Final Report Summary - APPRAISAL (Air Pollution Policies foR Assesement of Integrated Strategies At regional and Local scales)
Executive Summary:
Air quality in Europe is still facing a continued wide-spread of exceedances, particularly regarding PM, NOx and O3. In case of non-compliance the 2008 Air Quality Directive requests Member States (MS) to design local and regional plans and assess their impacts on air quality and human health. MS have therefore developed and applied a wide range of modelling methods to cope with these obligations. Today, with the revision of the EU air quality policy pending, there is a need to consolidate and assess the research results in the field of Air Quality and health Impact Integrated Assessment and make them accessible to policy makers.
In this context, the APPRAISAL project aims at:
1. Undertaking on overall review of the Integrated Assessment methodologies used in different countries at regional and local scale, from the simple (scenario analysis) to the more comprehensive (cost-benefit, cost-effectiveness analysis). This includes evaluating both top-down and bottom-up approaches to systematically analyse their strengths and weaknesses and to identify key areas to be addressed by further research.
2. Designing an Integrated Assessment Modelling framework, based on the information collected during the review process, for different policy-maker requirements, model capabilities and levels of data completeness.
3. Drawing guidelines on how to implement the defined Integrated Assessment System framework, based on identified strengths and weaknesses and best practice examples among the Integrated Assessment Systems in place within MS.
4. Communicating with key stakeholders and in particular to policy-makers the state-of-the-art scientific knowledge on emission abatement assessment.
APPRAISAL includes 15 highly experienced groups working on both air quality and health impacts assessment.
Partners come from all over Europe to guarantee the review process representativeness. A group of stakeholders are closely connected to the Consortium to ensure a direct line of communication with key policy makers. APPRAISAL contributes to improve knowledge on regional and local IA methodologies and supports the revision of EU air quality policies.
Project Context and Objectives:
Exceedances of air quality limit values in urban areas in Europe remain widespread, particularly for PM, NOX and O3. This is not simply a compliance issue but has significant implications for the health and wellbeing of European citizens.
The recent reports on the review of the Thematic Strategy on Air Pollution (Amann et al., 2013; Keiswetter et al., 2013) show the evolution trend of compliance from the base year 2010 to 2025 (assuming current legislation only), the improvement for the optimised A5 so-called ‘Central Policy Scenario’ by 2025 and the further compliance achieved in 2030, by implementing all technical measures (MTFR). The assessment of compliance with the annual mean NO2 limit value and the daily PM10 exceedances limit value are both shown in Figure 1 and Figure 2, respectively. In each case, the limit values used for assessing compliance are those of the current Ambient Air Quality Directive.
Figure 1. NO2 Annual mean compliance assessment (Amann et al. 2013).
Figure 2. PM10 compliance assessment (Amann et al. 2013).
Some important observations can be derived from these figures:
(i) The first observation in comparing the 2010 map with the 2025 CLE case is the clear move away from a general picture of non-compliance (2010) to few limited remaining areas of non-compliance. European wide measures (already mandated) will determine a significant improvement in compliance especially in the EU-15 Member States. What is also clear by comparing the 2025 CLE with the 2025 A5 (defined as ‘central policy scenario’) is the limited potential of further EU-wide measures to improve compliance; this is further underlined by comparing the 2025 A5 scenario with the 2030 MTFR scenario.
(ii) Introducing tougher European-wide measures to address residual non-compliance confined to 10% of the urban zones in Europe (the extent of NO2 non-compliance according to IIASA in the 2025 CLE scenario) would likely be significantly more costly than directly addressing the non-compliance areas with specifically designed measures based on bottom-up Integrated Assessment using regional/local data. This has significant implications for the role of regional/local ‘bottom up’ approaches to develop effective Air Quality Management Plans to efficiently achieving compliance.
(iii) In this regard, regional Integrated Assessment tools such as RIAT (Carnevale et al., 2012), LEAQ (Zachary et al., 2011), etc. with their ability to identify cost-optimised local strategies are already available to quantify the cost-effective split between further European wide measures and regional/local measures. They will inevitably need to find wider application and play an increasing role in these emerging ‘discrete islands of non-compliance’.
(iv) A further observation comes from comparing the 2025 CLE case with the 2025 A5 scenario. A5 is a high ambitious scenario (delivering 75% of the further health benefits of MTFR for the EU as a whole). At this high ambition EU level, a number of Member States are already forced to deploy all available pollution abatement measures (i.e. MTFR). Yet, from an AQ compliance perspective, this does not substantially change the picture from 2025 CLE. This points again to the key role of local targeted technical and non-technical measures in order to achieve compliance. As already noted, such measures (low emission zones, special fuels for captive fleets, captive fleet retrofitting etc.) can only be appropriately designed using ‘bottom-up’ tools.
These observations motivate the growing interest in IA models and tools for local and regional scale. Such interest is also due to the European and national directive requirements. As it is apparent from recitals 1 and especially 2 in the preamble to Air Quality Directive 2008/50/EC (AQD), European air quality legislation puts the main emphasis on protecting human health and the environment as a whole and stresses that “it is particularly important to combat emissions of pollutants at source and to identify and implement the most effective emission reduction measures at local, national and European level.” The basic principles have already been formulated in the former so-called Air Quality Framework directive (96/62/EC) and its daughter directives (1999/30/EC, 2000/69/EC, 2002/3/EC, 2004/1007/EC). The 2008/50/EC Directive encourages the use of air quality modelling, in combination with monitoring, as a scientifically relevant tool for a range of policy applications. Thus, “Air quality plans” according to AQD’s Art. 23 (formerly “Plans and Programmes”) are the strategic elements, to be developed with the aim to reliably meet ambient air quality standards considering not only the effect of emission reduction measure on ambient air quality, but considering aspects of cost-effectiveness as well. The importance of model based approaches for air management becomes apparent again in connection with Article 22 “Postponement of attainment deadlines and exemption from the obligation to apply certain limit values” commonly called “Notification for time extension”. For both, air quality plans and time extension, more elaborated requirements are formulated in Annex XV compared to former regulations. The 12 December 2011 (2011/850/EU) implementing decision already reflects this aspect, looking at the reporting obligations laid down there (Article 13, Annex II, Section H, I, J and especially K) and looking at the amount of information that has to be provided regularly. This important step forward will become real when e-reporting will enter in full operation mode, starting 1 January 2014.
Neither the Directive nor the implementing decision indicates what methodologies are required to devise efficient air pollution abatement measures.
In this context, the APPRAISAL objectives are:
I. to perform an overall review of the methodologies, from simple (e.g. scenario approach) to more comprehensive ones (e.g. full cost-benefit analysis), used in different countries to address and assess the impact of local and regional air quality plans and their health implications.
The proposed methodology to achieve this objective comprises two steps:
a. designing a database (keywords, tables and fields) in which strengths and weaknesses of the different methodologies are classified and organised around 4 main areas:
● Synergies among National, regional and local approaches, including emission abatement policies,
● Air quality assessment, (e.g. modelling, measurement, source apportionment...)
