• Adrian, G., and F. Fiedler, 1991: Simulation of unstationary wind and temperature fields over complex terrain and comparison with observations. Beitr. Phys. Atmos.,64, 27–48.

  • Anto, A. F., 1977: Observational studies on land–sea-breeze phenomena around the island of Sylt. Meteor. Rundsch.,30, 118–122.

  • Atkinson, B. W., 1981: Meso-scale Atmospheric Circulations. Academic Press, 495 pp.

  • Baer, M., and K. Nester, 1992: Parameterization of trace gas dry deposition velocities for a regional mesoscale diffusion model. Ann. Geophys.,10, 912–923.

  • Burt, W., H. Crew, N. Plutchak, and J. Dumon, 1974: Diurnal variation of winds over an upwelling region of Oregon. Bound.-Layer Meteor.,6, 35–45.

  • Businger, J. A., J. C. Wyngaard, Y. Izumi, and E. F. Bradley, 1971: Flux profile relationships in the atmospheric boundary layer. J. Atmos. Sci.,28, 181–189.

  • CETESB, 1985: Baixada santista, carta do meio ambiente e de sua dinimica. São Paulo, Brazil 33 pp. [Available from Frederico Hermann Jr., 345 Pinheiros, CEP 05489, São Paulo, Brazil.].

  • ——, 1987: Inventario de emissoes em fontes estacionarias de poluicao de poluicao do ar na regiao de Cubatão São Paulo, Brazil, 39 pp. [Available from Frederico Hermann Jr., 345 Pinheiros, CEP 05489, São Paulo, Brazil.].

  • ——, 1992: Acao da CETESB em Cubatão. Situacao em junho de 1992, São Paulo, Brazil, 40 pp. [Available from Frederico Hermann Jr., 345 Pinheiros, CEP 05489, São Paulo, Brazil.].

  • Carroll, J. J., and R. L. Baskett, 1979: Dependence of air quality in a remote location on local and mesoscale transports: A case study. J. Appl. Meteor.,18, 474–486.

  • Chang, J. S., R. A. Brost, I. S. A. Isaksen, S. Madronich, and P. Middleton, 1987: A three-dimensional Eulerian acid deposition model: Physical concepts and formulation. J. Geophys. Res.,92, 14681–14700.

  • Deardorff, J. W., 1978: Efficient prediction of ground surface temperature and moisture with inclusion of a layer of vegetation. J. Geophys. Res.,83, 1889–1903.

  • Degrazia, G. A., 1988: Anwendung von Ähnlichkeitsverfahren auf die turbulente Diffusion in der konvektiven und stabilen Grenzschicht. Ph.D. thesis, Universität Karlsruhe, 99 pp. [Available from Universität Karlsruhe, Kaiserstr. 12, 76131 Karlsruhe, Germany.].

  • Edinger, J. G., M. H. McCutchan, P. R. Miller, B. C. Ryan, M. J. Schroeder, and J. V. Behar, 1972: Penetration and duration of oxidant air pollution in the south coast air basin of California. J. Air Pollution Control Assoc.,22, 882–886.

  • Geiger, R., 1942: Das Klima der Bodennahen Luftschicht. Die Wissenschaft, 435 pp.

  • Haurwitz, B., 1947: Comments on the sea-breeze circulation. J. Meteor.,4, 1–8.

  • Koo, Y.-S., and D. D. Reible, 1995: Flow and transport modeling in a sea-breeze. Part II: Flow model application and pollutant transport. Bound.-Layer Meteor.,75, 209–234.

  • Kurita, H., H. Ueda, and S. Mitsumoto, 1990: Combination of local wind systems under gradient wind conditions and its contributionto the long-range transport of air pollutants. J. Appl. Meteor.,29, 331–348.

  • Lu, R., and R. P. Turco, 1994: Air pollutant transport in a coastal environment. Part I: Two-dimensional simulations of sea-breeze and mountain effects. J. Atmos. Sci.,51, 2285–2308.

  • ——, and ——, 1995: Air pollutant transport in a coastal environment–II: Three-dimensional simulations over Los Angeles Basin. Atmos. Environ.,29, 1499–1518.

  • Lyons, W. A., 1972: The climatology and prediction of the Chicago lake breeze. J. Appl. Meteor.,11, 1259–1270.

  • McElroy, J. L., and T. B. Smith, 1991: Lidar descriptions of mixing-layer thickness characteristics in a complex terrain/coastal environment. J. Appl. Meteor.,30, 585–597.

  • Munn, R. E., 1966: Descriptive Micrometeorology. Academic Press, 245 pp.

  • Neumann, J., and H. Savijarvi, 1986: The sea breeze on a steep coast. Beitr. Phys. Atmos.,59, 375–389.

  • Savijarvi, H., 1995: Sea breeze effects on large-scale atmospheric flow. Beitr. Phys. Atmos.,68, 335–344.

  • Schädler, G., N. Kalthoff, and F. Fiedler, 1990: Validation of a model for heat, mass and momentum exchange over vegetated surfaces using LOTREX-10E/HIBE88 data. Beitr. Phys. Atmos.,63, 85–100.

  • Schmidt, F. H., 1947: An elementary theory of the land- and sea-breeze circulation. J. Meteor.,4, 9–15.

  • Silva Dias, M. A. F., P. L. Vidale, and C. M. R. Blanko, 1995: Case study and numerical simulation of the summer regional circulation in São Paulo, Brazil. Bound.-Layer Meteor.,74, 371–388.

  • Simpson, J. E., 1994: Sea Breeze and Local Winds. Cambridge University Press, 234 pp.

  • Stockwell, W. R., P. Middelton, J. S. Chang, and X. Tang, 1990: The second generation regional acid deposition model—Chemical mechanism for regional air quality modeling. J. Geophys. Res.,95, 16343–16367.

  • Wakimoto, R. M., and J. L. McElroy, 1986: Lidar observation of elevated pollution layers over Los Angeles. J. Climate Appl. Meteor.,25, 1583–1599.

  • View in gallery

    Map of the area around Cubatão with the three measuring stations: São Bernardo, Barragem des Pedras, and Vicente de Carvalho.

  • View in gallery

    Land-use distribution in the area of Cubatão.

  • View in gallery

    Annual SO2 emission data in the area of Cubatão.

  • View in gallery

    Mean hodographs for the station of São Bernardo (dashed) and Vicente de Carvahlo (solid), as calculated from the November 1994–March 1995 dataset.

  • View in gallery

    Isolines of potential temperature (bottom) and wind vectors (top) at Vicente de Carvalho on 17 March 1995. The shaded areas indicate sea-breeze winds.

  • View in gallery

    Wind direction (top) (crosses: São Bernardo, triangles: Vicente de Carvalho, squares: Barragem des Pedras); wind speed (center) (dotted line: São Bernardo, dashed–dotted line: Vicente de Carvalho, solid line: Barragem des Pedras); and SO2 concentration (bottom) at Barragem des Pedras on 17 March 1995.

  • View in gallery

    Isolines of potential temperature (bottom) and wind vectors (top) at Vicente de Carvalho on 23 March 1995. The shaded areas indicate sea-breeze winds.

  • View in gallery

    Wind direction (top) (crosses: São Bernardo, triangles: Vicente de Carvalho, squares: Barragem das Pedras); wind speed (center) (dotted line: São Bernardo, dashed–dotted line: Vicente de Carvalho, solid line: Barragem des Pedras); and SO2 concentration (bottom) at Barragem des Pedras on 23 March 1995.

