Version classiqueVersion mobile
OpenEdition Books

The Mediterranean region under climate change

 | 
Jean-Paul Moatti
, 
Stéphane Thiébault

Part 1. Mechanisms, observed trends, projections

Sub-chapter 1.4.2. High atmospheric concentrations of aerosols, greenhouse gases and other pollutants

Texte intégral

1Evidence is growing for the deterioration of air quality over the Mediterranean basin. In this sub-chapter, we review the current situation over this semi-enclosed basin. We describe the spatio-temporal variabilities of the pollutants, greenhouse gases (GHGs) and aerosols, observed and modeled over the Mediterranean basin and how they help trace physical-chemical processes at regional and global scales through long-range transport. In a separate text box (see below), we focus in particular on secondary organic aerosols (SOA), a major component of fine particles, for which data and simulations were particularly scarce and uncertain especially over the western Mediterranean basin before ChArMEx. The Mediterranean basin is located in a transition zone between subtropical and mid-latitude climate regimes (Lionello, 2012), and is highly sensitive to climate change. In terms of sources of anthropogenic pollution, the basin is located at the intersection of three continents, Europe, Africa and Asia.

2Satellites and models (e.g., LELIEVELD et al. 2002; Nabat et al. 2013) together with campaigns such as the Mediterranean Intensive Oxidant Study (MINOS) (Ladstätter-Weissenmayer et al. 2003; Scheeren et al. 2003) show that, during the warm and dry Mediterranean summer season, air pollution above the Mediterranean often exceeds that observed over most parts of Europe. This is due to the convergence of European, African and Asian polluted air masses, to the absence of rain to clean up the atmosphere, and to the high insolation, which favors the formation of secondary pollutants like ultrafine particles or ozone. Natural aerosol pollution could originate from sources such as the African and Arabian deserts, active volcanoes, vegetation, or the sea surface. Pollutants, GHGs and aerosols originating from Asia can be trapped in the Asian monsoon and entrained to the upper troposphere before being redirected towards the eastern Mediterranean basin via the Asian monsoon anticyclone, where they accumulate and are subject to subsidence (Ricaud et al. 2014).

3Continental sources including industrial and densely populated coastal areas (Kanakidou et al. 2011; Im And Kanakidou, 2012) or forest fires (Cristofanelli et al. 2013) affect the ozone (O3) and carbon monoxide (CO) budgets with which methane emissions (CH4) interplay through complex reactions with nitrogen oxides (NOx) (Dentener et al. 2005), although the impacts of the respective source types are still not fully understood. Polluted air masses affecting the Mediterranean basin may originate from Europe (e.g., Pace et al. 2006), Asia (e.g., Lelieveld et al. 2002; Randel And Park, 2006), Africa (e.g.Ziv et al. 2004; Liu et al. 2009) and even North America (e.g., Formenti et al. 2002; Christoudias et al. 2012).

4The transport and dispersion conditions of atmospheric pollutants over the western and eastern Mediterranean basin differ, as illustrated in Fig. 4, which shows the contrasted vertical air motion in the western and eastern parts of the basin in summer. The subsiding air aloft induced by the descending branch of the Hadley global circulation cell affecting the eastern basin, and the depth of the Persian Trough (an extension of the Indian monsoon), control the spatio-temporal distribution of the boundary layer (BL) height during summer. The resulting shallow mixed layer and weak zonal flow, leads to poor ventilation rates, inhibiting the efficient dispersion of the pollutants. Several studies pointed to specific local (e.g. ventilation rates) and regional peculiarities (long-range transport) that enhance the building up of pollutant concentrations (Wanger et al. 2000; Matvev et al. 2002; Erel et al. 2007; Rudich et al. 2008; Drori et al. 2012) over the eastern Mediterranean basin.

Figure 4
Top panel: June to August 1948-2015 average of the NCEP/NCAR model omega (Pa s-1) at 500 hPa (about 5.5 km altitude) showing a maximum downward air motion of about 1.5 cm s-1 over Crete in the eastern Mediterranean, contributed by the descending branch of the African and Asian monsoon, and a maximum upward motion of about 0.5 cm s-1 over the westernmost Mediterranean basin. Bottom panel: spatial distribution of the summer mixed layer height over the Mediterranean (from Dayan et al. 1996).

