Ozone pollution: What can we see from space? A case study
2014; Wiley; Volume: 119; Issue: 13 Linguagem: Inglês
10.1002/2013jd021340
ISSN2169-8996
AutoresGilles Forêt, Maxim Eremenko, Juan Cuesta, Pasquale Sellitto, J. Barré, Benjamin Gaubert, Adriana Coman, G. Dufour, Xiong Liu, Mathieu Joly, C. Doche, Matthias Beekmann,
Tópico(s)Atmospheric and Environmental Gas Dynamics
ResumoJournal of Geophysical Research: AtmospheresVolume 119, Issue 13 p. 8476-8499 Research ArticleFree Access Ozone pollution: What can we see from space? A case study G. Foret, Corresponding Author G. Foret Laboratoire Interuniversitaire des Systèmes Atmosphériques, UMR7583, IPSL, CNRS, Université Paris Est Créteil, Université Paris Diderot, Créteil, France Correspondence to: G. Foret, [email protected]Search for more papers by this authorM. Eremenko, M. Eremenko Laboratoire Interuniversitaire des Systèmes Atmosphériques, UMR7583, IPSL, CNRS, Université Paris Est Créteil, Université Paris Diderot, Créteil, FranceSearch for more papers by this authorJ. Cuesta, J. Cuesta Laboratoire Interuniversitaire des Systèmes Atmosphériques, UMR7583, IPSL, CNRS, Université Paris Est Créteil, Université Paris Diderot, Créteil, FranceSearch for more papers by this authorP. Sellitto, P. Sellitto Laboratoire Interuniversitaire des Systèmes Atmosphériques, UMR7583, IPSL, CNRS, Université Paris Est Créteil, Université Paris Diderot, Créteil, France Laboratoire de Météorologie Dynamique, UMR8539, IPSL, CNRS, École Normale Supérieure, Paris, FranceSearch for more papers by this authorJ. Barré, J. Barré ACD, NESL, National Center for Atmospheric Research, Boulder, Colorado, USA CNRM/GAME UMR 3589 CNRS/Météo-France, Toulouse, FranceSearch for more papers by this authorB. Gaubert, B. Gaubert Laboratoire Interuniversitaire des Systèmes Atmosphériques, UMR7583, IPSL, CNRS, Université Paris Est Créteil, Université Paris Diderot, Créteil, FranceSearch for more papers by this authorA. Coman, A. Coman Laboratoire Interuniversitaire des Systèmes Atmosphériques, UMR7583, IPSL, CNRS, Université Paris Est Créteil, Université Paris Diderot, Créteil, FranceSearch for more papers by this authorG. Dufour, G. Dufour Laboratoire Interuniversitaire des Systèmes Atmosphériques, UMR7583, IPSL, CNRS, Université Paris Est Créteil, Université Paris Diderot, Créteil, FranceSearch for more papers by this authorX. Liu, X. Liu Harvard-Smithsonian Center for Astrophysics, Cambridge, Massachusetts, USASearch for more papers by this authorM. Joly, M. Joly CNRM/GAME UMR 3589 CNRS/Météo-France, Toulouse, FranceSearch for more papers by this authorC. Doche, C. Doche Météo-France/DIRSO/DEC/FDF, Mérignac, FranceSearch for more papers by this authorM. Beekmann, M. Beekmann Laboratoire Interuniversitaire des Systèmes Atmosphériques, UMR7583, IPSL, CNRS, Université Paris Est Créteil, Université Paris Diderot, Créteil, FranceSearch for more papers by this author G. Foret, Corresponding Author G. Foret Laboratoire Interuniversitaire des Systèmes Atmosphériques, UMR7583, IPSL, CNRS, Université Paris Est Créteil, Université Paris Diderot, Créteil, France Correspondence to: G. Foret, [email protected]Search for more papers by this authorM. Eremenko, M. Eremenko Laboratoire Interuniversitaire des Systèmes Atmosphériques, UMR7583, IPSL, CNRS, Université Paris Est Créteil, Université Paris Diderot, Créteil, FranceSearch for more papers by this authorJ. Cuesta, J. Cuesta Laboratoire Interuniversitaire des Systèmes Atmosphériques, UMR7583, IPSL, CNRS, Université Paris Est Créteil, Université Paris Diderot, Créteil, FranceSearch for more papers by this authorP. Sellitto, P. Sellitto Laboratoire Interuniversitaire des Systèmes Atmosphériques, UMR7583, IPSL, CNRS, Université Paris Est Créteil, Université Paris Diderot, Créteil, France Laboratoire de Météorologie Dynamique, UMR8539, IPSL, CNRS, École Normale Supérieure, Paris, FranceSearch for more papers by this authorJ. Barré, J. Barré ACD, NESL, National Center for Atmospheric Research, Boulder, Colorado, USA CNRM/GAME UMR 3589 CNRS/Météo-France, Toulouse, FranceSearch for more papers by this authorB. Gaubert, B. Gaubert Laboratoire Interuniversitaire des Systèmes Atmosphériques, UMR7583, IPSL, CNRS, Université Paris Est Créteil, Université Paris Diderot, Créteil, FranceSearch for more papers by