  • View in gallery

    Wind direction (modeled: squares, observed: circles); wind speed at Vicente de Carvalho; and SO2 concentrations at Barragem das Pedras (modeled: dashed lines, observed: solid lines) on 17 March 1995.

  • View in gallery

    Comparison of modeled and observed profiles of potential temperature (modeled: dashed lines, observed: solid lines) and wind direction (modeled: squares, observed: dots) at Vicente de Carvalho at 1200 (top) and 2400 LST (bottom) 17 March.

  • View in gallery

    Surface wind fields at 1200, 1800, and 2400 LST 17 March and at 0600 LST 18 March.

  • View in gallery

    Vertical cross sections from southeast to northwest (perpendicular to the coastline) at 1200 (top) and 2400 LST 17 March (bottom). The solid lines indicate the isolines of the potential temperature; the arrows indicate the component of the wind vector perpendicular to the coast.

  • View in gallery

    The SO2 concentration fields are shown at the lowest model level at 1200, 1800, and 2400 LST 17 March and at 0600 LST 18 March.

  • View in gallery

    Vertical cross section from southeast to northwest (perpendicular to the coastline) at 1200 LST 17 March (top), and vertical cross section from south to north at x = 15 km, 2400 LST 17, March (bottom). The solid lines indicate the isolines of the SO2 concentration; the arrows indicate the component of the wind vector in the corresponding plain.

  • View in gallery

    Daily sum of SO2 deposition on 17 March 1995.

  • View in gallery

    Wind direction (modeled: squares, observed: circles); wind speed at Vicente de Carvalho; and SO2 concentrations at Barragem des Pedras (modeled: dashed lines, observed: solid lines) on 23 March 1995.

  • View in gallery

    Comparison of modeled and observed profiles of potential temperature (modeled: dashed lines, observed: solid lines) and wind direction (modeled: squares, observed: dots) at Vicente de Carvalho at 1200 (top) and 2400 LST (bottom) on 23 March.

  • View in gallery

    Surface wind fields at 1200, 1800, and 2400 LST 23 March and at 0600 LST 24 March.

  • View in gallery

    SO2 concentration fields at the lowest model level at 1200, 1800, and 2400 LST 23 March and at 0600 LST 24 March.

  • View in gallery

    Vertical cross section from southeast to northwest (perpendicular to the coastline) at 1200 LST 23 March (top) and vertical cross section from south to north at x = 15 km 2400 LST 23 March (bottom). The solid lines indicate the isolines of the SO2 concentration; the arrows indicate the component of the wind vector in the corresponding plain.

  • View in gallery

    Daily sum of SO2 deposition on 23 March 1995.

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 12 12 4
PDF Downloads 12 12 8

The Impact of Secondary Flow Systems on Air Pollution in the Area of São Paulo

View More View Less
  • 1 Institut für Meteorologie und Klimaforschung, Forschungszentrum Karlsruhe/Universität Karlsruhe, Karlsruhe, Germany
© Get Permissions
Full access

Abstract

The area between the Atlantic Ocean and São Paulo is highly polluted due to high emission rates at Cubatão, a city situated 15 km inland at a steep slope. It was expected that secondary circulations would develop caused by the land–sea contrast and strong orographic changes, which influence the transport and diffusion of air pollutants. In 1994–95, surface stations were operated and radiosonde ascents were performed to analyze the characteristic features of the land–sea-breeze circulation.

The stations make evident a land–sea-breeze system that arrived in the suburbs of São Paulo in the early afternoon. The upslope winds favor the propagation of the sea breeze at the steep slope. During the measurement period, large-scale northwesterly winds prevailed that advected warm air from the plateau to the coastal area in the afternoon and resulted in a limitation of the boundary layer growth. The data were used to initialize a three-dimensional mesoscale model for calculation of the transport and deposition of SO2 emitted at Cubatão. The boundary layer height was found to be a limitation for vertical mixing of the air pollutants. However, a step between the coastal boundary layer and the boundary layer over the plateau causes SO2 to be vented into the free atmosphere at the slope and then transported toward the Atlantic Ocean with the large-scale northwesterly winds. Thus, over the coastal area, the SO2 concentrations in the free atmosphere were even higher than within the mixed layer. The deposition, summed up over a day, was calculated and found to be strongest at the slope and over the Atlantic Ocean.

Corresponding author address: Dr. N. Kalthoff, Institut für Meteorologie und Klimaforschung, Forschungszentrum/Universität Karlsruhe, Postfach 3640, D-76021 Karlsruhe, Germany.

Norbert.kalthoff@imk.fzk.de

Abstract

The area between the Atlantic Ocean and São Paulo is highly polluted due to high emission rates at Cubatão, a city situated 15 km inland at a steep slope. It was expected that secondary circulations would develop caused by the land–sea contrast and strong orographic changes, which influence the transport and diffusion of air pollutants. In 1994–95, surface stations were operated and radiosonde ascents were performed to analyze the characteristic features of the land–sea-breeze circulation.

The stations make evident a land–sea-breeze system that arrived in the suburbs of São Paulo in the early afternoon. The upslope winds favor the propagation of the sea breeze at the steep slope. During the measurement period, large-scale northwesterly winds prevailed that advected warm air from the plateau to the coastal area in the afternoon and resulted in a limitation of the boundary layer growth. The data were used to initialize a three-dimensional mesoscale model for calculation of the transport and deposition of SO2 emitted at Cubatão. The boundary layer height was found to be a limitation for vertical mixing of the air pollutants. However, a step between the coastal boundary layer and the boundary layer over the plateau causes SO2 to be vented into the free atmosphere at the slope and then transported toward the Atlantic Ocean with the large-scale northwesterly winds. Thus, over the coastal area, the SO2 concentrations in the free atmosphere were even higher than within the mixed layer. The deposition, summed up over a day, was calculated and found to be strongest at the slope and over the Atlantic Ocean.

Corresponding author address: Dr. N. Kalthoff, Institut für Meteorologie und Klimaforschung, Forschungszentrum/Universität Karlsruhe, Postfach 3640, D-76021 Karlsruhe, Germany.

Norbert.kalthoff@imk.fzk.de

Introduction

The transport, dispersion, and deposition of air pollutants is strongly influenced by meteorological conditions (Simpson 1994). The meteorological conditions, however, are a combination of large-scale flow and mesoscale or local circulation systems, such as land and sea breeze, slope and valley winds, and plain and plateau winds (e.g., Kurita et al. 1990; Savijarvi 1995; Lu and Turco 1994, 1995). The latter are caused by nonhomogeneous surface conditions such as differences in spatial surface temperature or orography. Often, the conditions for the evolution of more than one mesoscale circulation system exist at one site. So, merging of the secondary circulation systems may lead to long-range transport of air masses. And, as in many cases, large industrial accumulations are situated in coastal regions; secondary circulation systems play an important role in air pollutant transport. This was verified by observations by Kurita et al. (1990) for the Tokyo Bay area, by Wakimoto and McElroy (1986) for the Los Angeles Basin, and by Carroll and Baskett (1979) for the San Francisco area. The effect of the land–sea breeze on the air pollutant transport in a coastal area was, for example, modeled by Koo and Reible (1995), stressing out the effect of the return flow on air pollutant transport toward the sea. Beyond that, long-range transport leads to vegetation damage in areas far away from the emission sources; steep slopes near the coast may lead to pumping of pollutants from the boundary layer to the free atmosphere (chimney effect), as observed by Wakimoto and McElroy (1986). So, after venting of the air pollutants to elevated layers up the slope, backward advection of air pollutants toward the coastal areas was observed (McElroy and Smith 1991).