5Considering the long-range transport climatology that characterizes the Mediterranean basin, two different regimes can be observed: 1) From fall to spring with predominance in winter, and from the boundary layer to the upper troposphere, air masses mostly come from either Europe or the eastern Atlantic Ocean. 2) In summer, the origin of air masses affected by long-range transport over the eastern and western Mediterranean basin is distinct, more complex and altitude dependent. In the lower troposphere over the western basin, convective cells develop within the boundary layer with mostly air masses coming from Europe, northern Africa and eastern Atlantic Ocean. The coasts and mountains surrounding the western basin favor the development of mesoscale recirculations in summer that lead to the formation of ozone-and aerosol-rich layers above coastal areas and the sea (e.g. Millán et al. 2000). Over the eastern basin, the air masses originate from four major source regions: (i) west-north-west long fetch of maritime European air masses all year round, (ii) north-west flow originating in south-eastern Europe (etesian winds) in summer, (iii) south-east flow from the Arabian Peninsula in the fall, and (iv) south-west flow along the North African coast most frequent in late winter and spring (Dayan, 1986). In the mid-troposphere, whatever the season, air masses in both parts of the basin mainly come from the west. In summer, upper tropospheric air masses in the western basin mainly come from the west, but in the eastern basin, they also come from North Africa and the Arabian Peninsula (Ziv et al. 2004; Liu et al. 2009), and even farther away, from Asia (Lelieveld et al. 2002).

6Several airborne campaigns have recently been conducted above the Mediterranean basin as part of the ChArMEx program: TRansport and Air QuAlity (TRAQA) in 2012, Aerosol Direct Radiative Impact on the regional climate in the MEDiterranean region (ADRIMED) and Secondary Aerosol Formation in the MEDiterranean (SAFMED) in 2013, and SAFMED+ and Gradient in Longitude of Atmospheric constituents in the Mediterranean basin (GLAM) in 2014. They addressed different processes that impact pollutants, GHGs and aerosols: air quality, radiative impact of aerosols, secondary aerosol formation, and long range transport. Combined with spaceborne and modeling studies, airborne in situ measurements highlighted the strong pollutant, GHG, and aerosol gradients between the western and eastern Mediterranean basin in summer from the mid-to-upper troposphere. Maxima in ozone, carbon monoxide, methane, nitrous oxide are commonly observed in the eastern basin, although on some occasions, minima in carbon dioxide are detected. The gradients were mainly attributed to the impact of long range transport of air masses originating either from Asia, North America, Europe, or Africa (Fig. 5). In the case of Asia, the Asian monsoon and its associated anticyclone are the main cause of the gradient by (1) trapping lower tropospheric pollutants and GHGs in the Asian monsoon; (2) updrafting pollutants and GHGs in the Asian monsoon up to the upper troposphere; (3) building up pollutants and GHGs within the Asian monsoon in the upper troposphere; (4) re-distribution of the pollutants and GHGs at a large scale by the Asian monsoon anticyclone to the Middle East and North Africa in the upper troposphere; and (5) subsiding pollutants and GHGs into the middle troposphere above the eastern basin (RICAUD et al. 2014). In the case of North America, biomass burning by long-lasting forest fires is the main source of aerosols, pollutants and GHGs. They are updrafted into the mid-to-upper troposphere by pyro-convection and/or through meteorological systems generally located in the Atlantic Ocean (warm conveyor pool, strong depression, etc.) following the jet stream towards the western Mediterranean basin (ANCELLET et al. 2016; RICAUD et al. submitted). In the case of Africa, outbursts of Saharan desert dust are advected westward towards the Caribbean Sea, in the lower part of the troposphere, and, once in the central North Atlantic Ocean, are trapped within meteorological systems to be finally transported upward to the western Mediterranean basin like in the case of transport from North America.