● Health impact assessment approaches,
● Uncertainty and robustness, including Quality Assurance / Quality Control (QAQC).
b. collecting and classifying the available information according to the defined common format (data entry). The database is the main tool to perform the comparison of the existing Integrated Assessment Systems (IASs) on the basis of the defined common format (systematic review) and lead to the identification of existing limitations.
II. To identify the limitations of the currently available assessment methods as emerging from the review process.
III. To design an Integrated Assessment (IA) framework, based on the information collected during the review process, for different levels of data availability, policy-maker requirements and model capabilities and to provide guidelines on how to implement the defined IA framework and to test the application of such guidelines to real cases.
IV. To communicate with key stakeholders, and in particular to policy-makers state of the art scientific knowledge on emission abatement assessment.
V. To identify key areas to be addressed by research and innovation. This has been addressed by the review document, which reports strengths and weaknesses of the various methodologies used in the MS.
To achieve such objectives, the APPRAISAL activities have been structured into 5 Work Packages (WPs).
WP1 (Coordination and Management) guarantees the coordination and management of the project and will ensure that the project provides the expected deliverables according to the foreseen time-plan.
WP2 (Review and gaps identification in Air Quality and health assessment methodologies at regional and local scale) deals with the review of methodologies to assess the effects of emission abatement policy options and measures on the reduction of atmospheric pollutants concentrations and on human health (objective (I)). It takes into account the wide range of developed and applied methods by the EU Member States in the last decade. The review work is organized in four main tasks, covering the main important aspects of the assessment activities, namely (1) synergies among national, regional and local approaches, including emission abatement policies; (2) air quality assessment, including modelling, measuring and source apportionment; (3) health impact assessment approaches; and (4) uncertainty and robustness, including QAQC. WP2 also identifies the limitations of the current assessment methods (objective (II)) and the key areas to be addressed by research and innovation (objective (V)).
The revision phase (WP2) supports the design of an Integrated Assessment framework (project objectives (III)) at regional and local scales (WP3 - Designing IA systems interconnecting national, regional, local models and strategies). The design phase addresses in general what elements need to be considered in an IA framework and in which way these elements need to be connected to design efficient policies. The IA framework design phase also defines the data structure allowing a correct communication between the functional IA framework elements. The WP3 activities lead to identify key areas to be addressed by research.
In WP4 (Guidance on integrated air quality and health assessment systems) results of the review phase (WP2) and design phase (WP3) are condensed into a guidance document that can be used to assess and improve current integrated air quality and health impact assessment modelling practices in the EU (objective (III)).
The final WP4 document provides the stakeholders with answers to questions related to the choice, the setup of an Integrated Assessment Modelling tool and the evaluation of its output, i.e.:
● What is an IA framework? When and what can it be used for? What are its essential components? What are the different types of IAS and their strengths/weaknesses? Which IAS design should be chosen to answer a given problem?
● Which inputs does an IAS require? What are recommended values or procedures to obtain these?
● What can I get from an IAS, how accurate are these results? How can I use these results for designing policies?
To underpin the guidance document, guidelines have been evaluated in practice on real test-cases.
Within WP5 (Dissemination and policy support) dissemination, communication and capacity building activities occurred during the life cycle of the project through tasks such as the establishment of a network community, the provision of publishable deliverables and the setting up of conferences. The communication to key stakeholders of the state of the art scientific knowledge on air quality assessment, as well as the support to the AQ policy revision acted as drivers throughout the entire project structure and duration (objective (IV)).
Project Results:
According to the stated objectives and activities, the project results are:
1. The online database collecting the AQ Plans currently used at regional and local scale in EU MS
An overall review of Integrated Assessment (IA) methodologies currently used in EU Member States to assess regional and local scale air quality and its impacts has been performed.
The review is structured in the online “EU Air Quality Plans and Projects Database”. It focuses on
five topics:
- how emission abatement strategies at the national (Member State) level are considered/integrated in the definition of Regional (zone) and/or Local (agglomeration) Plans and Programmes (RLPPs) to improve the air quality, and vice-versa?
- what are the current modelling techniques used to assess the air pollutant concentrations and the impacts of RLPPs (e.g. air quality models used to produce what-if scenario-analysis or full cost-benefit, cost-effectiveness integrated assessments)?
- what is the usefulness of complementary methodologies which could be used either upstream or within the integrated assessment process, particularly source apportionment?
- are we using the state-of-the-art tools, methods and knowledge to assess the impacts of air pollution on human health?
- what are the available and current-practice methods to assess modelling uncertainty and robustness?
The structured on-line database was fully open to APPRAISAL partners and stakeholders to collect and classify methodologies and systems from member states current practices and from research funded projects. The APPRAISAL consortium contacted the National contact points in European Member States and stakeholders involved in the development of air quality plans, but also the users applying models in the frame of research projects.
Air quality experts and regional/local environmental agencies were invited to fill a questionnaire, detailing the methodologies they use to build their Air Quality Plan (AQP) or research project. More than 60 air questionnaires were finally completed. In particular, almost 60% were AQPs and 30% were research projects (10% other). Almost 70% concern AQ Planning, more than 20% concern AQ assessment and only 1% deal with health assessment. About the number of questionnaire completed in each Country: only Germany exceeds 10, 9 questionnaires are from Portugal and UK, 8 for Belgium and all the other Countries are below this threshold.
TOPIC 1 - Synergies among national, regional and local approaches, including emission abatement technologies
This topic is suitable to identify how emission abatement strategies at the national level are considered and integrated in the definition of regional and local air quality plans and programs.
In this respect, the main stakeholders were interviewed and more than half indicated that these synergies are taken into account.
The analysis on the type of measure included in local AQ plans shows that a greater number of non-technical measures are considered; looking at the National Strategies, the focus appears to be on transport.
Examining the responses to the questions in Topic 1, a common difficulty appears to be how balancing and reconciling emission inventories, especially between EU, national, regional and local scales. In addition, there is a recognition that policy needs properly consider the responsibility hierarchy required in the implementation phase; the policy written without taking into account the risks connected to these factors should become “undeliverable”, both in terms of attainability of targets and in terms of an appropriate executive function or body.
TOPIC 2 - Air quality assessment and planning, including modelling and measurement
This topic investigates the air quality assessment and planning processes, with special attention to how results are obtained at different scales and how these are combined.
Figure 3: IA methodologies used by MS in the scope of air quality plans (left) and by research projects (right).
Currently, air quality integrated assessment and planning are mainly done through scenario analysis, while optimization methods are still more used in research field. Furthermore, at present, many different models are applied for air quality assessment without any standard modelling tools. Moreover, assessing the air quality, the resolution of the emission inventory and other inputs are adapted to the geographical zone under study: this indicates that, in the most of cases, scale represents an issue for using the measurement data in model evaluation.