The eastern coast of Brazil, between the Atlantic Ocean and São Paulo, is characterized by both differences in surface temperature and a complex structure of the terrain and thus offers the conditions for the evolution of the secondary circulation systems mentioned above. Adjacent to the Atlantic Ocean, there is a coastal area of about 15 km wide, limited by the steep slope called Serra do Mar with an incline of about 700 m within 3 km. The slope ends in a plateau where São Paulo is situated, approximately 40 km inland. This complex terrain structure results in a variety of meteorological phenomena. First, the large-scale circulation system is connected to the southeast Passat that shows a strong subsidence inversion limiting vertical mixing. A second important flow system is the land–sea-breeze circulation, which is caused by the differences of thetemperature of the Atlantic Ocean and the land. Additionally, the steep slope gives rise to the evolution of katabatic downslope and anabatic upslope winds. Finally, the surfaces where the energy exchange takes place, that is, the plateau area and the coastal site, differ in height, which results in a horizontal temperature gradient and leads to advection effects in the coastal area under northwesterly flow conditions. All secondary circulation systems interfere with and enhance or diminish each other.

Within this complex terrain structure, the city of Cubatão is situated at the bottom of the steep slope of the Serra do Mar with high emission rates resulting from industrial activities. It is apparent that the location of the chemical industry in the area of Cubatão is not favorable to strong ventilation and intensive mixing of air pollutants. Thus, the conditions are comparable to those found in San Francisco, Los Angeles, and Tokyo.

It is the aim of this study to 1) investigate the secondary circulation systems detected in this area and 2) study by model simulations the transport, diffusion, and deposition of the released material and to demonstrate the loads on the ecosystems. Especially concerning the secondary circulation systems, the following questions arise: When is the onset of the sea breeze at the coast, what is the speed of propagation, and how far does the sea breeze move inland?

Due to the fact that in most cases the sea breeze has only a vertical extension of about 500 m associated with a return flow aloft (Atkinson 1981; Munn 1966), a sharp change of terrain height of about 700 m may block the sea-breeze front completely or may delay the inward penetration for several hours. This blocking effect was described by Neumann and Savijarvi (1986). On the other hand, mountain and valley winds (Geiger 1942) at the terrain barrier may trigger, both at day and night, the onset of the thermal land–sea-breeze flow and they may merge (Kurita et al. 1990; Silva Dias et al. 1995). Thus, the following questions arise: What is the vertical extension of the land–sea-breeze system, and are the sea-breeze winds able to overcome the steep terrain step?

Next, a site description and the experimental setup will be given. In section 3, observed data will be presented, both from permanent stations and from short-term measurements. Section 4 contains model results obtained with the Karlsruher nonhydrostatic mesoscale model (KAMM) (Adrian and Fiedler 1991) simulating the three-dimensional structure of the land–sea breeze and the related transport and deposition of the air pollutants in the area between the Atlantic Ocean and São Paulo.

Site description and experimental setup

The area investigated covers the zone between the Atlantic Ocean and the city of São Paulo. This area (Fig. 1) consists of the ocean, a flat coastal zone about15 km wide ending with a steep slope called Serra do Mar, with an increase in terrain height of 700 m within 3 km leading to the plateau area where São Paulo is situated, which is about 40 km inland. The orientation of the coastline and the slope is from southwest to northeast. The city of Cubatão, a rather large industrial agglomeration, is situated at the bottom of the Serra do Mar. Figure 2 shows the configuration of land use. In the coastal zone, sealed, forested, and agriculturally used areas can be found, which are interrupted by rivers and lakes in the surroundings of Santos. The slope as well as the plateau area between the slope and São Paulo are wooded. Also, several drinking water lakes for São Paulo can be found on the plateau. Figure 3 gives the annual emission rates for 17 of the most important SO2 emission sources at Cubatão; the sources are situated along the bottom of the Serra do Mar.

To study the characteristic features of the land–sea breezes, two permanent meteorological stations were installed in October 1994. The locations of the stations were selected allowing for the detection of the onset of the sea breeze near the coast and calculating the propagation speed of the sea breeze. The first station was installed near the coast at Vicente de Carvalho (23°55′S, 46°18′W) and the second at São Bernardo (23°42′S, 46°34′W), a suburb of São Paulo. These two stationslie on a line perpendicular to the coast (Fig. 1), 39 km apart from each other. A third station was run by the University of Frankfurt at Barragem das Pedras on top of the Serra do Mar to detect the temporal behavior of the up- and downslope winds. At Vicente de Carvalho and São Bernardo, the temperature and humidity at 2- and 10-m heights, the wind speed and direction at 10-m height, global and reflected radiations at 2-m height, and precipitation at 1-m height were measured. The sensors were scanned every 10 s, and the data were aggregated to 10-min mean values. At Barragem das Pedras, the meteorological parameters wind speed; wind direction; temperature; humidity at one altitude; and the chemical species SO2, NO2, NO, and O3 were measured, also on the basis of 10-min means.

For the analysis of the vertical structure of the land–sea-breeze circulation, an aerological station was installed at Vicente de Carvalho in March 1995. The radiosonde delivers temperature, humidity, air pressure, wind speed, and wind direction every 2 s. The wind speed and wind direction were calculated by a running mean over 30 s. The ascent rate of the radiosonde was selected to be within 2 to 3 m s−1. Radiosonde ascents were performed during convective weather conditions because under these conditions the development of a land–sea-breeze circulation could be expected and investigated. Ascents were made every 1.5 h. In March 1995, measurements were successively carried out during two periods of 63 h (16–18 March 1995) and 40 h (22–24 March 1995).

Observed results of the land–sea-breeze circulation

To get an overview of the characteristic features of the land–sea-breeze system, the mean behavior of the wind at the surface station was analyzed (section 3a). Data collected on 17–18 and 23–24 March 1995 were chosen to study the diurnal cycle and vertical structure of the land–sea-breeze circulation system (section 3b). The two periods were also the subject of the modeling activities performed to calculate transport and diffusion with land–sea-breeze conditions.

Mean behavior of the land–sea-breeze circulation

The results of 5 months of continuous measurements, from November 1994 to March 1995, were available at São Bernardo and Vicente de Carvalho and allowed the mean behavior of the land–sea-breeze circulation in southern summer to be investigated. The wind rose of São Bernardo shows that southeasterly winds (28%) dominate; that is, winds blew from sea to land. During 14% of the time, northwesterly winds occurred; that is, winds blew from land to sea. At Vicente de Carvalho, southeasterly winds were also dominant (16%), while a second maximum was found for northerly winds (12%). This means that at both stations, two wind directions prevail approximately perpendicular to the coast line.

