, the peak amount of pollutants from Silesia and Krakow reaching

the city is about four times of that from Plovdiv and about 40% of that from Bucharest. In our

simulation, other selected cities such as Kiev, Sofia, and Warsaw have almost no contribution

of pollutants to Istanbul in the first event due to the directions of low-level winds.

        Results from EXP2 (Figs. 8 and 9) are very similar to those from EXP1 (Figs. 6 and

7). The contribution of pollutants from Warsaw in Poland is greater than that from Plovdiv

and is, therefore, plotted for the second case (i.e., EXP2). For those plotted tracers, the

pollutant amount transported to Istanbul is smaller and the impact duration on the city is

shorter (cf. Figs. 7 and 9) compared with EXP1. Only a small amount of pollutants released

on the third day from Warsaw can reach Istanbul. The transport times of pollutants from

different cities are slightly longer but comparable with those from EXP1, and the contribution

of long-range transport from Bucharest is still the highest (Fig. 9) while that from Silesia and

Krakow, Poland is also non-negligible (recall that the same magnitude of emissions is used

for all cities).

        As mentioned above, vertical dispersion of boundary-layer pollutants is suppressed by

strong static stability at the top of the boundary layer. This suppression was simulated well by

MM5T. A vertical cross section from EXP1 shows that the pollutant plumes released from

Bucharest, Silesia, and Krakow all remained concentrated below 1.5 km as they were

advected towards Istanbul (Fig. 10a). Similar results were obtained for EXP2 (Fig. 10b),



                                                10
which took place during a period of comparably strong boundary layer capping. Note that the

topography in this area is very sophisticated (Fig. 10) and wind directions and magnitudes can

vary with time significantly. Therefore, to reduce the errors of pollutant transport calculations

the on-line tracer approach is more appropriate than the off-line one for this region.

       MM5T indicated that remote sources of pollution made relatively small contributions

to the total pollutant concentration in Istanbul during times of peak local emissions (Fig. 11).

However, it showed that pollutants from upstream could have dominated the local air quality

at other times of day (e.g., the 48 and 72 h times in EXP1). The results are qualitatively

consistent with Kindap et al. (2006) in showing that upstream sources of pollution could have

had a significant impact on the total concentration of pollutants in Istanbul during these two

periods in January 2002.



4. Concluding Remarks

       An on-line tracer model based on the fifth-generation Penn State/NCAR Mesoscale

model, called MM5T, was developed to identify the sources of long-range pollutant transport

more accurately and precisely than comparable off-line models.            The effects of tracer

transport due to advection, boundary layer mixing, cumulus convective mixing, and

sedimentation were taken into account. Two high-pollution events that occurred in Istanbul,

Turkey on 7-8 and 10-12 January 2002 were studied using the MM5T. Different tracers were

used to represent pollutants released from selected cities on different days, using a diurnal

cycle of emission rate that was maximized at 5 p.m. local time. In this semi-idealized study,

the same emission rate was applied to all selected cities.

       The first step in analyzing model results was to verify accurate simulation of

meteorological fields, especially the low-level wind fields that play a crucial role in transport

calculations. Time-series of low-level wind and temperature fields in the vicinity of Istanbul



                                               11
showed good agreement with local observations. Furthermore, the model also reproduced the

larger-scale patterns well. In particular, it simulated a surface high-pressure system over

central Europe and a surface low over western Russia, with a substantial pressure gradient

between these two systems. This gradient induced strong north-northwesterly low-level flow

capable of transporting upstream pollutants towards Istanbul.            Moreover, the model

reproduced a strong frontal inversion over the path of tracer transport, a feature that

suppressed mixing at the top of the planetary boundary layer and effectively trapped low-level

pollutants near the ground. These weather conditions created a favorable environment for

long-range transport and limited dilution of pollutants.