Another crucial point concerns the challenge posed by modelling air quality at a local scale and especially the integration of these local scale results with results at larger scales.
TOPIC 3 - Health Impact Assessment (HIA)
“HIA is a mean of assessing the health impacts of policies, plans and projects using quantitative, qualitative and participatory techniques” (http://www.who.int/hia/en/). HIA addresses different methodological issues, which must be clarified for proper interpretation of results by policymakers. Different methodologies exist for developing a HIA. The aims and objectives of the assessment, data availability, resources and time-frames have an influence for the HIA methodologies used in particular assessment. With HIA, policymakers will be able to make a more balanced or scientifically weighted decision for the different policy options or scenarios. HIA is usually performed in a relative way by which different options, and interventions can be weighted against each other.
The questionnaire results pointed out that the most common approach used is the predictive approach, with time-series focused on both short-time and long-term exposure. The most frequent air pollutants included in the health impact assessments are the “traditional” pollutants, such as particles (PM10 and PM2.5) nitrogen oxides (NOx) and ozone (O3).
For accurately assessing health effects of air pollution, detailed exposure estimates need to be available. Aggregating monitored data collected by different monitoring stations, or concentrations measured at central monitoring stations do not seem to reflect the personal exposure in many cases. Analyzing epidemiological studies, it is also evident that, to fully assess the health impacts, we must take a multiple pollutant exposure approach and consider also that air pollution exposure has both physical and psychological effects. This latter dimension is less documented and is more difficult to measure. Subjective indicators constitute an appropriate alternative.
TOPIC 4 - Uncertainty and robustness
The topic analyzes model validation and uncertainty estimation describing their limitations. In particular, the key point concerns the use of model for regulatory purposes, and the different uncertainty approaches in Air Quality (AQ) Assessment, Health Impact Assessment (HIA) and Integrated Assessment Modelling (IAM), also with respect to the EU legislation requirements.
The main outcome from the analysis of the questionnaires indicates that model evaluation and uncertainty estimation is more regularly performed in air quality modelling, while it is not often applied in other components, like in HIA applications. The needs that emerged from the questionnaires relate to the quality and quantity of input and validation data, to the improvement of modelling tools and the use of better modelling practices. Many respondents reported the need for the establishment of an evaluation protocol to standardize and harmonize validation and uncertainty estimation methods in EU countries.
TOPIC 5 - Source Apportionment Methodologies
This topic analyzes the studies on source apportionment in Europe and their increasing number in recent years, closely related to the development of improved tools with better functionalities and performance. However, the lack of a European network of aerosol pollutants monitoring sites is becoming a limiting factor for a further growth and consolidation of source apportionment techniques. The partial information on European sources emission factors and fingerprints is limiting sources identification.
Moreover, the definition of European methodological protocols to guarantee a minimum level of quality and to make results from different studies comparable is required.
From the methodological point of view, both Receptor models and Eulerian models appear as the most dynamic areas in the source apportionment field.
The most important source categories to target in order to abate exceedances of air quality limits are traffic, (e.g. nitrogen oxides), agriculture (e.g. ammonia) and biomass burning during the cold season.
According to these results, future studies on source apportionment in Europe should focus on the relevance of gas-to-particle conversion and photochemical processes as the main contributors to recalcitrant pollutants like PM and ozone.
2. The analysis of the limitations of the currently available assessment methods
The systematic review of the AQ plans and projects collected in the APPRAISAL database allowed to identify the limitations of the current methodologies.
● A recurring theme is the challenge that is being posed by local scale modelling, especially with respect to the integration of these local scale results within larger scale ones. This is certainly true also for the integrated assessment tools, which currently are still lacking when considering local scale integrated assessment. Further research should be devoted to this topic.
● When speaking about weaknesses in IAM, the main issue (reported in 70% of the studies) is related to the emission inventories. Even if standard emission inventory (such as EMEP) are available at the European level, these are still lacking when moving to local scale inventories and the consistency between the emissions used at different scales is a major concern.
● For air quality measurements the question was raised on how model results can be combined with observed values in scenario calculations.
● In the overall IAM framework, source apportionment methodologies can bring added values. In fact, in recent years the number of studies on source apportionment in Europe has steadily increased. This is closely related to the continuous development and improvement of tools with respect to functionality and performance. Nevertheless, the lack of an appropriate European network of urban monitoring sites with detailed chemical and physical characterization of aerosols is becoming a limiting factor for a further growth and consolidation of SA techniques. This is true for both receptor models and Eulerian models. From the methodological point of view, a combination of both receptor models and Eulerian models possibly complemented with other techniques (mainly Lagrangian models) appears as the most dynamic area in the further development of source apportionment tools.
● The IAM optimization approach fully responds to the AQ Directives. The emission reduction measures are selected by an optimization algorithm that assesses their impact on air quality, health exposure and implementation costs. Such optimization algorithms require thousands of air quality assessments; in these cases, AQ systems cannot directly be used because of the high computing time demand, so they are used to provide a limited number of simulations that are then processed to identify ‘simple’ emissions-AQ links (source-receptor relationships/models) that are able to capture the specific features of a region.
● There is a need to establish an evaluation protocol in order to standardise and harmonise validation and uncertainty estimation methods in EU countries. Within an IAM framework, model evaluation and uncertainty estimation are more regularly performed in air quality modelling, while they are rarely applied in other IAM components such as, for example, in the health impact assessment. Operational and diagnostic evaluations are the methods preferred both in the case of modelling for the purpose of air quality planning as well as for research projects. For the purpose of air quality plans, expert judgement is also frequently used. Uncertainty propagation methodologies are used, although not so often, to quantify confidence levels of air quality model results.
3. The design of an IA framework interconnecting national, regional, local models
An Integrated Assessment System framework, at the regional and local scale, has been designed starting from the results of the revision phase. It is based on the EEA DPSIR scheme. The building blocks and the connections among them are represented in Figure 4.
Figure 4: the DPSIR scheme adapted to IAM at regional/local scale.
The meaning of each block is as follows (quoting from EEA glossary):
- DRIVERS: this block describes the “action resulting from or influenced by human/natural activity or intervention”. Here we refer to variables (often called “activity levels”) describing traffic, industries, residential heating, etc...
- PRESSURES (Emissions): this block describes the “discharge of pollutants into the atmosphere from stationary sources such as smokestacks, and from surface areas of commercial or industrial facilities and mobile sources, for example, motor vehicles, locomotives and aircrafts.” PRESSURES depend on DRIVERS, and are computed as function of the activity levels and the quantity of pollution emitted per activity.
- STATE (Air Quality): this block describes the “condition of different environmental compartments and systems“. Here we refer to STATE as the concentrations of air pollutants resulting from the PRESSURES defined in the previous block. In IAM implementations, STATE can sometimes be directly measured, but more often it is computed using some kind of air quality model.