An insight into the daily cycle and strength of the land–sea-breeze circulation can be gained from the hodographs (Fig. 4). As in former observations (Lyons 1972; Burt et al. 1974; Anto 1977) and theoretical studies (Haurwitz 1947; Schmidt 1947), an elliptic shape was detected at Vicente de Carvalho. Calm northerly winds of less than 1 m s−1 prevailed during the night. The land breeze turned to sea breeze between 0900 and 1000 LST. At 1500 LST, the sea breeze was directed perpendicular to the coast with a strength of more than 2 m s−1. The land breeze again set in after 2300 LST. A different hodograph was found at São Bernardo (Fig. 4). The onset of the sea breeze occurred at 1330 LST, that is, about 4 h later than at the coast. This gives a mean propagation speed of the sea breeze of about 10 km h−1 or 3 m s−1. At São Bernardo, the maximum wind speed during sea breeze with values of 1.6 m s−1 was observed at 1800 LST. The land breeze starts in the morning hours between 0600 and 0700 LST. A similar behavior was observed by Silva Dias et al. (1995) who found that the sea breeze in summer arrived at the city of São Paulo between 1400 and 1500 LST. From these data it can generally be concluded that the sea breeze overcomes the steep mountain step and enters the area of São Paulo. Therefore, it can be expected that polluted air from the bay area will be transported to the area of São Paulo.

Diurnal cycle of the land–sea breeze on 17 and 23 March 1995

Two days were selected in March 1995. Although the detail structure was complex, some overall characteristic features of boundary layer evolution should be emphasized. The first period was characterized by calm, the second by strong synoptic westerly winds, which at the surface resulted in a different propagation speed of the sea breeze, and at the coast in a different warm-air advection above the sea-breeze layer.

Figure 5 shows the horizontal wind vectors and the temperature field on 17 March 1995. There was an upper inversion at about 2800 m that slightly increased during the day. At night, a surface inversion was observed, with the strongest gradient in the lowest 300 m. In the morning, at approximately 0800 LST, the surface inversion began to dissolve, and at 1500 LST, when the mixed layer had fully developed, the top of the mixed layer had reached a height of 800 m. Compared to the temperature at night, the temperature not only increased in the boundary layer but also above the inversion. This can be explained by the advection of warm air with northwesterly winds from the plateau. The warm-air advection above the inversion restricted the boundary layer growth and hence influenced vertical mixing of air pollutants. During the day, there was no significant increase in temperature in the mixed layer; thus, heating of the mixed layer by divergence of the net radiation and the sensible heat flux were almost compensated by the advection of cold air from the sea. The inversion strength during the day increased to ΔT = 3 K. The top of the sea-breeze flow nearly corresponded to the top of the mixed layer. Within the mixed layer, the southeasterly sea-breeze winds reached maximum values of up to 6.5 m s−1 at 1330 LST. At the top of the inversion, the wind speed dropped to 1 m s−1, and the wind changed to a northerly flow. In the night, the wind profile was threefold with a weak land-breeze layer up to the top of the surface inversion, that is, about 400 m, a layer with easterly winds above this up to a height of 1200 m, and westerly winds above. Most striking were the strong easterly winds during the night in the intermediate layer that were observed each night during the two measurement periods. This layer with easterly winds, blowing from sea to land, is not yet completely understood. The layer probably consists of the remaining part of the sea-breeze system and of the return current of the land breeze because the wind speed was too strong to be the return current of the land breeze, which was predicted from theorectical calculations and observations (e.g., Schmidt 1947). However, a similar threefold structure was also observed by Kurita et al. (1990) during the night.

Figure 6 shows the wind direction measured at the three surface stations on 17 March 1995, reflecting the cyclic behavior of the land–sea breeze. At Vicente de Carvalho, at approximately 0900 LST, the wind began to turn from the northerly land winds to southwesterly sea winds, and at 1000 LST, the change was completed. The larger scattering in the wind direction during the night compared to the scatter during daytime results from the fact that the land breeze is much weaker (1 m s−1) than the sea breeze (3 m s−1). In the course of the day, the wind blew from the south. Later on, at 1900 LST, the wind began to turn again, and 1 h later, a land–wind system had developed. At São Bernardo, the wind change from land to sea wind took place at approximately 1300 LST, whereas the reversal occurred at around 2100 LST. This resulted in a propagation speed of the sea breeze of about 3 m s−1 and a duration of the sea breeze of 8 h at the coast and 7 h at the plateau. The onset of the sea breeze can also be detected from other parameters. The change of the wind direction coincided with an increase in wind velocity (Fig. 6). Simultaneously, at São Bernardo, the decrease in the relative humidity stopped, and a sharp increase in humidity of about 15%–20% at 1400 LST clearly demonstrates that wet air from the sea had reached the area of São Paulo (not shown). No indication of the sea breeze was found from the temperature signal at the surface. At Barragem das Pedras, the station at the top of the slope, two things are notable. First, there was a strong katabatic downslope wind during the night. At 0900 LST, the wind turned into southeasterly upslope winds at about thesame time the sea breeze at the coast set in. Comparable observations were made by Kurita et al. (1990). The beginning of the upslope wind was accompanied by a significant reduction of the wind speed. Second, the onset of the upslope wind was correlated with a sharp increase in the SO2 concentration up to values of 130 ppb. This means that SO2 was advected from the area of Cubatão, where a high concentration had accumulated during the night. A few hours later, the SO2 concentration dropped, which was caused by enhanced vertical mixing and an increase in the boundary layer height (Fig. 5). A second SO2 maximum with values of 95 ppb was detected in the early evening. At 1800 LST, global radiation became zero, and the surface layer adopted a stable stratification (Fig. 5). However, the wind still blew from the sea. Thus, as vertical mixing was suppressed, high concentrations again occurred in the area of Cubatão and were then transported to the plateau until the wind changed to downslope winds at 2200 LST. In comparison with the mean behavior of the sea breeze calculated from values collected during 5 summer months, this day was very typical.

The second case study of the diurnal cycle of boundary layer evolution was taken from the second measurement period (23 March 1995). Again, large-scale but stronger westerly winds were observed. The upper inversion was at approximately 2400 m with less change in time during the day (Fig. 7). At 1330 LST, the mixed layer had fully developed; however, with a height of zi = 350 m, the vertical extension was much less than in the first case study. This low mixed layer resulted from a strong advection of warm air in a layer between 400 and 2000 m by northwesterly winds from the plateau. This resulted in a temperature increase of about 4 K within 5 h, suppressing a further increase in the mixed layer and leading to an enhancement of the inversion strength. This feature has an important impact on the diffusion of the air pollutants because vertical mixing is surpressed and limited to a much smaller height than is normally expected under horizontally homogeneous conditions. In the evening, the sea breeze disappeared;up to 2230 LST, a land breeze of zi = 250-m extension had developed. Again, between the land-breeze layer and the large-scale wind, a strong easterly flow prevailed; that is, the sea-breeze winds survived at higher levels.

The surface observations on 23 March are shown in Fig. 8. On the coast, the sea breeze again set in at 0930 LST. However, there was a delay at Barragem das Pedras and São Bernardo as compared to 17 March. At Barragem das Pedras, easterly winds started at 1200 LST. This resulted in an inland propagation speed of 2 m s−1 of the sea breeze between Vicente de Carvalho and Barragem das Pedras. In São Bernardo, a change in the wind direction took place at 1900 LST. However, this could not clearly be related to the sea breeze. This again agrees with observations from Silva Dias (1995) who found that the inland propagation of the sea breeze was slowed down on days with strong westerly winds. Figure 8 again shows the SO2 concentration maxima at Barragem das Pedras in the morning and evening. They aremuch lower than those on 17 March, which was probably due to the reduced emission rates.