       Results showed that pollutants originating in Bucharest, Plovdiv, Silesia, and Krakow

(and Warsaw for the second event) could have contributed to two high-pollution episodes on

7-8 and 10-12 January 2002 in Istanbul. Pollutants originating in Bucharest and Plovdiv

could have arrived in Istanbul after 18-24 h, while those emitted in Warsaw, Silesia, and

Krakow would have appeared within 36-48 h. Given the meteorological conditions associated

with these events, and an assumption that emission rates were identical in all cities, MM5T

indicated that pollutants from Bucharest would have made the largest contribution to remotely

generated pollution during these events, with those from Silesia and Krakow not far behind.

Our model suggests that locally generated pollutants would be present in much greater

concentrations than those from remote sources during times of maximum local emissions

(assuming, again, that emission rates from all cities are equal), but that imported pollutants

could have become predominant when local emissions were near minimum values.            These

results are qualitatively consistent with those from Kindap et al. (2006).

       This study documents the potential utility of MM5T.               Compared to off-line

calculations of pollutant transport and diffusion, this on-line model mitigates temporal

interpolation errors and is more accurate.      In the application described here, it uses a



                                               12
simplified approach that excludes chemical reactions and other important processes such as

wet deposition, but MM5T provides a powerful framework for much more sophisticated

applications. For example, it could be converted into a valuable tool for air pollution studies

or emergency response planning (say, in the event of an accidental or explosive release of

chemicals).   Thus, it could play an important role in policy decision-making and disaster

response, among other applications.



Acknowledgements: The authors would like to thank Mr. Jhih-Ying Chen and Dr. Kemal

Gurer who helped generate Fig. 4. This study was supported by the NASA-NAG5-13678

project.




                                              13
References

Bergametti, G., L. Gomes, G. Coudegaussen, P. Rognon, and M. N. Le Coustumer, 1989:

       African dust observed over Canary Islands- source-regions identification and transport

       pattern for some summer situation. J. Geophys. Res., 94, 14855–14864.

Dudhia, J., 1989:     Numerical study of convection observed during the winter monsoon

       experiment using a mesoscale two-dimensional model. J. Atmos. Sci., 46, 3077-3107.

Grell, G. A., J. Dudhia, and D. R. Stauffer, 1994: A description of the fifth-generation Penn

       State/NCAR mesoscale model (MM5). NCAR/TN-398+STR, National Center for

       Atmospheric Research, Boulder, CO, 122 pp.

Hogrefe, C., J. Biswas, B. Lynn, K. Civerolo, J-Y Ku, J. Rosenthal, C. Rosenzweig, R.

       Goldberg, and P. L. Kinney, 2004: Simulating regional-scale ozone climatology over

       the eastern United States: model evaluation results. Atmos. Environ., 38, 2627-2638.

Hong, S.-Y. and H.-L. Pan, 1996: Nonlocal boundary layer vertical diffusion in a medium-

       range forecast model. Mon. Wea. Rev., 124, 2322-2339.

Jimenez, P., R. Parra, S. Gasso, and J. M. Baldasano, 2005: Modeling the ozone weekend

       effect in very complex terrains: a case study in the Northeastern Iberian Peninsula.

       Atmos. Environ., 39, 429-444.

Kain, J. S., 2004: The Kain-Fritsch convective parameterization: An update. J. Appl. Meteor.,

       43, 170-181.

Kallos, G., V. Kotroni, K. Lagouvardos, and A. Papadopoulos, 1998: On the long-range

       transport of air pollutants from Europe to Africa. Geophys. Res. Lett., 25, 619-622.

Kindap, T., A. Unal, S.-H. Chen, Y. Hu, M. T. Odman, and M. Karaca, 2006: Long-range

       aerosol transport from Europe to Istanbul, Turkey. Atmos. Environ., in press.




                                              14
Kubilay, N., S. Nickovic, C. Moulin, and F. Dulac, 2000: An illustration of the transport and

       deposition of mineral dust onto the eastern Mediterranean. Atmos. Environ., 34, 1293-

       1303.

Lee, S-M, S-C Yoon, and D. W. Byun, 2004: The effect of mass inconsistency of the

       meteorological field generated by a common meteorological model on air quality

       modeling. Atmos. Environ., 38, 2917-2926.