- IMPACT: this block describes “any alteration of environmental conditions or creation of a new set of environmental conditions, adverse or beneficial, caused or induced by the action or set of actions under consideration”. In the proposed framework, we refer to IMPACT on human health, vegetation, ecosystem, etc... derived by a modification of the STATE. Again the calculation of the IMPACT may be based on some measure, but normally requires a set of models (e.g. health impacts are often evaluated using dose-response functions).
- RESPONSES: this block describes the “attempts to prevent, compensate, ameliorate or adapt to changes in the state of the environment”. In our framework, this block describes all the measures that could be applied, at a regional/local scale, to improve the STATE and reduce IMPACTS. In particular, 3 different group of responses can be considered:
o Energy/Structural measures, impacting on the DRIVERS block and including both energy efficiency and population behaviour measures.
o End of Pipe measures, impacting on EMISSIONS and related to the introduction of new technologies affecting the level of emission for each activity.
o Technical measures, impacting directly on STATES and including technological improvements allowing for direct concentration reductions (e.g. ecological tarmac/concrete).
Each of these blocks is analysed in terms of input, functionality, output, synergies among different scales and uncertainties related to the block.
Drivers
The basic function of the DRIVERS block is to model the development of key activities (i.e. road traffic, residential combustion, centralized energy production/industry, agriculture) over time. For instance, for traffic, it means to analyze and/or project the fuel consumption due to traffic, the driven km by cars, the current and future fuel, etc. This information is then used as an input to compute emissions (PRESSURES).
More in general, the input parameters for drivers are factors that represent causes of emission-wise essential activities. Important input parameters include population, general economic activities (e.g. in the form of GDP), more specific activity factors (e.g. sector specific production intensities, transport demand, energy demand, etc.) and technology change factors (e.g. vehicle stock structure, energy efficiency of buildings, etc.) that may be driven by international, national or local requirements or goals (RESPONSES block) or “natural”, non-forced development.
Pressure
Air pollutant emissions act as pressures on the environment. Thus, the PRESSURES block corresponds to the computation of the quantity of pollutants emitted into the atmosphere from stationary sources (such as stacks of large industrial facilities, e.g. power plants), surface areas (residential, commercial or industrial facilities), and mobile sources (for example locomotives, aircrafts, ships, cars, etc.).
The emission of a pollutant at a source (emitter) can be measured (as in large point sources) or estimated. In the last case, it is generally calculated as the product of the activity of this emitter (DRIVERS) and an emission factor, which is the quantity of pollutant emitted per unit of activity.
State
In the case of air quality, the STATE describes the ambient concentrations of targeted pollutants. In the case of traffic, it means to consider the air concentrations due to the emissions from cars. The air quality state can be described as gridded concentrations over the studied area, or as local concentrations on receptor sites, depending on the objectives and on the available tools. Also, in addition to the spatial dimension, the air quality state has a temporal dimension, considering that a pollutant can be monitored/ modelled with a temporal resolution of hours/days, etc. Once concentrations are estimated in space and time with different available approaches, one can calculate various air quality indicators, such as aggregation of the initial air quality data e.g. to provide the number of daily exceedances of particulate matter concentration on a cell, the annual mean of nitrogen oxides aggregated over a domain, etc. In this document we will focus on concentrations as a state indicator, even if the content would be basically the same for deposition of atmospheric pollutants.
Impact
IMPACT describes the consequences of any alterations or modifications of environmental conditions related to the STATE of air quality, being either beneficial or adverse. Among the various impacts, we could distinguish between impacts on human health, on environment (vegetation and ecosystems), on social and economic aspects, on climate or on visibility. Moreover some impact could be derived from another, such as economic valuation of human health or of ecosystems. This means, for instance, to evaluate a monetary equivalent of the increased morbidity/mortality related to traffic emissions. In this context, we specifically focus on impacts on human health. In general, the following input is needed to compute IMPACTS:
- air pollution concentrations;
- population data;
- dose/concentration-response functions.
Responses
The RESPONSES block represents the decision framework, that is to say the set of techniques/approaches that can be used to take decisions on actions (mostly, emission reduction measures) to be implemented.
The main components of a decision framework are defined considering: control variables (these represent, for instance, the emission reduction measures that can be applied by the regional/local authority, as car-sharing, congestion charge, etc.); objectives (these represent what a Decision Maker would like to improve); constraints (these can be of different types, as legislative, economic, physical, etc. They can be mathematically formalized, if using a formal approach to take decisions, or they can be taken into account when making decisions, but without explicitly modelling them).
The designed Integrated Assessment framework classifies (in broad terms) two possible decision pathways:
● scenario analysis (Figure 5, a). This is currently the most commonly used approach to design “Plans and Programmes” at regional/local scales. Expert judgment or Source Apportionment are used to identify potential emission reduction measures, the effects of which are then tested by running a regional/local air quality model. This approach has the advantage of simplicity but does not guarantee that cost-effective measures are selected. An evaluation of costs and impacts can only be performed “ex-post”.
● optimization (Figure 5, b). This approach identifies cost-effective measures to improve air quality by solving a mathematical programming problem. During such a process, abatement costs and impacts are continuously compared until the least costly set of measures is found, so that a given air quality target is met (or vice versa, minimizing the Air Quality target given a maximum available budget). This approach guarantees the selection of the set of policies that are cost-effective for a given domain and, in principle, can incorporate all costs and impacts in the optimization procedure.
Figure 5: the DPSIR scheme adapted to IAM at regional/local scale. The red arrow in the Figure represents the “feedback on cost-effectiveness”, provided by the optimization approach.
4. The taxonomy of Air Quality plans
The IAM framework can be used to classify air quality plans according to the complexity of the available data and the methodologies used. An overview of these different levels and their main characteristics for the different blocks is given in Table 1.
Table 1 Different levels of detail for the different DPSIR blocks.
The analysis of individual AQPs can be summarized using a radar chart. For each block of the framework, five levels of complexity have been defined: N/A – information not available, Level 0 – the block is not considered in the AQP, Level 1 – low level of complexity in the implementation, Level 2 – medium level of complexity, and Level 3 – high level of complexity.
The radar chart in Figure 6 represents the “average graph” computed considering all plans available in the database.
Figure 6: A radar graph representing the average complexity level of AQ plans.
5. AQ policies uncertainty assessment
Uncertainty and Sensitivity analyses are key issues in the definition and evaluation of emission control strategies. An AQ plan should include information about the source of uncertainty related to each part of the IA modelling system used to define it. Such assessment is a very complex task due to the fact that an IAM is implemented by a set of integrated models, whose input are usually uncertain and whose output can become the input of the next stages. This fact makes the analytical and partial derivative related methods for both uncertainty and sensitivity analysis quite impossible to be used on the overall system, and sometimes also for a portion of it.
In the next subsections, a brief discussion about the specific uncertainty sources for each block of the IAM framework is presented.