this authorA. Coman, A. Coman Laboratoire Interuniversitaire des Systèmes Atmosphériques, UMR7583, IPSL, CNRS, Université Paris Est Créteil, Université Paris Diderot, Créteil, FranceSearch for more papers by this authorG. Dufour, G. Dufour Laboratoire Interuniversitaire des Systèmes Atmosphériques, UMR7583, IPSL, CNRS, Université Paris Est Créteil, Université Paris Diderot, Créteil, FranceSearch for more papers by this authorX. Liu, X. Liu Harvard-Smithsonian Center for Astrophysics, Cambridge, Massachusetts, USASearch for more papers by this authorM. Joly, M. Joly CNRM/GAME UMR 3589 CNRS/Météo-France, Toulouse, FranceSearch for more papers by this authorC. Doche, C. Doche Météo-France/DIRSO/DEC/FDF, Mérignac, FranceSearch for more papers by this authorM. Beekmann, M. Beekmann Laboratoire Interuniversitaire des Systèmes Atmosphériques, UMR7583, IPSL, CNRS, Université Paris Est Créteil, Université Paris Diderot, Créteil, FranceSearch for more papers by this author First published: 24 May 2014 https://doi.org/10.1002/2013JD021340Citations: 11AboutSectionsPDF ToolsRequest permissionExport citationAdd to favoritesTrack citation ShareShare Give accessShare full text accessShare full-text accessPlease review our Terms and Conditions of Use and check box below to share full-text version of article.I have read and accept the Wiley Online Library Terms and Conditions of UseShareable LinkUse the link below to share a full-text version of this article with your friends and colleagues. Learn more.Copy URL Abstract Due to its impact on environment, tropospheric ozone received particular attention since several decades. Ground-based networks associated with regional chemical transport models are used to monitor and forecast surface ozone concentrations, but coverage, representativeness, and accuracy issues remain important. Recent satellite observations have demonstrated the capacity to probe tropospheric ozone, but there has been no explicit attempt to quantify their ability to measure ozone pollution near ground. We propose here to assess the ability of ozone sounders to detect a photochemical ozone pollution event that is supposed to be a favorable situation for satellite detection. We have chosen ozone pollution event over Europe associated with a warm conveyor belt that efficiently transports photochemically produced ozone upward. Ozone satellite products from Global Ozone Monitoring Experiment-2, Infrared Atmospheric Sounding Interferometer (IASI), and Ozone Monitoring Instrument are analyzed here for their capacity to capture such an event. Also, in situ observations and regional chemical-transport models show increasing ozone concentrations in the continental and Mediterranean boundary layer and further transport to central Europe and Scandinavia associated with upward transport. Satellite observations do not detect high ozone concentrations within the boundary layer due the weak sensitivity near the surface. Nevertheless, we have shown that the IR sounder IASI was able to detect, qualitatively and quantitatively, the ozone plume transported upward by the warm conveyor belt, suggesting that a quantification of upward transport of ozone pollution could be possible using current satellite observations. This should encourage us to further explore approaches more sensitive to surface ozone (such as the multispectral approach) and to prepare the next generation of still more sensitive spaceborne instruments. Key Point Ability of satellite to detect ozone pollution is evaluated 1 Introduction Tropospheric ozone is associated with well-known environmental issues. It is recognized as the third greenhouse gas and also as a strong oxidant involved in the budget of methane as a source of the OH radical [Forster et al., 2007]. In addition, ozone near the surface is a pollutant with strong adverse effects on human health and vegetation [Hayes et al., 2007; Sitch et al., 2007; World Health Organization, 2013]. To understand, quantify, and if possible mitigate its impact on the environment, we need to be able to know its concentrations (and their evolution) from global to local scale. In Europe, background ozone concentrations are controlled on one hand by the import of ozone-rich air masses from the stratosphere into the troposphere [Crutzen et al., 1999] and on