Mlawer, E. J., S. J. Taubman, P. D. Brown, M. J. Iacono, and S. A. Clough, 1997: Radiative

       transfer for inhomogeneous atmospheres: RRTM, a validated correlated-k model for

       the longwave. J. Geophys. Res., 102, 16663-16682.

Nickovic, S. and S. Dobricic, 1996: A model for long-range transport of desert dust. Mon.

       Wea. Rev., 124, 2537–2544.

Rodriguez, S., X. Querol, A. Alastuey, G. Kallos, and O. Kakaliagou, 2001: Saharan dust

       contributions to PM10 and TSP levels in Southern and Eastern Spain. Atmos.

       Environ., 35, 2433-2447.

Seaman, N. 2000: Meteorological modeling for air-quality assessments. Atmos. Environ., 34,

       2231-2259.

Seigneur, C., B. Pun, K. Lohman, and S.-Y. Wu, 2003: Regional modeling of the atmospheric

       fate and transport of benzene and diesel particles. Environ. Sci. Technol., 37, 5236-

       5246.

Toon, O. B., R. P. Turco, D. Westphal, R. Malone, and M. S. Liu, 1988: A multidimensional

       model for aerosols: Description of computational analogs. J. Atmos. Sci., 45, 2123–

       2143.

Westphal, D.L., O. B. Toon, and T. N. Carlson, 1988: A case study of mobilization and

       transport of Saharan dust. J. Atmos. Sci., 45, 2145–2175.




                                             15
Figure Captions

Fig. 1: Time series of measured PM10 concentrations at Umraniye, Uskudar, Besiktas, and

       Sarachane observation stations in Istanbul from 00 UTC January 5 to 00 UTC January

       12, 2002 (Courtesy Kindap et al. 2006).

Fig. 2: The idealized emission rate that was used for MM5T simulations.

Fig. 3: 24 h simulation from EXP1, showing sea level pressure (shaded; hPa), 1.5-km

       potential temperature (solid lines; K), and 950-mb wind vectors at 00 UTC 6 January,

       2002. Marked cities were used for tracer experiments in this study.

Fig. 4: Observed (dashed lines) and 72 h MM5T simulated (solid lines) 10-m winds (knots; a

       full barb equals 10 knots and a half barb equals 5 knots) and 2-m temperature (oC) for

       (a) Gokceada and (b) Uzunkopru from EXP1 at stations near Istanbul.

Fig. 5: Vertical sounding at the location S1 in Fig. 6a (white dot) after a 12 h simulation from

       EXP1. A cold front inversion is shown between 900 mb and 800 mb.

Fig. 6: Snapshot of simulated tracers and wind vectors at the 100 m height at (a) 24 h and (b)

       48 h simulation from EXP1. Only the three most significant amounts of tracers, i.e.,

       from Bucharest (dashed lines), Plovdiv (solid lines), and Silesia and Krakow (shaded),

       that contributed pollutants to Istanbul are plotted. Tracers in (a) were released during

       the first day. In (b) tracers from Bucharest (dashed lines) and Plovdiv (solid lines)

       were released during the 2nd day and tracers from Silesia and Krakow (shaded) were

       released during the first 2 days. The interval is 0.02 units. S1 (white dot) in (a) is the

       location for the plot of the vertical sounding in Fig. 5.

Fig. 7: Time series of tracer in Istanbul. Tracers were emitted from (a) Bucharest (Romania),

       (b) Plovdiv (Bulgaria), and (c) Silesia and Krakow (Poland) for EXP1. White, gray,

       and black colors indicate tracers that were emitted on day 1 (0-24 h), day 2 (24-48 h),




                                               16
       and day 3 (48-72 h) from each particular city. Note that the plotted ranges of the y

       axis are different.

Fig. 8: Snapshot of simulated tracers and wind vectors at the 100 m height at (a) 24 h and (b)

       54 h simulation from EXP2. Only the three most significant amounts of tracers, i.e.,

       from Bucharest (dashed lines), Warsaw (solid lines), and Silesia and Krakow (shaded),

       that contributed pollutants to Istanbul are plotted. Tracers in (a) were released during

       the first day. In (b) tracers from Bucharest (dashed lines) were released during the 2nd

       day and tracers from Warsaw (solid lines) and Silesia and Krakow (shaded) were

       released during the first 2 days.