Main uncertainties of DRIVERS block components are:
● Road traffic: Traffic models and/or detailed road segment specific traffic information are relatively commonly available. Technological parameters are relatively well known at least at national level. Parameters required for reliable non-exhaust emission assessment (e.g. road surface type and condition) can be a considerable source of uncertainty.
● Non-road traffic and machinery: For some forms of non-road activities, e.g. sea vessels, trains and airplanes, activities and spatial patterns are often relatively well known. For many other forms of machinery, in contrast, the activity data can be much more uncertain.
● Residential combustion: Residential wood combustion activities and technology information are often uncertain because a lot of the wood fuel is used from private stock directly, and house-hold level heating system stock is poorly known. Furthermore, spatial assessment (i.e. gridding) of residential combustion activities is often uncertain because of the lack of building registers for residential heating appliances.
The uncertainties associated to emissions inventories (PRESSURES) are directly related to accuracy. This accuracy can be split into two main contributions:
- structural inaccuracy, which is due to the structure of the inventory. The structural accuracy estimates the inventory structure ability to calculate as precisely as possible the real emissions. This uncertainty can be split into 3 contributions
o inaccuracy due to aggregations: the emissions are calculated on defined spatial and time scales and for some of them these scales are different from the real emissions ones. This can be due to lack of information on the emission processes or on the variability of the real emissions;
o incompleteness, which means that an emission inventory may be inaccurate due to the absence of emission sources because of a limited understanding of the emission processes;
o inaccurate mathematical formulation and calculation errors: the mathematical formulation used is generally inaccurate (simplified), for example by considering that the relation between emission and activity is supposed linear, which is generally not true.
- inaccuracy on the input data (i.e. activity data, emission factors).
The uncertainties on the input data are mainly due to the lack of information on the different parameters used to estimate the emissions of an inventory.
When the AQ STATE is evaluated through measurements only, uncertainties are related to the measurements themselves, to the geo-statistical methods used to interpolate point measurements and to the representativeness of measurement sites to characterize the area under study. Intrinsic uncertainties of AQ modelling are mainly related to errors in the physical formulation of the model, and to uncertainties in the input data. An operational validation of the AQ model by comparison with measurements is required, opening the question of the representativeness of the chosen measurement sites in relation to the model scale. Evaluating the indefiniteness of prospective study is more challenging and would require the use of diagnostic evaluation (e.g. sensitivity tests) or probabilistic evaluation (e.g. errors propagation). Furthermore, as mentioned earlier, for prospective IAM, estimating the AQ state over a relatively short temporal period (up to one year) introduces uncertainties on the representativeness of the AQ state itself.
Health IMPACT analysis relies on two main processes, namely exposure assessment and epidemiological analysis relating exposure to the health outcome. These two processes include a number of basic steps, finally leading to the quantification of the expected atmospheric pollution induced health burden in the target population, most commonly expressed in terms of years of life lost attributable to the exposure to the atmospheric pollutants under study. Assumptions and uncertainties related to each process may significantly influence the result of the analysis. The main sources of uncertainty in HIA studies can be summarised as follows:
● Uncertainties in estimating the impact for each health outcome. This uncertainty is mainly related to the health-outcome frequencies data. Mortality may be considered generally accurate, but frequency measures of morbidity and data on health-care systems contain uncertainties. Furthermore, in contrast to directly countable events listed in national health statistics (eg, deaths or injuries due to traffic accidents), it is not possible to directly identify the victims of mixtures with cumulative toxicity, such as smoking or air pollutants. Also, the health outcomes may not be specifically linked to air pollution due to synergistic effects with other factors.
● Uncertainties in exposure assessment. Poor exposure assessment is an important source of uncertainty in HIA and can result from errors and biases in either air quality models or in exposure models. The different sources of error and uncertainties in the exposure models result from variability not modelled or incorrectly modelled, inaccurate inputs, errors in coding, simplifications of physical, chemical and biological processes to form the conceptual models, and flaws in the conceptual model. Emission and meteorological input data accuracy and physical/chemistry assumptions and parameterisations in the air quality model largely affect the reliability of its results on the spatial distribution of ambient pollutant concentrations. Furthermore, statistical methods (e.g. kriging) used to produce higher resolved air pollution fields starting from air quality model results and other inputs (local observations, emissions etc) may also introduce uncertainties at specific locations far away from the observations.
● Uncertainties related to the concentration-response functions, estimated by epidemiological models: Some of the formal approaches for uncertainty analysis in epidemiological concentration-response models include Bayesian analysis, Monte Carlo analysis and model intercomparison.
● Uncertainties concerning the temporal scale of effects, i.e. the latency times from exposure to adverse event. This is an uncertainty mainly associated with long-term exposure studies, as acute effects follow exposure by a few days.
It is important that RESPONSE decision approach focuses on robust strategies, that is to say on “policies that do not significantly change due to changes in the uncertain model elements”. This issue is linked to the need of defining a set of indexes and a methodology to measure the sensitivity of the decision problem solutions. It is in fact worth underlining that, while for air quality models the sensitivity can be measured by referring in one way or the other to field data, for IAMs this is not possible, since an absolute “optimal” policy is not known and most of the times does not even exist. The traditional concept of model accuracy must thus be replaced by notions such as risk of a certain decision or regret of choosing one policy instead of another.
6. Guidance document to support the IAM design
The guidance document, focuses on the different topics that need to be addressed to set up an AQ Integrated Assessment Modelling (IAM) system or methodology.
The document has been structured around the building blocks of the EEA DPSIR scheme. It does not provide a single set of recommendations but a range of solutions, so that each of the building blocks can be elaborated to a different level of detail, adapted to the data and tools that are available in practice or what is required to solve the specific air quality problem that is targeted by the IAM (‘fit for purpose’). The different topics of the document are addressed through a number of questions. Some of these questions are specific to the building block while others relate to topics that reappear in all blocks such as how the uncertainty or the interaction between scales should be tackled.
The guidance document has been evaluated on existing AQPs with the aim to identify which guidance was currently lacking and should be further extended .
Eight case studies have been analysed in detail and are listed below:
• AQP for Antwerp
• AQP for Athens
• AQP for the Northern Region of Portugal
• Preliminary AQP for Emilia Romagna
• Research project for the Alsace Region
• AQP for the Warsaw Agglomeration
• Research project for the Warsaw Agglomeration
• AQP for Helsinki
The scheme to explore these ‘practical applications’ has been always the same: the AQP was interpreted in terms of the DPSIR blocks, and a level of detail was assigned to the Drivers, Pressures, State, Impact, Response, according to the way they have been analysed. Then, possible improvements to the DPSIR blocks used were derived together with (possibly) missing or underestimated topics in the guidance document.
In all the considered cases (Figure 7) the drivers and, to a somewhat lesser degree, the pressures and state are well elaborated, while in none of the studies the health impact has been considered with the highest level of detail and also the choice of abatement measures has mostly been based on expert judgment.