the other hand by continental and intercontinental transport [Fiore et al., 2009]. In the remote troposphere, ozone is chemically produced by the slow oxidation of methane and carbon monoxide [Crutzen et al., 1999], in the presence of nitrogen oxide (NO) and UV radiation. In the polluted planetary boundary layer (PBL), the same processes occur at larger speed due to the oxidation of short-lived volatile organic compounds (VOCs). This photochemical ozone production takes place over the most populated continental areas where strongest emissions of ozone precursors occur. Finally, the balance is ensured by the dry deposition of ozone over solid surfaces (especially vegetation). Surface measurement networks over populated areas and at some remote locations enable an accurate (but spatially incomplete) monitoring of surface ozone concentrations. These networks are complemented by vertical soundings and measurements made on board of commercial aircraft [Marenco et al., 1998] and balloons that allow, for example, a better view of long-range transport. Regional chemical transport models (rCTMs) are efficient monitoring tools for ozone concentrations providing also forecasting capabilities. Based on such tools, operational platforms exist now in several European countries (see Kukkonen et al. [2012] for a review). The goals of these platforms are manifold ranging from operational forecasting of severe ozone episodes to verification of long-term exposure and mitigation scenarios. As a consequence, to improve the accuracy and operational aspects of such tools is now mandatory. This is the purpose of several international programs (Seventh Framework Programme/Monitoring Atmospheric Composition and Climate-Interim Implementation (MACC-II) (http://www.gmes-atmosphere.eu/) and Air Quality Model Evaluation International Initiative [Rao et al., 2011]). Following approaches used in meteorology and oceanography, the monitoring and forecasting tools developed for air quality are now using in a synergistic manner model and observations based on assimilation techniques [Elbern and Schmidt, 2001; Blond and Vautard, 2004; Hanea et al., 2004; Wu et al., 2008]. In this context, the use of satellite observations for pollution monitoring and evaluation is becoming more and more popular. For the case of ozone, the last generation of spaceborne sounders is offering increased opportunities to observe tropospheric ozone, either using UV/VIS (ultraviolet/visible) instruments like Global Ozone Monitoring Experiment-2 (GOME-2) [see European Organisation for the Exploitation of Meteorological Satellites, 2006] or Ozone Monitoring Instrument (OMI) [Levelt et al., 2006] or using TIR (thermal infrared) instruments like Tropospheric Emission Spectrometer (TES) [Worden et al., 2007] or Infrared Atmospheric Sounding Interferometer (IASI) [Clerbaux et al., 2009]. Concerning UV instruments, Liu et al. [2010] have shown, for the case of the OMI, the potential of this sounder to detect and follow tropospheric ozone plumes. Nevertheless, direct vertical sensitivity to boundary layer concentrations is rarely achieved due to reduced sensitivity to lower tropospheric ozone and coarse vertical resolution. Kar et al. [2010] demonstrated the capability of these UV/VIS instruments to occasionally “see” urban plume signatures but at monthly temporal scales and more efficiently for isolated urban centers. Also, Wang et al. [2011] shows a good agreement between sondes and OMI in the middle troposphere but large discrepancies between OMI and surface observations in the case of high ozone concentrations. On the other hand, several studies have already shown the potential of TIR instruments to monitor ozone concentrations in the troposphere and also in the lower troposphere. In the case of IASI observations, the accuracy of the surface to 6 km partial columns has been demonstrated with errors generally lower than 20% at midlatitudes [Keim et al., 2009; Dufour et al., 2012]. TES observations have already been used to analyze particular ozone features such as the Eastern Mediterranean and Middle East maxima during summer in the free troposphere [Liu et al., 2009; Worden et al., 2009; Richards et al., 2013]. This is also the case of IASI observations that have been used to investigate ozone pollution over