Figure 5
Very long-range intercontinental transport of air masses over the Mediterranean basin. Dots identify source regions and arrows identify transport pathways.

7In addition to intercontinental transport of pollutants, GHGs and aerosols, the Mediterranean basin is a region that favors stratosphere-to-troposphere transport (ZBINDEN et al. in prep.). Stratospheric intrusions result in high ozone and low carbon monoxide or methane penetration into the troposphere. The depth, the irreversibility and the frequency of the stratospheric penetrations modify the tropospheric climatology and trends of these chemical species (Tyrlis et al. 2014).

8Although measured and modeled data are still being analyzed in the ChArMEx program, their comparison underlines the difficulties models face representing 1) the significant fine structures observed both horizontally and vertically, 2) some processes that might be missing in the models (e.g. pyro-convection), 3) the mixing ratios of pollutants and GHGs with sufficient accuracy to estimate trends. Finally, chemical analyses that consist in assimilating satellite data with model data, widely used during the ChArMEx airborne campaigns, are efficient tools that benefit studies of the distribution and evolution of pollutants, GHGs, and aerosols over time.

9The ChArMEx program allowed us to investigate the variabilities of pollutants, GHGs and aerosols in the Mediterranean basin in space and over time with particular emphasis on the summer period combining in situ airborne, spaceborne and model data, and to attribute these variabilities to several processes. Pyroconvection, convection, monsoon and the Asian monsoon anticyclone, the jet stream, etc., are dynamic processes that redistribute pollutants, GHGs and aerosols originating from desert dusts, biomass burning, etc., from Asia, Europe, Africa and North America over the Mediterranean basin. Some potentially key geographical areas that impact the eastern basin were revealed, e.g. the Arabian Sea, where future airborne campaigns may be deployed.

Box 1
The secondary organic aerosol (SOA)
The secondary organic aerosol (SOA) is made up of thousands of organic compounds originating from a wide range of natural and human sources (combustion of fossil fuel and biomass, gaseous emissions from the continental biosphere, marine emissions, etc.). SOA accounts for from 20% to 60% of very fine airborne particulate matter PM1 (particles whose aerodynamic diameter is <1 µm) (Zhang et al. 2007). This aerosol fraction is the most relevant for health issues (Pope And Dockery, 2006) but also climate effects. Organic aerosol particles scatter solar radiation, which cools the Earth’s atmosphere, since part of the radiation is reflected back to space, but more recently their absorbing properties have also been put forward (Zhang X. et al. 2011). This affects the radiative properties of aerosol observed over the Mediterranean basin (Di Biagio et al. 2016). Depending on its hygroscopicity and mixing state, organic aerosols affect cloud micro-physics (Lohmann And Feichter, 2005; Zhu et al. 2016). SOA forms from semi-volatile organic compounds (SVOCs) by nucleation or condensation. Oxidative processes are most efficient in lowering the volatility of initially volatile compounds of anthropogenic and biogenic origin (Kroll et al. 2008). Anthropogenic emissions play a role in the formation of oxidants and hence in the formation of SOA, so that SOA of biogenic origin may be strongly reduced by reducing anthropogenic emissions, especially above large cities around the Mediterranean Sea (Sartelet et al. 2012). Research to elucidate the formation of SOA is still very active. Once formed, SOA can become highly viscous, which prevents later evaporation of the organic material (Virtanen et al. 2010).
During ChArMEx, and for the first time in the western Mediterranean, intensive and longterm aerosol mass spectrometer measurements filled the gap of missing observations of SOA over this region. On a yearly average, organic aerosol made up about 50% of PM1 at Cape Corsica (POM in Fig. B1, left), among which around 90% were of secondary origin (LV-OOA and SV-OOA in Fig. B1, right). Surprisingly, aged (highly oxidized) SOA (LV-OOA) concentrations were found throughout the year, dominating OA even in winter when photochemical conditions are low. This result provides further evidence for the highly oxidative atmosphere of the Mediterranean. Additional isotopic 14C measurements during the intensive campaign in the summer of 2013 showed that the majority of the organic aerosol was of biogenic origin (from the continental biosphere), although still 20% at Cape Corsica and 35% at Mallorca were of anthropogenic origin (fossil fuel combustion).
SOA modelling is still very challenging as the many chemical reaction pathways are not explicitly known and need to be parameterized in 3D models. Including a volatility basis-set (VBS) scheme including multi-step functionalization and fragmentation of organic compounds and formation of non-volatile SOA in the CHIMERE regional chemistry-transport model (Shrivastava et al. 2015) made it possible to retrieve the SOA mass and its fossil vs. biogenic fractions observed at surface stations at Cape Corsica and Mallorca (Cholakian et al. 2016), and recently in the Paris area (Zhang et al. 2015). Simulations with the Polyphemus model including the new multiphase SOAP organic aerosol scheme were in good agreement with measurements of concentrations of organic aerosol at Cape Corsica. The formation of extremely low-volatility organic aerosols has been added to the model to better represent organic aerosol properties (Chrit et al., 2016). Well evaluated chemistry-transport models is a prerequisite for simulating the regional scale interaction of SOA with climate (radiation, micro-physics) over the Mediterranean region as the next step. Ongoing developments will be constrained and validated by the large number of field observations made recently in the western Mediterranean as well as new measurements in the eastern basin as part of the ChArMEx program.