Figure 7: Summary of complexity levels for the eight studies considered.
Based on the analysis for the AQP studies the following observations can be made:
● Much effort is put into quantifying the drivers and pressures (emissions) in all the studies that were considered. In all examples, the drivers are treated at the highest level of complexity and only in one of the examples the emissions are based on a top-down and thus ‘low level’ methodology. Regardless of this already high level of detail, the drivers and the emissions are still seen by some studies as the topic where most of the additional guidance would be welcome; and more in particular the guidance document should be extended with practical references to additional data sources with an emphasis on: (i) ‘real life’ emission factors for traffic, domestic heating (wood/coal burning appliances) and agriculture; (ii) projections and future estimates and (iii) practical examples on combining emission inventories at different scales.
● The state (concentrations) is in all the examples handled with one or more models. For some of the AQPs that lack detailed local scale concentration assessments, it is also acknowledged that this is a point for improvement. The document needs to be extended with guidance on using data assimilation and a better reference should be made to FAIRMODE initiative for the evaluation of model results.
● Uncertainty, both in the emissions and concentrations, is not addressed except in a single research study, which was included in this analysis. Based on the feedback given from the 8 cases, there is however need for additional guidance on how to perform such analysis.
● Even if health impact is considered in most of the examples, this is never done based on a detailed temporal and spatial resolution for the exposure and population data (in three of the plans health impact is not addressed at all). Also for this topic more practical guidance would be welcome with references to real case applications and to where input data (dose-response, ..) can be found. As this is a subject unfamiliar to most of those involved in developing air quality plans, some of the topics addressed in the guidance document should be better explained.
● Only the two studies using the RIAT+ modelling system rely on a multi-objective optimisation method to identify the optimal mix of abatement measures. It can however be noted that for both applications, even if they imply the highest level of complexity for the response block, some of the identified improvements relate to the database, e.g. the costs and emission reduction efficiencies for the abatement measures or the weights attributed to single objectives when doing a multi objective optimisation. It should be clear that – as in any modelling activity – the results of the optimisation process heavily rely on the quality of the input data that are used in this process.
7. The design and the assessment of AQ plans for Porto and Brussels regions
The final step of the APPRAISAL project was the application of the IAM tool RIAT+ in two test cases: one in the Brussels Capital Region in Belgium and in the region of Porto in the North of Portugal. The experience obtained through these two test cases was used to further improve the guidance document.
The RIAT+ system, developed during the OPERA project (www.operatool.eu LIFE09 ENV/IT/092, Best LIFE Environment projects Award 2014), is an IAM tool designed to help regional decision makers to select optimal air pollution reduction policies that will improve the air quality at minimum costs. To achieve this, the system incorporates explicitly the specific features of the area of interest with regional input data-sets for the:
● precursor emissions of local and surrounding sources;
● abatement measures (technical and non-technical) described per activity sector and technology with information on application rates, emission removal efficiency factor and cost;
● the effect of meteorology and prevailing chemical regimes through the use of site specific source-receptor functions.
The system runs as a stand-alone desktop application and can be downloaded (free license) from the project web-site (http://www.operatool.eu/download/). RIAT+ has been already applied in Emilia-Romagna Region (IT) and in Alsace (FR) during the OPERA project and in Lombardy Region to evaluate AQP measures.
The tool allows the two possible decision pathways already mentioned in the DPSIR scheme: scenario analysis and optimisation. The main outputs from RIAT+ are a summary of emission reductions on the domain, a table of the application rates for the different measures, maps of a set of relevant air quality indexes (AQIs) and, for the optimization pathway, the Pareto Curve providing the efficient solutions of a specific AQI ranked by costs.
A source receptor (S/R) model is used, internally, to link emissions (pressure) to an air quality indicator, that represent the state. A S/R model can be as simple as a linear relationship, or as complex as a chemical transport model (CTM). To limit the computational time, RIAT+ normally uses a nonlinear relations identified by means of Artificial Neural Networks (ANNs), tuned to replicate the results of a limited set of simulations performed by the users with deterministic air quality model calibrated for the specific site.
The selection of which simulations to use is an important aspect in setting up the ANN. They must be representative of the range of emissions/concentrations that can be encountered during the optimization procedure. The definition of this training data set is typically referred to as the ‘Design of Experiment’: this establishes the configurations for the CTM simulations.
Both the test cases are presented below with the same scheme: a quick introduction on the area and the proposed abatement measures, followed by the application of RIAT+, with a focus on the technology database used, the chemical transport modelling runs, the Design of Experiment and the identification of the source-receptor models and, in the end, the results obtained.
Bruxelles case study
The Brussels Capital Region (BCR) has an area of 161 km2 and is home to more than 1.1 million people. The region consists of 19 municipalities, one of which is the Brussels Municipality, the capital of Belgium (Figure 8).
Figure 8: Location of the BCR (red zone) in Belgium.
The proposed abatement measures were provided by BIM (Brussels Environment, http://www.ibgebim.be). They are a list of 13 measures consisting of 9 traffic measures and 4 domestic heating measures, that have been approved by Brussels authorities. For these abatement measures, BIM provided order-of-magnitude estimations of the costs and emission reductions.
The RIAT+ database with abatement technologies that are available for the macro-sectors of interest – non-industrial combustion (2) and transport (7) – was derived from GAINS Europe (http://gains.iiasa.ac.at/gains/EUN/index.login?logout=1) and the OPERA project database (www.operatool.eu).
For air quality modelling of the BCR, the AURORA chemical transport model (see: http://pandora.meng.auth.gr/mds/showlong.php?id=167) was used with a domain of 49 x 49 grid cells at 1 km resolution (Figure 9) with base emissions for the year 2009 and a hourly time step. For the vertical discretization, 20 layers were used for a domain extending up to 5 km.
Figure 9: Model grid used for the CTM calculations.
The results model setup were validated by comparison to the observed values at the measurement stations inside the model domain. For the model validation, the FAIRMODE methodology (http://fairmode.jrc.ec.europa.eu/) was adopted.
For the Design of Experiment phase, three levels of emission reduction were distinguished: base case (B), high emission reductions (H) and low emission reductions (L). In order to determine the emission reduction scenarios for the ANN training, the three levels B, H, L were combined to produce 14 emission scenarios. These scenarios were applied to the emissions both in and outside the policy application domain (PAD), which, in this case, is the BCR.
In this study, the AQIs considered are:
● PM10: yearly average of PM10 concentrations
● NO2: yearly average of NO2 concentrations.
RIAT+ was then run with this S/R model to look for optimal policies beyond the 2020 Current LEgislation. As the emission changes, that can be obtained with the selected set of measures, are limited, unsurprisingly, the concentration changes are also limited. Figure 10 shows an example of NO2 concentration changes due to the emission abatement measures.