Simulation of the land–sea-breeze circulation and deposition of SO2

Model description and initialization

The model used here has been described by Adrian and Fiedler (1991) and consists of four components:

  • an atmosphere model (KAMM)
  • a soil-vegetation model
  • the transport and dispersion model (DRAIS)
  • the chemical model Regional Acid Deposition Model (RADM2) (Chang et al. 1987; Stockwell et al. 1990)

The nonhydrostatic KAMM model is based on the equations of motion, the continuity equation, the first law of thermodynamics, and an equation for the specific humidity. The whole model system needs as input data information from the large-scale synoptic weather situation as well as the topography and land use and emission data. The larger-scale flow is assumed to be geostrophic and hydrostatic and is described by the components of the geostrophic wind uG and υG, the large-scale field of the potential temperature ΘG, and πG (π:Exner function), representing the large-scale pressure field. The model’s output are space- and time-dependent distributions of wind (u, υ, w), potential temperature Θ, specific humidity s, concentrations, and deposition. The following equations of motion for the horizontal and vertical velocity components u, υ, and w are applied in KAMM:
i1520-0450-37-3-269-e1
The variable π describes the deviations from the large-scale flow due to the mesoscale processes, for example, due to the influence of the orography. The mesoscale pressure field is subdivided into a dynamic part πd and a thermal part πh. The equation of heat is used as an equation for the potential temperature Θ and degenerates to a transport equation if diabatic processes are excluded:
i1520-0450-37-3-269-e4
The equation for the specific humidity s is as follows:
i1520-0450-37-3-269-e5

The unprimed variables denote the mesoscale mean motion, and the barred variables denote the turbulent fluxes. The prognostic equations must be closed by parameterizations of the Reynolds stresses. Within the KAMM model system, they are formulated using the eddy diffusivity concept. The turbulent stress tensor is assumed to be proportional to the deformation tensor of the mean flow, where the proportionality factor is theeddy diffusion coefficient. Under stable atmospheric conditions, the eddy diffusion coefficient is determined by profile functions taken from Businger et al. (1971), which depend on the local Richardson number. In case of an unstable stratified atmosphere, a nonlocal closure scheme of Degrazia (1988) is applied. In this case, the turbulent diffusion coefficient is a function of the boundary layer height which is determined by a prognostic equation. The equations are transformed into a coordinate system, which follows the topography. The soil-vegetation model (Schädler et al. 1990), which has been integrated into the KAMM system, uses a concept similar to that of Deardorff (1978); that is, the canopy is considered as one “big leaf” placed between the soil surface and the lower atmosphere.

The model DRAIS (three-dimensional regional diffusion simulation model), which has also been integrated in the KAMM model (Baer and Nester 1992), simulates the transport, dispersion, and deposition of air pollutants for given source positions and strengths. The following diffusion equation is solved:
i1520-0450-37-3-269-e6
where ci stands for the concentration of the different species, K denotes the tensor of the diffusion coefficients, and Qi is the source term of the species. The values of Pi and Li are the rates of production and lossby chemical reactions that are determined using the RADM2 mechanism. The lower boundary conditions applied in (6) include dry deposition with the dry deposition rate determined by the pollutant deposition velocity υD. The flux of material of the different species F, directed toward the lower boundary surface, is defined by
FυDczr
where c(zr) is the concentration at a reference height zr. The dry deposition velocity involves a complex linkage between turbulent diffusion in the surface boundary layer, molecular scale motion at the air–ground interface, and interaction of the material with the surface. To parameterize the velocity of dry deposition of gases to vegetated surfaces, a “big leaf” multiple resistance model is used. The most important individual resistances to pollutant transfer are the aerodynamic resistance ra, associated with atmospheric turbulence; a quasi-laminar boundary resistance rb, which takes into account the resistance to mass transfer through the quasi-laminar layer of air in contact with the surface elements and is influenced by the diffusivity of the material being transferred; and a net canopy resistance rc, which is dominated by biological surface factors and includes the stomatal, mesophyll, and cuticular uptake resistances. Thus, υD is determined by
i1520-0450-37-3-269-e8

To run the simulation model, topography and land-use data have to be provided as well as meteorological and emission data. The topography and land-use data were obtained by digitizing a map provided by Companhia de Tecnologia de Saneamento Ambiental (CETESB) (1985) covering an area of 40 km (east–west)by 74 km (north–south). The resolution of the data is 500 m × 500 m, and it was transferred to a 1 km × 1 km grid, which is presently the horizontal grid size of the numerical model. In the vertical direction, a much denser grid size is used with separations of approximately 20 m close to the ground and approximately 400 m in the upper levels. The model domain in the vertical extends up to a height of 4 km. The land-use data were also taken from CETESB (1985). Six classes of vegetation type were defined: water, marsh land, open grassland, agricultural land, forest, and urban areas (Fig. 2). As far as soil type is concerned, four separate classes are distinguished: water, peat soil, sandy loam, and urban areas. A set of parameters are associated with theland-use classes: leaf area index (LAI), vegetation cover σf, upper limit for the stomatal resistance Ro, depth of the root zone zroot, displacement height d, roughness length zo, vegetation albedo αf, vegetation emissivity εf, and cveg = uaf/u*, the ratio of the velocity uaf in the canopy to the friction velocity u*. The vegetation parameters employed in the simulation model are summarized in Table 1.

To study conditions of relevance to the dispersion conditions in the Cubatão area, the numerical model needs observation data for initialization. These input data comprise horizontal pressure gradients on the synoptical scale. In addition, information on the vertical variation of temperature, humidity, and wind have to beprovided. The necessary input data concerning the large-scale pressure gradient, vertical profiles of wind, temperature, and humidity, were taken from the 0900 LST radiosounding of Vicente de Carvalho and interpolated to the whole area. To initialize the soil-vegetation model, the sea surface temperature and vertical profiles of temperature and water content in soil are required. In some cases, the sea surface temperature was provided by the Department of Atmospheric Sciences of the University of São Paulo (USP), which also provided profiles of the soil temperature measured at a USP monitoring station located on the plateau east of São Paulo. If no data were available, the initial data for the soil model were taken from climatological data. For the subsequent calculations, the sea surface temperature is set to 27°C. To obtain reliable results by simulating the transport, diffusion, chemical reactions, and deposition of pollutants released in the Cubatão area, one of themost crucial prerequisites is a detailed emission inventory as a function of time and space. As this detailed source information was missing, the data were taken from CETESB (1987, 1992) and CETESB (1994, personal communication). The reports include yearly sums of the emissions of CO2, SO2, NOx, HC, aerosol, and CO from different sources. So far, only the transport, diffusion, and dry deposition of SO2 has been simulated. The SO2 source data used—only main emissions are considered—are shown in Fig. 3. Due to the fact that only yearly sums of the emissions were available, the source strength did not vary in time in the numerical simulations but was kept constant during the whole day.

Model simulations of 17 and 23 March 1995

The results of the model simulation have been divided into two parts. First, model simulations of the meteorological conditions have been compared with observations from Vicente de Carvalho, where both surface and boundary layer observations are available. The simulations for SO2 were compared with measurements from Barragem das Pedras, where pollutant data are available. Then air pollution modeling for SO2 has been carried out for the whole area. The simulations were carried out on 17 and 23 March 1995.