polluted regions like China [Dufour et al., 2010] or Europe [Zyryanov et al., 2012]. Due to its twice daily coverage and a large across-track swath, IASI has proven to be capable of qualitatively probing events of elevated tropospheric ozone concentrations at daily scale, i.e., a photochemical pollution event over eastern Europe [Eremenko et al., 2008], a deep stratospheric intrusion over Europe [Zyryanov et al., 2012], the predominance of stratospheric intrusions that explain large ozone observed over Beijing region in winter than in spring and summer, enhanced partial columns of ozone are observed in coincidence with pollution alerts [Dufour et al., 2010]. This ability is, of course, of primary importance in the context of air quality. Moreover, these observations can be used to correct models. Parrington et al. [2008] have shown how the assimilation of TES observations could improve the representation of modeled-free tropospheric concentrations over a summer period. On the other hand, Parrington et al. [2009] show that if the assimilation of TES observations was modifying surface modeled fields, this did not result in a general model improvement. More recently, using not only TES but also OMI ozone observations, Huang et al. [2013] investigated the ability of these observations to detect a long-range transport episode and to further correct (via assimilation) the representation of boundary layer ozone in a CTM. They conclude that these observations were not very efficient in correcting the model in this case because of a lack of spatial coverage and maybe inadequate accuracy of these satellite data observations. Similarly, Coman et al. [2012] have assimilated IASI surface to 6 km ozone partial columns, allowing to reduce the model bias not only in the free troposphere but also at the surface. Improvement at the surface was partially due to downward transport of free tropospheric air masses, where the fields are preferentially corrected by IASI observations. Such situations occur when persistent subsiding anticyclonic conditions are present [Foret et al., 2009]. In spite of these numerous applications, the capability of satellite ozone observations to detect photochemical pollution occurring events within the boundary layer has not been clearly and definitively demonstrated and quantified. In the case of UV instruments, it has been demonstrated that while OMI has a significant sensitivity to the tropospheric ozone column, it has small sensitivity to lower tropospheric ozone [Sellitto et al., 2011]. In fact, the sensitivity of OMI retrievals to the boundary layer is not fully exploited, due to interference with aerosols and surface albedo [Liu et al., 2010]. Nevertheless, in the case of low cloud cover the sensitivity can be significantly improved in the lower part of the troposphere above clouds [Liu et al., 2010]. In the case of TIR measurements, observations are generally not very sensitive to surface ozone concentrations. Nevertheless, pollution events, which are one of the main targets of air quality monitoring and forecasting, present some peculiarities that are favorable to their detection by TIR sounders. Associated with intense photochemistry and air mass stagnation, anticyclonic situations favor the occurrence of strong ozone concentrations within the planetary boundary layer and possibly above. In such cases, cloud cover is weak and the number of available pixel (and correspondingly the amount of information) is significant. Moreover, large boundary layer heights favor mixing of surface ozone to altitudes where it can be detected. Also, a strong thermal contrast between the ground and the first atmospheric layer allows improving significantly the sensitivity of TIR instruments in lower layers. The objective of this work is to evaluate the capability of satellite instrument to observe surface and lower tropospheric ozone concentrations in the case of a high ozone event and the way through which it can be transported upward. To do so, a photochemical pollution case of moderate intensity, i.e., representative of current European pollution, associated with a warm conveyor belt has been chosen. In this case, ozone concentrations in the boundary layer are relatively high and are further lifted into the lower free troposphere by a warm conveyor