Figure 10: Yearly average NO2 concentration changes (µg/m3) for all traffic and all non-industrial heating measures as well as for the combination of these two in 2020 compared to the reference (CLE 2020). The number in parentheses is the maximum concentration change.
Porto case study
The Great Porto Area is a Portuguese NUTS3 subregion involving 11 municipalities. It covers a total area of 1024 km2 with a total population of more than 1.2 million inhabitants (Figure 11).
Figure 11: Location of the Great Porto Area in Portugal and in the Northern Region of Portugal.
The GAINS database of reduction technologies, which contains a large data set collected for Portugal by IIASA (http://www.iiasa.ac.at) was used. It contains values for the years 2010, 2015, 2020 and 2025, including costs and emissions effects. The reference scenario «TSAP» of March 2013 was considered in the study.
The TAPM model (http://pandora.meng.auth.gr/mds/showlong.php?id=120) was applied to the Great Porto Area (150x150 km) for one entire reference year (2012) with a 2x2 km2 spatial resolution using disaggregated emissions from the Portuguese 2009 emission inventory, which is the most recent available. Extending previous evaluations of the same model for the Great Porto Area, the results of the TAPM simulations were compared to the measured values at the monitoring stations inside the model domain. As in the Brussels case, the FAIRMODE methodology was used for the validation (http://fairmode.jrc.ec.europa.eu/).
The Design of the Experiment was influenced by the computational time constraints, so just 10 emission reduction scenarios were simulated. Processing the TAPM simulation results, low and high concentration levels were obtained. Each TAPM simulation is a full year with a hourly time step. The target considered in this application was the PM10 annual mean.
RIAT+ was applied in the optimization mode and Figure 12 shows the efficient solutions obtained for the Great Porto domain. On the horizontal axis of the figure, there are industrial costs (i.e. those related to the implementation of end-of-pipe measures), considered over CLE (i.e. those mandatory under the current legislation) and expressed in M€. On the vertical axis, there is the spatial average of PM10 annual mean (the AQI value for this particular case) estimated over the entire study area.
Figure 12: Pareto curve for the optimization of PM10 yearly mean concentration.
The Pareto Curve shows that a PM10 mean concentration of 28.8 μg/m3 can be reached by adopting emission reduction technologies costing around 7.6 Million € per year, but shows at the same time, that a mean concentration below 28 μg/m3 cannot be attained even investing a very large sum of money in local reductions. To further improve air quality, a much larger application domain must be considered for reduction measures.
For the point of the Pareto curve mentioned above, Figure 13 shows the spatial distribution of average PM10 annual concentration values. The largest reduction of PM10 emissions and concentration levels are expected over the Porto municipality where the population has the highest density.
Figure 13: Mean PM10 concentration resulting from RIAT+ application (point C of the Pareto curve).
Main comments to the test cases
The main aim of these applications was to confront the practical application of RIAT+ with the guidance document.
The following main conclusions can be drawn:
- In practice, the list of options for abatement measures is restricted not only by what is technically and economically feasible, but possibly even more by political and social acceptance. IAM tools should therefore be extended to allow their users to take into account the implications of political and social acceptance at an early stage of the decision process.
- The applications demonstrate that tools exist which can be practically applied in an integrated assessment of air quality that does not only consider compliance of concentration to limit values but also efficiently take into account internal and external costs (e.g. health impact) of different available abatement options.
- The biggest task when implementing such a comprehensive IAM is - as is also the case in regular air quality modelling applications – to obtain high quality input data i.e.. information on local emissions and the cost and effectiveness of possible abatement measures. When such data are lacking, one can still rely on existing European inventories and databases with data on abatement measures such as EMEP and GAINS well keeping in mind the assumed validity of such data for the region of interest and the implications for the results obtained using the IAM.
- If an IAM system uses source-receptor relationships (artificial neural networks, linear regression, ...) to relate emission changes to concentration changes, such relationships should be carefully tested to ensure that they not only correctly replicate the concentration values obtained through more complex modelling tools (e.g. CTMs) but also capture the dynamics i.e. the concentration changes calculated by the model for which they are a surrogate.
In the Brussels case, a lot of time was put into estimating precise measures while the impact on air quality of these measures is rather limited due to the dimension of the area selected. A first screening step such as a simple scenario to check the importance of the impacts should be done before using a complex methodology as the latter has limited added value in such cases.
In the Porto case a list of available technologies from an existing database was used and the main sectors were selected and identified. Nevertheless, a more detailed list of locally available measures needs to be decided and discussed with stakeholders and policy makers. With the optimization approach, it was possible to have a first idea of the optimal investment costs and benefits to achieve a given PM10 air quality objective, and to identify the complete range of their feasible values.
8. The identification of the key research areas
The key areas to be addressed by research have been identified and organized following the DPSIR scheme building blocks.
Drivers (Activities)
Considerable deficiencies were identified for the DRIVERS block, for all activity sectors contributing to local scale emissions, not only for power plants, road traffic and residential combustion, but also for agriculture, non-road traffic and machinery.
An important future research line should be devoted to the integration of activity inventories at different spatial scales. A further key issue for future research is related to activity evolution. On the one side, one would certainly like to improve the estimation of how local economic sectors will develop and adapt in the future, taking into account both internal factors, such as economic downturns, and external ones, as climate changes. This means considering new land use policies (activity location) as part of the IAM problem. On the other side, since a perfect prediction of activity evolution is out of question, new methods to deal with uncertain predictions (ensemble modelling, risk aversion,...) have to be developed and possibly become standard.
Pressures (Emissions)
In the IAM database collected by APPRAISAL, 70% of the respondents identified emission values as the main weakness of their modelling approach. Quantifying the effectiveness of specific abatement measures within a zone presumes that the emission inventory is disaggregated with sufficient details both spatially and per categories to properly consider the emission abatement measures. This level of detail is unfortunately lacking in inventories, leading to unreliable estimates of the effect of measures; i.e. the official national and European (EMEP) emission inventories only contain detailed (per category) emission totals for the member state as a whole, or alternatively, gridded data with only SNAP 1 detail. So this issue can cause, as already said, inconsistencies when integrating data coming from different scales.
More in detail, key areas to be investigated by research and innovation concern various topics:
- emission inventory harmonization;
- emission scenario projection, emission factors estimation.
State (Concentration levels)
Key areas to be addressed by research and innovation, in the STATE module, are the following:
- integration of ground-based and remote-sensing monitoring methods, to precisely assess the “current” AQ situation;
- better understanding of composition of the various PM fractions;
- regional and local scale CTM implementation and validation;
- source-receptor model identification approaches.
Impacts (Human health)
In such field the key areas to be addressed by research and innovation are:
- health impact assessment integration in IAM;
- exposure-response functions;
- acute effects and short-term impact on mortality;
- mortality and morbidity factors due to long term NO2 and O3 exposure.
Responses (Methodologies to design measures)
Most of the APPRAISAL recommendations on key areas for future research and innovation are related to the decision process (or RESPONSES block within the DPSIR cycle), mainly considering:
- Development of more detailed information on abatement measures.