Fig. 9: Time series of tracer in Istanbul. Tracers were emitted from (a) Bucharest (Romania),

       (b) Warsaw (Poland), and (c) Silesia and Krakow (Poland) for EXP2. White, gray,

       and black colors indicate tracers that were emitted on day 1 (0-24 h), day 2 (24-48 h),

       and day 3 (48-72 h) from that particular city. Note that the plotted ranges of the y axis

       are different.

Fig. 10: Vertical cross-sections of simulated tracers below 3 km along the lines AB and CD in

       Figs. 7a and 7b for (a) EXP1 (48 h) and (b) EXP2 (54 h), respectively. In (a)

       Bucharest (dashed lines), Plovdiv (solid lines), and Silesia and Krakow (shaded) are

       plotted. In (b) Bucharest (dashed lines), Warsaw (solid lines), and Silesia and Krakow

       (shaded) are plotted.

Fig. 11: Time series of tracer in Istanbul. Tracers are contributed from locally emitted (gray

       color) and transported (black color) for (a) EXP1 and (b) EXP2.




                                              17
                          400
                                        Umraniye           Uskudar            Besiktas        Sarachane
                          350
 Concentration ( mg/m3)




                          300

                          250

                          200

                          150

                          100

                          50

                           0
                             0
                           Jan 5   12      24
                                          Jan 6    36     48
                                                        Jan 7   60    72
                                                                     Jan 8    84     96         120
                                                                                    Jan 9 108 Jan 10 132 Jan 11 156 Jan 12
                                                                                                          144         168
                                                                         Time (hr)
                                                                             Time




Fig. 1: Time series of measured PM10 concentrations at Umraniye, Uskudar, Besiktas, and
        Sarachane observation stations in Istanbul from 00 UTC January 5 to 00 UTC January
        12, 2002 (Courtesy Kindap et al. 2006).




                                                                             18
                                    4
Emission rate (x 10 -3 units s-1)




                                    3


                                    2


                                    1


                                    0
                                        0        12          24            36        48          60           72
                                                                     Time (h)


                                    Fig. 2: The idealized emission rate that was used for MM5T simulations.




                                                                      19
                                                        L




                     H




Fig. 3: 24 h simulation from EXP1, showing sea level pressure (shaded; hPa), 1.5-km
        potential temperature (solid lines; K), and 950-mb wind vectors at 00 UTC 6
        January, 2002. Marked cities were used for tracer experiments in this study.




                                        20
  (a)




   (b)




Fig. 4: Observed (dashed lines) and 72 h MM5T simulated (solid lines) 10-m winds (knots; a
        full barb equals 10 knots and a half barb equals 5 knots) and 2-m temperature (oC) for
        (a) Gokceada and (b) Uzunkopru from EXP1 at stations near Istanbul.



                                             21
Fig. 5: Vertical sounding at the location S1 in Fig. 6a (white dot) after a 12 h simulation from
        EXP1. A cold front inversion is shown between 900 mb and 800 mb.




                                              22
                       (a)




                                                                                unit

                       (b)




                                                                                unit




Fig. 6: Snapshot of simulated tracers and wind vectors at the 100 m height at (a) 24 h and (b)
        48 h simulation from EXP1. Only the three most significant amounts of tracers, i.e.,
        from Bucharest (dashed lines), Plovdiv (solid lines), and Silesia and Krakow (shaded),
        that contributed pollutants to Istanbul are plotted. Tracers in (a) were released during
        the first day. In (b) tracers from Bucharest (dashed lines) and Plovdiv (solid lines)
        were released during the 2nd day and tracers from Silesia and Krakow (shaded) were
        released during the first 2 days. The interval is 0.02 units. S1 (white dot) in (a) is the
        location for the plot of the vertical sounding in Fig. 5.