belt associated with a classical frontal situation. Such conditions are recognized to be important processes to ventilate boundary layer pollution into the free troposphere at midlatitudes [Sinclair et al., 2008]. In situ ozone measurements (surface and profiles) and meteorological analysis are used to characterize this event. Available satellite observations are then confronted to these observations and simulations made with rCTM. More precisely, in this study we have used ozone retrievals from GOME-2, OMI, and IASI instruments. In this case of a few days event, the use of TES retrievals was not appropriate due to the very weak sampling of the instrument. In the following, we first give a short review of satellite products available (section 2); then we make a quick description of modeling systems (section 3). In section 4, we describe the case study from the meteorological point of view (section 4.1) and from the “ozone” point of view (section 4.2). In section 5, we discuss the capability of available satellite products to observe this pollution event. Finally, conclusions and perspectives are given (section 6). 2 Characteristics of Available Satellite Instruments and Products Several spaceborne instruments have already evidenced their capability to observe tropospheric ozone with reasonable accuracy. However, it is still unclear if such observations can be used to observe either qualitatively or quantitatively photochemical pollution events. We present here the different instruments that could be used for such a purpose. GOME-2 (Global Ozone Monitoring Experiment-2) Based onboard the MetOp satellites, GOME-2 is a nadir-viewing UV/VIS spectrometer which measures daytime earth reflectance (i.e., the ratio between backscattered radiance and solar irradiance) with ground pixels of 40 km × 80 km over a swath of about 2200 km (similar to IASI). The spectral resolution is ∼ 0.24 nm after convolution by the instrument response function and the sampling interval ∼ 0.12 nm. The ozone retrieval scheme used in the present paper is described and validated by Cuesta et al. [2013]. It is originally based on the work of Cai et al. [2012] for forward model calculations, but with spatial resolution refined by a factor of 8 (40 km × 80 km here). The scheme implements a Tikhonov-Phillips-type regularization, for which constraints are set in a way that they match approximately the retrieval noise for the IASI approach described below. This UV ozone retrieval considers two microwindows between 290 and 345 nm accounting for the Hartley and Huggins bands (from, respectively, channels 1 and 2 of GOME-2). Measurements below 290 nm are not used because of their sensitivity essentially to stratospheric ozone and due to low signal-to-noise ratios. IASI (Infrared Atmospheric Sounding Interferometer) The IASI instruments [Clerbaux et al., 2009] are nadir-viewing Fourier-transform spectrometers designed for operation on the meteorological MetOp satellites launched by European Organisation for the Exploitation of Meteorological Satellites (EUMETSAT) and European Spatial Agency. The first instrument, considered in this study, was launched onboard the satellite MetOp-A on 19 October 2006 and started operational measurements in June 2007. In addition to meteorological products (surface temperature and humidity profiles, and cloud information), the large spectral coverage (645–2760 cm−1), the high radiometric sensitivity and accuracy, and the rather high spectral resolution (the apodized spectral resolution is 0.5 cm−1) of the instrument allow deriving global distributions of several important atmospheric trace gases among which are ozone [e.g., Boynard et al., 2009], CO [e.g., George et al., 2009], and ammoniac [Clarisse et al., 2009]. The nadir field of view for one IASI pixel has the diameter of 12 km at the surface. The maximum scan angle of 48.3° from nadir corresponds to coverage of about 2200 km across track for one swath as for GOME-2. The retrieval of ozone profiles from IASI spectra used in the present study is based on the method described and validated in Eremenko et al. [2008] and Dufour et al. [2012]. OMI (Ozone Monitoring Instrument) OMI is a Dutch/Finnish UV/VIS nadir-viewing push-broom spectrometer, which is embarked on the NASA-Aura spacecraft [Levelt et al., 