Figure 9 shows the time series of modeled and observed wind directions and speeds at Vicente de Carvalho as well as the SO2 concentration for Barragem das Pedras on 17–18 March. In both cases, southerly winds developed in the morning. The sea breeze turned into the land breeze at 2100 LST. The modeled diurnal cycle of the land–sea-breeze circulation fits well with the observations. The wind speed was higher during the sea-breeze period and lower during the land-breeze period. The modeled onset of the downslope winds at Barragem das Pedras is about 1 h earlier than observed, and hence, the SO2 maximum is also predicted earlier by the simulations as well as higher values on the second day. The comparison of the vertical profiles of potential temperature and wind direction for Vicente de Carvalho is given in Fig. 10. Although the general features of the temperature andwind direction profiles are well reproduced, the mixed-layer height at 1200 LST, and thus the height of the sea breeze, are overestimated by the model. At night, the main difference between the observed and modeled temperature profiles is restricted to the surface layer, that is, the model results show a more stable layer up to a height of about 200 m. The profile of the wind direction reveals the threefold structure of the atmosphere, that is, northeasterly winds at the surface, easterly winds in the residual layer, and westerly winds above.

The modeled surface wind field of a 35 km × 65 km subarea of the whole model domain is shown in Fig. 11. The diagram includes the wind field at four different times: 1200 LST, 1800 LST, and 2400 LST 17 March, and 0600 LST 18 March. At 1200, the sea breeze has propagated about 25 km inland. A propagation speed of 3.5 m s−1 was calculated from the model and fits with the observed value of 3 m s−1. The sea breeze stillcontinued at 1800 LST. At that time, a change in the wind direction on the slope could be detected. At 2400, the land breeze had propagated into the Atlantic Ocean. The strongest downslope winds were found at the Serra do Mar, which was also confirmed by the observations (Fig. 6). A similar flow field existed at 0600 LST. Two vertical cross sections were chosen perpendicular to the coast, that is, from southeast to northwest through Vicente de Carvalho for 1200 LST and 2400 LST 17 March (Fig. 12). The sea breeze in the coastal area covered the whole mixed layer up to 1000 m, as can be seen from the profiles (Fig. 10). Over the plateau, the sea breeze had a smaller vertical extension. At night, a small and weak land-breeze layer developed up to a height of 200 m. The highest wind speed developed on the slope of the Serra do Mar, as can be seen from Fig. 11 and from observations (Fig. 9). Above the land-breeze layer, southeasterly winds between 200 and1500-m height blew at both the coastal and the plateau sites. Above this layer, the northwesterly synoptic-scale flow was observed.

Corresponding to this daily variation of the flow field, the transport of SO2 altered during the day. Figure 13 shows the near-surface SO2 concentration field, about 10 m above ground. At 1200, SO2 is transported with the southeasterly flow of the sea breeze from the source areas of Cubatão to the northwest. The highest concentrations were found on the slope. However, due to enhanced vertical mixing, surface concentration values were moderate. In the evening, when convection was suppressed, but southeasterly winds still prevailed, higher near-surface concentrations were detected on the slope and the plateau. This agrees with measurements at Barragem das Pedras, where the highest SO2 concentrations were found in the evening and after the onset of the land breeze (Fig. 6). The next morning, at 0600 LST, the highest SO2 concentrations, transported by the land breeze, existed over the Atlantic Ocean. Vertical cross sections of the concentration field are shown in Fig. 14. The cross section at 1200 LST corresponds to the vertical cut in the temperature field, that is, perpendicular to the coast. The cross section at 2400 LST is from south to north at x = 15 km through the maximum of the surface concentration over the Atlantic Ocean. At 1200 LST, SO2 was transported up the Serra do Mar.The inland propagation of the SO2 front is clearly correlated with the sea-breeze front. Additionally, vertical diffusion within the mixed layer over the plateau up to 2800-m height led—together with northwesterly winds—to a southeasterly transport of SO2. This means that SO2 was vented from the mixed layer over the plateau into the free atmosphere over the coastal zone. Hence, over the ocean, the SO2 concentrations over the mixed layer were higher than within the mixed layer itself. A similar process has already been described by Kurita et al. (1990), Edinger et al. (1972), and Wakimoto and McElroy (1986). During the night, most of the SO2 is trapped within the nocturnal surface layer of the coastal zone and the ocean. Over the Atlantic Ocean, a second, elevated SO2 maximum between 1700 and 2700 m still exists. The deposition accumulated over a day is given in Fig. 15. There are two main areas where deposited SO2 can be distinguished. These are the slope areas north of the emission sources and the Atlantic Ocean and lakes in the south of Cubatão. Maximum deposition values accumulated over the day on the slope are 30 mg m−2 and 16 mg m−2 over the Atlantic Ocean.

The second case study was carried out for 23 March 1995. Two main differences were detected in comparison to 17 March 1995. On account of the higher large-scale northwesterly winds, the propagation of the sea-breeze front was delayed. Second, in the afternoon, when the boundary layer over the plateau heated up, thenorthwesterly winds led to warm-air advection and the accompanying reduction of the mixed-layer height to about 400 m. Both facts are important in the transport of air pollutants from the Cubatão area to São Paulo.

Figure 16 shows the comparison of the simulated and observed wind directions at Vicente de Carvalho and the SO2 concentration for Barragem das Pedras. Again, the onset of the sea breeze as well as the onset of the land breeze are well reproduced. However, the simulated SO2 concentration is much higher than observed, probably due to lower emissions from Cubatão. Figure 17 shows the comparison of the profiles of potential temperature and wind direction for the same site for 1200 LST and 2400 LST 23 March. Again, the agreement at both times was satisfactory. The reduction of propagation speed of the sea breeze can be seen in Fig. 18. At 1200 LST, the sea breeze just arrived at the top ofthe Serra do Mar, while on 17 March the sea breeze was about 10 km farther inland. The speed of propagation between Vicente de Carvalho and Barragem das Pedras was calculated to be 2 m s−1, that is, much less than on 17 March. The corresponding SO2 concentrations at lower levels are shown in Fig. 19. Highest values were found on the slope of the Serra do Mar. In the early morning, no trace of SO2 was found over the Atlantic Ocean. This was due to the northeasterly land winds transporting the SO2 plume out of the modeling domain. The strong influence of warm-air advection on the mixed-layer height can be seen in Fig. 20. At 1200 LST, the mixed-layer top was about 500 m and vertical mixing of SO2 was limited to that height. However, due to the northwesterly winds at upper levels and the influence of the orography on the air pollutant transport, high SO2 concentrations were also observed above the top of thecoastal inversion. This was a result of sea-breeze winds transporting the SO2 up the slope of Serra do Mar to the plateau. At the top of the Serra do Mar, this air mass was already vented to higher levels than the top of the mixed layer over the coast. The large-scale northwesterly winds on the plateau transported the SO2 through the step between the boundary layer over the coast and over the plateau back to the coastal site. Thus, air pollutants were transferred from the boundary layer to the free atmosphere in this process and could be regarded as an effective exchange over heterogenous surfaces. Finally, the daily sum of the SO2 deposition is indicated in Fig. 21. High deposition rates again appeared on the slope of the Serra do Mar.

Conclusions

The meteorological conditions in the area of Cubatão are strongly influenced by secondary circulation systemslike the land–sea breeze and up- and downslope winds. The aim of this study was to investigate the influence of the secondary circulation systems on the transport and deposition of SO2 emitted from Cubatão. Model simulations were performed and observations were made to analyze the evolution of the land and sea breeze in space and time. The main results are as follows.