                                               23
                                (a)                    Bucharest
                          120
                                      Day 1    Day 2   Day 3
                          100

                           80
         Unit (x 10 )
         -3




                           60

                           40

                           20

                            0
                                 0        12      24      36
                                                       Plovdiv     48   60   72
                                (b)                     Time (h)
                          20
                                      Day 1    Day 2   Day 3
           Unit (x 10 )
         -3




                           0
                                0        12       24      36       48   60   72
                                (c)                      and (h)
                                                 Silesia Time Krakow
                          60
                                      Day 1    Day 2   Day 3


                          40
           Unit (x 10 )
         -3




                          20



                           0
                                0        12       24      36       48   60   72
                                                        Time (h)


Fig. 7: Time series of tracer in Istanbul. Tracers were emitted from (a) Bucharest (Romania),
        (b) Plovdiv (Bulgaria), and (c) Silesia and Krakow (Poland) for EXP1. White, gray,
        and black colors indicate tracers that were emitted on day 1 (0-24 h), day 2 (24-48 h),
        and day 3 (48-72 h) from each particular city. Note that the plotted ranges of the y
        axis are different.



                                                         24
                        (a)




                                                                              unit




                        (b)




                                                                              unit




Fig. 8: Snapshot of simulated tracers and wind vectors at the 100 m height at (a) 24 h and (b)
        54 h simulation from EXP2. Only the three most significant amounts of tracers, i.e.,
        from Bucharest (dashed lines), Warsaw (solid lines), and Silesia and Krakow (shaded),
        that contributed pollutants to Istanbul are plotted. Tracers in (a) were released during
        the first day. In (b) tracers from Bucharest (dashed lines) were released during the 2nd
        day and tracers from Warsaw (solid lines) and Silesia and Krakow (shaded) were
        released during the first 2 days.


                                              25
                                (a)                   Bucharest
                           80
                                      Day 1   Day 2   Day 3

           Unit (x 10 )
           -3              60


                           40


                           20


                            0
                                0        12      24         36    48   60      72
                                (b)                   Sarsaw
                                                       Time (h)
                           20
                                      Day 1   Day 2   Day 3
           Unit (x 10 )
           -3




                            0
                                0        12      24       36     48    60      72
                                (c)                     Time Krakow
                                                Silesia and  (h)
                           40

                                      Day 1   Day 2   Day 3
            Unit (x 10 )
           -3




                           20




                            0
                                0        12      24      36       48   60      72
                                                       Time (h)

Fig. 9: Time series of tracer in Istanbul. Tracers were emitted from (a) Bucharest (Romania),
        (b) Warsaw (Poland), and (c) Silesia and Krakow (Poland) for EXP2. White, gray,
        and black colors indicate tracers that were emitted on day 1 (0-24 h), day 2 (24-48 h),
        and day 3 (48-72 h) from that particular city. Note that the plotted ranges of the y axis
        are different.



                                                       26
                 (a)
                                                                                  unit




                 (b)
                                                                                  unit




Fig. 10: Vertical cross-sections of simulated tracers below 3 km along the lines AB and CD
         in Figs. 7a and 7b for (a) EXP1 (48 h) and (b) EXP2 (54 h), respectively. In (a)
         Bucharest (dashed lines), Plovdiv (solid lines), and Silesia and Krakow (shaded) are
         plotted. In (b) Bucharest (dashed lines), Warsaw (solid lines), and Silesia and
         Krakow (shaded) are plotted.




                                             27
                               (a)                     EXP1
                        1200
                                       Istanbul         Transport

                         900
         Unit (x 10 )
         -3




                         600


                         300


                           0
                                0      12         24      36      48     60   72
                                                         Time (h)



                               (b)                      EXP2
                        1200
                                     Istanbul          Transport

                         900
         Unit (x 10 )
         -3




                         600


                         300


                           0
                                0      12         24      36        48   60   72
                                                       Time (h)




Fig. 11: Time series of tracer in Istanbul. Tracers are contributed from locally emitted (gray
         color) and transported (black color) for (a) EXP1 and (b) EXP2.




                                                        28

				
DOCUMENT INFO
Shared By:
Categories:
Stats:
views:55
posted:7/11/2011
language:English
pages:28