Chemical Transport Models (CTM) are an essential computational method used to predict air quality, which are closely aligned to weather prediction models. They also incorporate the complex reactions of pollution emissions, their dispersal in the atmosphere, the chemical transformations they undergo and their removal processes. Contemporary CTM solely or combined with other modern data sampling and processing facilities are the computational core of broad range of European initiatives such as GMES (Europe’s initiative for Global Monitoring of Environment and Security), HARMO (Harmonisation within Atmospheric Dispersion Modelling for Regulatory Purposes), ACCENT (Atmospheric Composition Change, the European Network) and others. GMES is the name of the Earth Observation programme launched by the European Commission and the European Space Agency. The objective of GMES is to monitor and forecast the state of the environment on land, at sea and in the atmosphere and to improve the security of the citizens in a world facing an increased risk of natural and other disasters. GMES does not replace existing European capacities, but rather complements them with a view to fulfilling user needs and guaranteeing sustainability and European Earth Observation autonomy in the long term. MACC-II - Monitoring Atmospheric Composition and Climate - Interim Implementation (http://www.gmes-atmosphere.eu/services/raq) - is the current pre-operational GMES Atmosphere Service. MACC-II provides data records on atmospheric composition for recent years, data for monitoring present conditions and forecasts of the distribution of key constituents for a few days ahead. MACC-II combines state-of-the-art atmospheric modelling with Earth observation data to provide information services covering European air quality, global atmospheric composition, climate forcing, the ozone layer and UV and solar energy, and emissions and surface fluxes. The principal goal of other initiative, the PASODOBLE project, (http://www.myair-eu.org) is to develop and demonstrate user-driven information services (MYAIR services) for the regional and local air quality sector by combining space based data, in-situ data and models in four thematic service lines. Through this project European citizens will directly benefit from earth observation, measurement networks and air quality modelling. The project infrastructure includes interfaces to GMES Core Services – satellite based measurements and in-situ data.
The present work is based on the primary results obtained in a modelling study preformed in the Bulgarian National Institute of Meteorology and Hydrology (NIMH), in the frame of the EC FP6 Project “Central and Eastern Europe Climate Change Impact and Vulnerability Assessment” (CECILIA) program. CECILIA’s primary mission was to improve the understanding of local climate change in Central and Eastern Europe and its impacts on forestry, agriculture, hydrology and air quality. The main objective of the project was to deliver a climate change impact and vulnerability assessment in targeted areas of Central and Eastern Europe. Emphasis is given to applications of regional climate modelling studies at a resolution of 10 km for local impact studies in key sectors of the region Juda-Rezler et al (2012). Specific tasks of the activities in CECILIA’s workpackage 7 (WP7) was to study the impacts of climate change on health and air quality (photochemistry of air pollution, aerosols). For this purpose special modellig system (Syrakov et al 2010; Syrakov et al 2011) was build in NIMH - BAS Sofia, which computational core is the US EPA Models-3 (Byun and Ching 1999, Byun and Schere 2006) air quality modelling, consisting of:
-
CMAQ - Community Multi-scale Air Quality model being the CTM of the Models-3 System;
-
MM5 - the 5th generation PSU/NCAR Meso-meteorological Model used as meteorological pre-processor to CMAQ, and
-
SMOKE - Sparse Matrix Operator Kernel Emissions Modelling System being the emission pre-processor to CMAQ.
A number of interfaces (Linux scripts and FORTRAN codes) are created as to link those models with different types input information in a system able to perform long term calculations. Meteorological driver of the system is the ALADIN weather forecasting model Spiridonov et al. (2005) operating in this case in special mode. In spite MM5 performs full meteorological modelling, here it is used as a kind of dynamical pre-processor in space and time. The domain’s vertical profile contained 23 σ-levels of varying thickness, extending up to 100 hPa height. Proper physical options are set in MM5. The speciation procedure is dependent on the Chemical Mechanism (CM) used. CMAQ supports different CMs. Here, the Carbon Bond, v.4 (CB4) is exploited. The basis of the CB-4 mechanism is that reactivity of organic compounds in the atmosphere can reasonably be simulated by mechanism species that represent different carbon bond types. In the time passing CB4 has undergone several changes since its publication. In particular particle mater chemistry is added. In the used Version 4.6 of CMAQ the CB4 is upgraded with the Version 1.7 of ISORROPIA aerosol model Nenes et al. (1998). According to this combined mechanism (CB4-aero3) 10 organic and 5 PM2.5 spices are input in addition to the other inorganic gases. Most of the organic species in CB4 represent carbon-carbon bond types, but ethene (ETH), isoprene (ISOP) and formaldehyde (FORM) are represented explicitly. A new method for computing the vertical velocity has been implemented which follows the omega calculation in the Weather Research and Forecasting (WRF) model but using CMAQ’s advection schemes to compute horizontal mass divergence. It’s a two step process where first the continuity equation is integrated through the column to get the change in column mass and then solve for omega layer-by-layer using the horizontal mass divergence. The new scheme is much less diffusive in the upper layers because it is constrained to have zero flux at the model top. The dry deposition is modeled using the electrical resistance method. The wet removal processes can proceed first with the cloud droplet formation via several mechanisms including heterogeneous nucleation and aerosol activation, then with in-cloud scavenging by existing cloud droplets or below-cloud scavenging by falling precipitation or both. All of these processes influence the amount and composition of the ground-level rainwater. Therefore, an accurate parameterization scheme, describing all these processes is incorporated in the system.