2006]. It is in the heritage of European Space Agency Global Ozone Monitoring Instrument (GOME) [Burrows et al., 1999] and the Scanning Imaging Absorption Spectrometer for Atmospheric Chartography [Bovensmann et al., 1999] instruments. OMI operates in the spectral interval 270–500 nm, with a spectral resolution of 0.42–0.63 nm and a direct nadir spatial resolution of 13 km along the track × 24 km across the track, in the so-called “Global Mode,” while it can reach the spatial resolution of 13 km × 12 km in its “Zoom Mode.” The spatial resolution at the borders of the swath is substantially lower than the nominal nadir resolution. The telescope of the OMI instrument has a wide field of view (114°), which corresponds to a 2600 km wide swath on the surface and global coverage time of 1 day. Since May 2007, a major degradation in the performances of OMI has emerged as the so-called “row anomaly,” caused by a problem in the OMI nadir port, that affects OMI radiance spectra measured for particular viewing directions of the instrument. The row anomaly initially only affected a few cross-track pixels but has become more serious since January 2009, affecting more than 1/3 of the cross-track pixels. More info can be found at http://www.knmi.nl/omi/research/product/rowanomaly-background.php. The tropospheric ozone columns analyzed in the present work are derived from the ozone profile product by Liu et al. [2010] with several major modifications described in Kim et al. [2013]. To speed up processing, the retrieval is done here at a nadir resolution of 52 km × 48 km by averaging (co-adding) 4/8 OMI UV1 (270–310 nm)/UV2 (310–330 nm) pixels. A major change to the retrieval presented by Liu et al. [2010] is the constraint on measurement error. Recent downward revision of the OMI measurement error (smaller by a factor of 1.4–2.3) [Braak, 2010], together with further reduction of this error through co-adding, results in unrealistically small observational error specification (0.035% at 320 nm under tropical clear conditions) that causes spurious variability in the retrieval. Therefore, a minimum measurement error of 0.2% in the spectral region of 300–330 nm is imposed. Although this error specification stabilizes the retrievals, it significantly reduces the retrieval sensitivity compared to that in Liu et al. [2010]. Pixels with effective cloud fraction greater than 0.3 are considered as “cloudy” and are screened out. The remaining pixels are quality checked by means of the quality flags in the Level 2 data, and pixels affected by the row anomaly are further excluded from the analysis. 3 Description of the Modeling System For this study, we used two state-of-the-art rCTMs, CHIMERE [Menut et al., 2013] and MOCAGE [Barré et al., 2013]. Configurations of both models are derived from the operational versions that have been set up for the MACC-II project [Barré et al., 2012; Zyryanov et al., 2012]. The geographical domain covered is roughly western Europe as shown in Figure 1. The horizontal resolution is 0.2° × 0.2° for MOCAGE and 0.25° × 0.25° for CHIMERE. MOCAGE covers a vertical domain from the surface to the upper stratosphere (5 hPa) using 23 hybrid (σ, p) vertical levels within the troposphere. CHIMERE covers the troposphere (from the surface to 200 hPa) with a similar vertical discretization (20 hybrid (σ, p) levels). Different parameterizations used in both models are summarized in Table 1. For advection, the MOCAGE model uses a semi-Lagrangian approach [Williamson and Rash, 1989], while CHIMERE uses a sixth-order [Colella and Woodward, 1984] or third-order [Van Leer, 1979] scheme in an Eulerian framework. Chemical modules are also different, since in the MOCAGE model, tropospheric and stratospheric chemistry of ozone are represented, which is not the case in the CHIMERE model with only tropospheric chemistry. Dry deposition schemes both follow the “resistance” approach [Wesely, 1989]. Among external forcings, only anthropogenic emission inventories are similar for both models (TNO inventory) [Visschedijk et al., 2007]. Other model forcings are different: MOCAGE is using the meteorological analysis from ARPEGE [Courtier et al., 1991] at a 3 h frequency, and chemical boundary conditions are provided by the MOCAGE-CTM itself; for both meteorology and the