  • A land–sea-breeze system was found from the 5-month average of the southern summer. The sea breeze passed the coast at 0930 LST and arrived in the area of São Paulo about 4 h later. The propagation speed of the sea-breeze front is about 3 m s−1.
  • The vertical extension of the sea-breeze layer was about 800 m at the coastal site and less over the plateau. The nocturnal land breeze reached up to about 400 m.
  • Apart from the land–sea breeze, up- and downslope winds developed at the slope of the Serra do Mar. The upslope winds appeared at the coast at about the same time as the sea breeze. These upslope winds support the propagation of the sea-breeze front to overcome the steep slope and penetrate into the area of São Paulo.
  • Northwesterly synoptic winds during the measurement period in March 1995 resulted in advection of warm air from the plateau to the coast. This led to low mixed-layer heights at the coast (about 400 m during the second measurement period). This low mixed-layer height was of considerable significance for the vertical diffusion of air pollutants.
  • These upper-layer northwesterly winds also had another effect. During the day, SO2 in the lower layers was transported with the sea breeze from the source areas of Cubatão on the plateau. Then, with these northwesterly winds at the plateau, SO2 is transported back to the coastal area. As the plateau was higher than the mixed layer over the coast, the air pollutants were transferred from the boundary layer into the free atmosphere. Thus, over the ocean, the SO2 concentration within the free atmosphere was higher than within the boundary layer.
  • Due to the land–sea breeze and the two accompanying dominating southeasterly and northwesterly wind directions, deposition of SO2 took place mainly on the slope of the Serra do Mar and into the Atlantic Ocean. As the sea breeze penetrated into the area of São Paulo during the day, air pollutants were also transported from Cubatão to São Paulo.

Acknowledgments

We would like to thank A. Wenzel, U. Corsmeier, IMK, Professor Massambani and his group from the University of São Paulo for their help during the different measurements, and Professor Jaeschke and his group from the University of Frankfurt for providing data from Barragem das Pedras. The project was supported by the “Bundesministerium für Forschung und Technologie” under Contract BMFT 07INT05A within the frameworks of “Deutsch-brasilianische Zusammenarbeit in Forschung und Technologie BRA ENV 3/2.”

REFERENCES

  • Adrian, G., and F. Fiedler, 1991: Simulation of unstationary wind and temperature fields over complex terrain and comparison with observations. Beitr. Phys. Atmos.,64, 27–48.

  • Anto, A. F., 1977: Observational studies on land–sea-breeze phenomena around the island of Sylt. Meteor. Rundsch.,30, 118–122.

  • Atkinson, B. W., 1981: Meso-scale Atmospheric Circulations. Academic Press, 495 pp.

  • Baer, M., and K. Nester, 1992: Parameterization of trace gas dry deposition velocities for a regional mesoscale diffusion model. Ann. Geophys.,10, 912–923.

  • Burt, W., H. Crew, N. Plutchak, and J. Dumon, 1974: Diurnal variation of winds over an upwelling region of Oregon. Bound.-Layer Meteor.,6, 35–45.

  • Businger, J. A., J. C. Wyngaard, Y. Izumi, and E. F. Bradley, 1971: Flux profile relationships in the atmospheric boundary layer. J. Atmos. Sci.,28, 181–189.

  • CETESB, 1985: Baixada santista, carta do meio ambiente e de sua dinimica. São Paulo, Brazil 33 pp. [Available from Frederico Hermann Jr., 345 Pinheiros, CEP 05489, São Paulo, Brazil.].

  • ——, 1987: Inventario de emissoes em fontes estacionarias de poluicao de poluicao do ar na regiao de Cubatão São Paulo, Brazil, 39 pp. [Available from Frederico Hermann Jr., 345 Pinheiros, CEP 05489, São Paulo, Brazil.].

  • ——, 1992: Acao da CETESB em Cubatão. Situacao em junho de 1992, São Paulo, Brazil, 40 pp. [Available from Frederico Hermann Jr., 345 Pinheiros, CEP 05489, São Paulo, Brazil.].

  • Carroll, J. J., and R. L. Baskett, 1979: Dependence of air quality in a remote location on local and mesoscale transports: A case study. J. Appl. Meteor.,18, 474–486.

  • Chang, J. S., R. A. Brost, I. S. A. Isaksen, S. Madronich, and P. Middleton, 1987: A three-dimensional Eulerian acid deposition model: Physical concepts and formulation. J. Geophys. Res.,92, 14681–14700.

  • Deardorff, J. W., 1978: Efficient prediction of ground surface temperature and moisture with inclusion of a layer of vegetation. J. Geophys. Res.,83, 1889–1903.

  • Degrazia, G. A., 1988: Anwendung von Ähnlichkeitsverfahren auf die turbulente Diffusion in der konvektiven und stabilen Grenzschicht. Ph.D. thesis, Universität Karlsruhe, 99 pp. [Available from Universität Karlsruhe, Kaiserstr. 12, 76131 Karlsruhe, Germany.].

  • Edinger, J. G., M. H. McCutchan, P. R. Miller, B. C. Ryan, M. J. Schroeder, and J. V. Behar, 1972: Penetration and duration of oxidant air pollution in the south coast air basin of California. J. Air Pollution Control Assoc.,22, 882–886.

  • Geiger, R., 1942: Das Klima der Bodennahen Luftschicht. Die Wissenschaft, 435 pp.

  • Haurwitz, B., 1947: Comments on the sea-breeze circulation. J. Meteor.,4, 1–8.

  • Koo, Y.-S., and D. D. Reible, 1995: Flow and transport modeling in a sea-breeze. Part II: Flow model application and pollutant transport. Bound.-Layer Meteor.,75, 209–234.

  • Kurita, H., H. Ueda, and S. Mitsumoto, 1990: Combination of local wind systems under gradient wind conditions and its contributionto the long-range transport of air pollutants. J. Appl. Meteor.,29, 331–348.

  • Lu, R., and R. P. Turco, 1994: Air pollutant transport in a coastal environment. Part I: Two-dimensional simulations of sea-breeze and mountain effects. J. Atmos. Sci.,51, 2285–2308.

  • ——, and ——, 1995: Air pollutant transport in a coastal environment–II: Three-dimensional simulations over Los Angeles Basin. Atmos. Environ.,29, 1499–1518.

  • Lyons, W. A., 1972: The climatology and prediction of the Chicago lake breeze. J. Appl. Meteor.,11, 1259–1270.

  • McElroy, J. L., and T. B. Smith, 1991: Lidar descriptions of mixing-layer thickness characteristics in a complex terrain/coastal environment. J. Appl. Meteor.,30, 585–597.

  • Munn, R. E., 1966: Descriptive Micrometeorology. Academic Press, 245 pp.

  • Neumann, J., and H. Savijarvi, 1986: The sea breeze on a steep coast. Beitr. Phys. Atmos.,59, 375–389.

  • Savijarvi, H., 1995: Sea breeze effects on large-scale atmospheric flow. Beitr. Phys. Atmos.,68, 335–344.

  • Schädler, G., N. Kalthoff, and F. Fiedler, 1990: Validation of a model for heat, mass and momentum exchange over vegetated surfaces using LOTREX-10E/HIBE88 data. Beitr. Phys. Atmos.,63, 85–100.

  • Schmidt, F. H., 1947: An elementary theory of the land- and sea-breeze circulation. J. Meteor.,4, 9–15.