As shown in Chervenkov et al. (2008), the long range transport of pollutants can play relevant role even for not long-living pollutants like SO2. The modelling system needs adequate boundary conditions (BC). The necessary for the BC condition data are prepared with the Comprehensive Air quality Model with eXtensions (CAMx) Katragkou et al. (2008) by means of off-line procedure that takes place in Aristotle University of Thessaloniki in Greece. The results are uploaded to a dedicated server in Sofia. There, this data is processed online in order to produce the current day CMAQ-ready boundary condition file. Due mainly to the specifics of the coordinate system, the CMAQ do not requires upper boundary conditions.
CMAQ demands its emission input in specific format, reflecting the time evolution of the release of all pollutants accounted for by the used chemical mechanism. The emission inventory is usually made on annual basis for, as a rule, big territories (municipalities, counties, countries, etc.) and many pollutants are estimated as groups like NOx, SOx, VOC (Volatile Organic Compounds), PM2.5 In preparing CMAQ emission file a number of specific estimates must be done. Firstly, all this information must be gridded. Secondly, time variation profiles must be over-posed on these annual values to account for seasonal, weekly and daily variations. Finally, organic gases emission estimates, and to a lesser extent SOx, NOx and PM2.5, must be split, or ‘speciated’, into more defined compounds in order to be properly modeled for chemical transformations and deposition. The different types of sources: Area Sources (AS), Large Point Sources (LPS), mobile and Biogenic Sources (BS) are treated in specific way. Obviously, emission models are needed as reliable emission pre-processors to the chemical transport models. Such a component in EPA Models-3 system is SMOKE. Unfortunately, it is highly adapted to the US conditions – emission inventory, administrative division, motor fleet etc. Many European scientific groups are working now for adapting SMOKE to European conditions Borge et al (2008). SMOKE is partly used here, only for to calculate biogenic emissions and to merge AS-, LPS- and BS-files into a single CMAQ emission input file. The anthropogenic emission files (AS and LPS) are prepared by interface programs. Input to these interfaces is gridded emission inventories. Here, TNO high resolution inventory Visschedijk et al (2007) is exploited. The TNO inventory resolution is 0.25° × 0.125° longitude-latitude, that is on average 15 × 15 km. GIS technology is applied to transform this data to the CMAQ grid and so to produce the correct input. It must be mentioned that the TNO inventory is elaborated for AS and LPS separately, distributed over 10 SNAPs (Selected Nomenclature for Air Pollution). The temporal allocation is made on the base of daily, weekly and monthly profiles, provided by Peter Builtjes Builtjes et al., (Builtjes et al. 2003). The temporal profiles are country-, pollutant- and SNAP-specific.
SMOKE is used to produce biogenic emission file. The SMOKE system uses a more advanced emissions modelling approach for biogenic processing than it uses for the other source types. For biogenic emissions, the temporal processing is a true simulation model driven by ambient meteorology and other data. SMOKE currently supports BEIS (Biogenic Emissions Inventory System) mechanism, versions 2 and 3 (here version 3.13, see Schwede et al, 2005). BEIS2 and BEIS3 are fed with spatial allocation of land-use data as the first processing step. They subsequently compute normalized emissions for each grid cell and land use category. The final step is adjusting the normalized emissions based on gridded, hourly meteorology data and assigning the chemical species to output a model-ready biogenic emissions file. To isolate the effect of climate change on air quality, anthropogenic emissions scenario for year 2000 was used in all simulations, while climate sensitive biogenic emissions were allowed to vary with the simulated climate. Due to the generally decreasing trend in the anthropogenic share in the emissions in the decade, it is reasonable to expect that this approach underestimates to a certain extend the real air pollution situation. Nevertheless the study is appropriate both because the emission abatement in Bulgaria is not so well expressed as in other European countries. On the other hand, the calculations are performed with very precise emission inventory and high spatio-temporal resolution with modern simulation model system. The achieved results can be treated as estimation of the lower limit of the real Syrakov et al. (2011)) The model grid consists of 54 × 40 grid cells with size 10 × 10 km and covers Bulgaria entirely, together with the border regions of the neighboring countries and the most western part of the Black sea. From the modelling system output only values for the surface layer are stored. Saved on hourly basis are 17 most important pollutants, including all under consideration. The calculated datasets for every day in the decade 1991–2000 are saved in the so-called “Control Run” data base. The needed hourly concentrations for the computation of the statistical quantities, required for the comparison from the directives, are taken from this data base and processed. On the next step the additional condition, namely this for the tolerated number of exceedances, is checked. If, due to any reason (for instance different stages of one directive) for some pollutant we have “duplicate” formulation of the AQLV and the attached additional condition, the more severe one is checked first. This is done in the following manner: If, say, for a given pollutant the directive allows m exceedances of the corresponding statistical average, we find the (m + 1) th maximum from all possible values. If this value is greater than the prescribed AQLV, the legislation is breached. This task, namely to find the m-th largest among n concentrations has to be repeated many times in each grid cell and that’s why it is essential to optimize the corresponding numerical routine. This is done by setting especially effective procedure by m < <n, because this condition is very well expressed. In all cases we find the maximal value also due to its relevance as indicative of the air pollution. All statistical quantities are calculated for every year in the period 1991–2000 and the averaged slice results over the whole time are presented in the next chapter.