boundary conditions, CHIMERE is using European Centre for Medium-Range Weather Forecast (ECMWF) products, i.e., 3-hourly forecast of Integrated Forecasting System (IFS) and Model for Ozone And Related chemical Tracers (MOZART)-IFS. In conclusion, both rCTMs are significantly different in their formulation and their input data, and their simultaneous use will give robustness to the analysis and comparisons performed in this work. Figure 1Open in figure viewerPowerPoint Two-meter temperature fields and surface winds from IFS at 3 P.M. for (top left) 18 August 2009, (top right) 19 August 2009, (bottom left) 20 August 2009, and (bottom right) 21 August 2009. Table 1. Main Parameterizations Used for CHIMERE and MOCAGE Models Natural Emissions Chemistry Advection Convection Diffusion Dry Deposition CHIMERE [Menut et al., 2013] Guenther et al. [2006] MELCHIORII; Schmidt et al. [2001] Colella and Woodward [1984]; Van Leer [1979] Tiedtke [1989] Troen and Mahrt [1986] Zhang et al. [2003] MOCAGE [Josse et al., 2004; Bousserez et al., 2007] Guenther et al. [1995]; Dentener et al. [2005] RACM; Stockwell et al. [1997]; + REPROBUS; Lefèvre et al. [1994] Semi-Lagrangian [Williamson and Rash, 1989] Bechtold et al. [2001] Louis [1979] Michou and Peuch [2002] 4 Description of the Case Study For this study, we have chosen an ozone pollution event of moderate intensity. It is characterized by high surface ozone concentrations sometimes exceeding 90 ppb (i.e., the information threshold of ozone in Europe). It is one of the most important events during summer 2009 [European Environmental Agency, 2010], but if compared to ozone pollution events occurring between 2007 to 2012, we consider it as a rather typical event for western and central Europe [European Environmental Agency, 2013a] far from high pollution events occurring in 2003 and 2006 [European Environmental Agency, 2013b]. It occurred in August 2009 (19–21). Moreover, during this period, air masses are lifted from the boundary layer upward due to the presence of a warm conveyor belt transferring polluted air masses into the lower free troposphere as it is frequently observed at northern midlatitudes in the vicinity of frontal systems. Meteorological Situation To describe the meteorological situation, we use analysis (every 12 h) and forecast (3-hourly between analyses) from the Integrated Forecasting System (IFS) of the European Centre for Medium-Range Weather Forecast (ECMWF). On 19 August 2009, an anticyclonic situation set up over western Europe and especially over France; it was characterized by low wind speed (and associated weak dispersion), high surface temperatures above 30°C (Figure 1), generally a good proxy of ozone pollution, and low cloud cover. The next day (20 August 2009), a low-pressure system arrived at the western tip of France. On 21 August 2009, this structure propagated eastward with still high temperature on its eastern side. Also, geopotential height at 500 hPa confirms this evolution, with the presence of an anticyclonic situation over Europe followed by a more and more baroclinic circulation with the formation of a through over western Europe and a ridge over central Europe (Figure 2). A frontal structure extending from southwest to northeast along geopotentiel isolines is observed (Figure 2), presenting strong temperature gradients at 3 km altitude (Figure 3). It is linked to a warm conveyor belt ahead of the low-pressure system associated with a cloud band along the front (not shown). Figure 2Open in figure viewerPowerPoint Geopotential height fields at 500 hPa from IFS at noon for (top left) 18 August 2009, (top right) 19 August 2009, (bottom left) 20 August 2009, and (bottom right) 21 August 2009. Figure 3Open in figure viewerPowerPoint Three-kilometer height temperature and wind fields from IFS at 3 P.M. for (top left) 18 August 2009, (top right) 19 August 2009, (bottom left) 20 August 2009, and (bottom right) 21 August 2009. Warm conveyor belts favor the transport of air masses from the boundary layer to the free troposphere. They are characterized by warm air streams that are uplifted along and ahead a cold front [Bethan et al., 1998; Kowol-Santen et al., 2001; Agustí-Panareda et al., 2005, 2009]. HYSPLIT back trajectories [Rolph, 2013] show that air masses arriving at
Referência(s)