  • Silva Dias, M. A. F., P. L. Vidale, and C. M. R. Blanko, 1995: Case study and numerical simulation of the summer regional circulation in São Paulo, Brazil. Bound.-Layer Meteor.,74, 371–388.

  • Simpson, J. E., 1994: Sea Breeze and Local Winds. Cambridge University Press, 234 pp.

  • Stockwell, W. R., P. Middelton, J. S. Chang, and X. Tang, 1990: The second generation regional acid deposition model—Chemical mechanism for regional air quality modeling. J. Geophys. Res.,95, 16343–16367.

  • Wakimoto, R. M., and J. L. McElroy, 1986: Lidar observation of elevated pollution layers over Los Angeles. J. Climate Appl. Meteor.,25, 1583–1599.

Fig. 1.
Fig. 1.

Map of the area around Cubatão with the three measuring stations: São Bernardo, Barragem des Pedras, and Vicente de Carvalho.

Citation: Journal of Applied Meteorology 37, 3; 10.1175/1520-0450-37.3.269

Fig. 2.
Fig. 2.

Land-use distribution in the area of Cubatão.

Citation: Journal of Applied Meteorology 37, 3; 10.1175/1520-0450-37.3.269

Fig. 3.
Fig. 3.

Annual SO2 emission data in the area of Cubatão.

Citation: Journal of Applied Meteorology 37, 3; 10.1175/1520-0450-37.3.269

Fig. 4.
Fig. 4.

Mean hodographs for the station of São Bernardo (dashed) and Vicente de Carvahlo (solid), as calculated from the November 1994–March 1995 dataset.

Citation: Journal of Applied Meteorology 37, 3; 10.1175/1520-0450-37.3.269

Fig. 5.
Fig. 5.

Isolines of potential temperature (bottom) and wind vectors (top) at Vicente de Carvalho on 17 March 1995. The shaded areas indicate sea-breeze winds.

Citation: Journal of Applied Meteorology 37, 3; 10.1175/1520-0450-37.3.269

Fig. 6.
Fig. 6.

Wind direction (top) (crosses: São Bernardo, triangles: Vicente de Carvalho, squares: Barragem des Pedras); wind speed (center) (dotted line: São Bernardo, dashed–dotted line: Vicente de Carvalho, solid line: Barragem des Pedras); and SO2 concentration (bottom) at Barragem des Pedras on 17 March 1995.

Citation: Journal of Applied Meteorology 37, 3; 10.1175/1520-0450-37.3.269

Fig. 7.
Fig. 7.

Isolines of potential temperature (bottom) and wind vectors (top) at Vicente de Carvalho on 23 March 1995. The shaded areas indicate sea-breeze winds.

Citation: Journal of Applied Meteorology 37, 3; 10.1175/1520-0450-37.3.269

Fig. 8.
Fig. 8.

Wind direction (top) (crosses: São Bernardo, triangles: Vicente de Carvalho, squares: Barragem das Pedras); wind speed (center) (dotted line: São Bernardo, dashed–dotted line: Vicente de Carvalho, solid line: Barragem des Pedras); and SO2 concentration (bottom) at Barragem des Pedras on 23 March 1995.

Citation: Journal of Applied Meteorology 37, 3; 10.1175/1520-0450-37.3.269

Fig. 9.
Fig. 9.

Wind direction (modeled: squares, observed: circles); wind speed at Vicente de Carvalho; and SO2 concentrations at Barragem das Pedras (modeled: dashed lines, observed: solid lines) on 17 March 1995.

Citation: Journal of Applied Meteorology 37, 3; 10.1175/1520-0450-37.3.269

Fig. 10.
Fig. 10.

Comparison of modeled and observed profiles of potential temperature (modeled: dashed lines, observed: solid lines) and wind direction (modeled: squares, observed: dots) at Vicente de Carvalho at 1200 (top) and 2400 LST (bottom) 17 March.

Citation: Journal of Applied Meteorology 37, 3; 10.1175/1520-0450-37.3.269

Fig. 11.
Fig. 11.

Surface wind fields at 1200, 1800, and 2400 LST 17 March and at 0600 LST 18 March.

Citation: Journal of Applied Meteorology 37, 3; 10.1175/1520-0450-37.3.269

Fig. 12.
Fig. 12.

Vertical cross sections from southeast to northwest (perpendicular to the coastline) at 1200 (top) and 2400 LST 17 March (bottom). The solid lines indicate the isolines of the potential temperature; the arrows indicate the component of the wind vector perpendicular to the coast.

Citation: Journal of Applied Meteorology 37, 3; 10.1175/1520-0450-37.3.269

Fig. 13.
Fig. 13.

The SO2 concentration fields are shown at the lowest model level at 1200, 1800, and 2400 LST 17 March and at 0600 LST 18 March.

Citation: Journal of Applied Meteorology 37, 3; 10.1175/1520-0450-37.3.269

Fig. 14.
Fig. 14.

Vertical cross section from southeast to northwest (perpendicular to the coastline) at 1200 LST 17 March (top), and vertical cross section from south to north at x = 15 km, 2400 LST 17, March (bottom). The solid lines indicate the isolines of the SO2 concentration; the arrows indicate the component of the wind vector in the corresponding plain.

Citation: Journal of Applied Meteorology 37, 3; 10.1175/1520-0450-37.3.269

Fig. 15.
Fig. 15.

Daily sum of SO2 deposition on 17 March 1995.

Citation: Journal of Applied Meteorology 37, 3; 10.1175/1520-0450-37.3.269

Fig. 16.
Fig. 16.

Wind direction (modeled: squares, observed: circles); wind speed at Vicente de Carvalho; and SO2 concentrations at Barragem des Pedras (modeled: dashed lines, observed: solid lines) on 23 March 1995.

Citation: Journal of Applied Meteorology 37, 3; 10.1175/1520-0450-37.3.269

Fig. 17.
Fig. 17.

Comparison of modeled and observed profiles of potential temperature (modeled: dashed lines, observed: solid lines) and wind direction (modeled: squares, observed: dots) at Vicente de Carvalho at 1200 (top) and 2400 LST (bottom) on 23 March.

Citation: Journal of Applied Meteorology 37, 3; 10.1175/1520-0450-37.3.269

Fig. 18.
Fig. 18.

Surface wind fields at 1200, 1800, and 2400 LST 23 March and at 0600 LST 24 March.

Citation: Journal of Applied Meteorology 37, 3; 10.1175/1520-0450-37.3.269

Fig. 19.
Fig. 19.

SO2 concentration fields at the lowest model level at 1200, 1800, and 2400 LST 23 March and at 0600 LST 24 March.

Citation: Journal of Applied Meteorology 37, 3; 10.1175/1520-0450-37.3.269

Fig. 20.
Fig. 20.

Vertical cross section from southeast to northwest (perpendicular to the coastline) at 1200 LST 23 March (top) and vertical cross section from south to north at x = 15 km 2400 LST 23 March (bottom). The solid lines indicate the isolines of the SO2 concentration; the arrows indicate the component of the wind vector in the corresponding plain.

Citation: Journal of Applied Meteorology 37, 3; 10.1175/1520-0450-37.3.269

Fig. 21.
Fig. 21.

Daily sum of SO2 deposition on 23 March 1995.

Citation: Journal of Applied Meteorology 37, 3; 10.1175/1520-0450-37.3.269

Table 1.

Vegetation parameters used in the simulations.

Table 1.
Save