Figure B1
PM1 composition in µg m-3 (left) and source apportionment (right) of carbonaceous aerosols at Ersa, Cape Corsica (42° 58’N, 9° 23'E, alt. 530 m), averaged over two years (from June 2012 to July 2014) from off-line filters and on-line aerosol mass spectrometer (Q-ACSM) measurements, respectively (after Nicolas, 2014). POM stands for particulate organic matter, NO3 for nitrate, nss-SO4 for non-sea salt sulfate, NH4 for ammonium end EC for elemental carbon. LV-OOA (low-volatility oxygen-like organic aerosol) is assimilated to aged secondary organic aerosol, SV-OOA (semi-volatile oxygen-like organic aerosol) to freshly formed secondary organic aerosol, HOA (hydrogen-like organic aerosol) to primary organic aerosol.

Table des illustrations

Légende Figure 4Top panel: June to August 1948-2015 average of the NCEP/NCAR model omega (Pa s-1) at 500 hPa (about 5.5 km altitude) showing a maximum downward air motion of about 1.5 cm s-1 over Crete in the eastern Mediterranean, contributed by the descending branch of the African and Asian monsoon, and a maximum upward motion of about 0.5 cm s-1 over the westernmost Mediterranean basin. Bottom panel: spatial distribution of the summer mixed layer height over the Mediterranean (from Dayan et al. 1996).
URL http://books.openedition.org/irdeditions/docannexe/image/23373/img-1.jpg
Fichier image/jpeg, 280k
Légende Figure 5Very long-range intercontinental transport of air masses over the Mediterranean basin. Dots identify source regions and arrows identify transport pathways.
URL http://books.openedition.org/irdeditions/docannexe/image/23373/img-2.jpg
Fichier image/jpeg, 155k
Légende Figure B1PM1 composition in µg m-3 (left) and source apportionment (right) of carbonaceous aerosols at Ersa, Cape Corsica (42° 58’N, 9° 23'E, alt. 530 m), averaged over two years (from June 2012 to July 2014) from off-line filters and on-line aerosol mass spectrometer (Q-ACSM) measurements, respectively (after Nicolas, 2014). POM stands for particulate organic matter, NO3 for nitrate, nss-SO4 for non-sea salt sulfate, NH4 for ammonium end EC for elemental carbon. LV-OOA (low-volatility oxygen-like organic aerosol) is assimilated to aged secondary organic aerosol, SV-OOA (semi-volatile oxygen-like organic aerosol) to freshly formed secondary organic aerosol, HOA (hydrogen-like organic aerosol) to primary organic aerosol.
URL http://books.openedition.org/irdeditions/docannexe/image/23373/img-3.jpg
Fichier image/jpeg, 84k

© IRD Éditions, 2016

Conditions d’utilisation : http://www.openedition.org/6540