- Inclusion of socio-economic aspects in the analysis.
- Better account for “ Energy Efficiency / non-technical measures”. These measures are now limited to scenario analysis because of the difficulty to estimate their associated costs. Still progress is required to include these measures in optimized IAMs.
- Multi-scale interactions: as already done for Chemical Transport Models, there is the need to develop IAMs nesting capabilities (one-way/two-ways) to include EU/national constraints in regional analyses, and at the same time providing feedbacks from the regional to the EU/national scale.
- Provide guidance for developing and using IAM approaches and favour harmonization.
- Better integrate Air Quality and Climate Change policies in the IAM framework. In a “resource limited” world, efficiency (to get win-win solutions from Air Quality and Climate Change actions) is of extreme importance and requires guidelines to better integrate climate change policies (normally established at national or even international levels) into air quality plans (developed at regional/local level).
- Develop dynamically evolving impact assessments. Current approaches aim at solutions to be reached at a given time horizon and are therefore “static”. But the system is non-stationary (see the effect of the current economic crisis) and it becomes important for decision makers to know where to invest in priority with potential to adapt decisions with time.
9. The contribution to the Air Quality review
One of the main results of the project is the contribution to the AQ review process.
In general the Air Quality Directive 2008/50 and more recently the Commission Implementing Decision 2011/850 do not specify what methodology is required to devise efficient measures to improve the quality of the air. The contents of the template provided for reporting however indicate that a scenario approach supported by source apportionment can be useful addressing the following:
1. Source apportionment: Which are the main emission sources responsible for the pollution, distinguishing local and regional (transboundary) contributions? With which accuracy is the emission source base case known?
2. Air quality assessment for the current situation: In which zones (location, type) are exceedances of the limit values of a pollutant observed and how large is the population that is exposed?
3. Air quality assessment for future years or emission scenarios:
● What is the baseline level i.e. the concentration to be expected in the year when the limit value comes into force without any measures beyond those already agreed or implied by existing legislation.
● Which measures are currently in place beyond those required by current legislation and what is their effect on the air quality?
● Which additional measures are planned and what is their effect on the air quality?
With respect to emissions, the Directive 2008/50/EC requires an air quality plan reporting the origin of pollution (Annex XV) by providing a list of the main emission sources responsible for pollution (map) and the total quantity of emissions from these sources (tonnes/year). The Commission Implementing Decision of 12 December 2011 requires the AQP to report on the emission scenario and the total emission for both the baseline and for the projection as well as the reduction in annual emissions due to the applied measures.
In fact, the Directive acknowledges the importance “to identify and implement the most effective emission reduction measures at local, national and Community level” (article 2). This presumes that:
1. the emission inventory used for the AQP is sufficiently detailed to allow mapping measures to the specific emissions managed at the different administrative levels that have to be considered.
2. the costs of emission reduction technologies are available.
3. a suitable optimization approach to select effective policies can be implemented.
Emission inventories and projections as needed for the assessment and planning at the local scale are currently developed ad-hoc. It is recommendable to take an initiative to harmonize the criteria and the procedures for developing such local emission inventories. Further fixing and specifying these procedures might improve emission data necessarily needed for air quality modelling and in consequence will improve modelling results for this part.
Moreover, the effectiveness of any type of remediation measure strongly depends on the reliability of the pollution source identification and quantification process. Hence, the use of methodologies with minimum biases and uncertainties certainly contributes to focusing valuable resources and time on the most contributing sources in the area of interest.
Article 25 of the Directive deals with the problem of transboundary air pollution. To be effective an air quality plan should appropriately take into account the contribution of sources outside the zone considered in the plan. This is especially true for long lived and secondary pollutants and where the zone that is modelled is small as in local and street level models. In those cases larger scale modelling is needed to properly incorporate the effect of the boundary conditions or at least a sensitivity analysis should be required to quantify the importance of the boundary conditions. If results at different scales are combined, the consistency of the inputs used should be checked and care should be taken to account for differences between the models.
On the other hand the problem of transboundary air pollution can be read as the issue to assess the impact of regional-local emissions, in other words, to quantify the effective potential of regional-local policies in a specific domain. Methodologies should be formalized and developed to fill this gap.
Integrated Assessment Modelling (IAM) should support the air quality authorities in selecting efficient mitigation strategies by providing tools for assessing and solving air quality planning problems at different spatial scales. AQ modelling is the IAM component explicitly mentioned in EU legislation. The Directive recognizes that modelling can be used in combination with measurements to obtain a better representation of the spatial distribution: “Where possible modelling techniques should be applied to enable point data to be interpreted in terms of geographical distribution of concentration” (Article 6). As population density is not necessarily homogeneous within a zone, the air quality plan report could be improved by replacing the single values for the concentration and population within the zone in the report by a map showing the spatial distributions for the concentration and the population.
As the role of modelling in understanding the influence of physical and chemical processes on the dispersion and transformation of pollutants is increasingly being recognised, and MS are already using models in their current assessment techniques it is recommended to further promote the use of modeling tools in the scope of the nowadays AQD revision. Moreover, there is no alternative to modelling for assessing the effectiveness of emission reduction measures in future years. Thus, modelling should become an essential part of air quality planning and any such modelling based report should include a complete description of the model and inputs used as well as an evaluation to quantify the reliability of the AQ assessment.
Today many different modelling tools exist that are being used for AQ assessment and planning so that there is currently no obvious standard model that could be imposed as a ‘preferred’ model for each of the different scales and pollutants considered. Preferably however the model as a whole or at least its subcomponents should have undergone a scientific peer review or a report should exist in which the model has been submitted to a diagnostic analysis.
The need to incorporate uncertainty estimation in air quality modelling is also recognised by policy makers and is required by the AQD, which specifies modelling quality objectives. As the directive does not provide guidelines on how to carry out model evaluation to achieve the quality requirements imposed, the development of relevant guidelines is necessary for modellers and authorities. Several attempts have been made for the establishment of uncertainty assessment guidelines within a number of projects, including AIR4EU and FAIRMODE. The Guidance Document that was elaborated within FAIRMODE is the current reference point for model users and regulators to ensure that their air quality model meets the quality criteria required by EU legislation.
Another important issue for a proper model application when developing an AQP, which is currently not covered by the AQD, concerns the representativeness of the simulation period for air quality planning. Currently an AQP usually relies on the model results for a single meteorological year. This year is often the year for which the exceedances were observed that triggered the demand for an AQP in the first place but which is not necessarily a year that is representative for the time horizon of the AQP. An alternative could be to select a single year (or years) based on criteria (e.g. typical meteorological year, critical meteorological conditions year) that assure it represents the air quality for the time horizon of the AQP. Another option could be to use a longer multiyear meteorological period for the AQP modelling work. The latter could however well be impractical considering the increase in computer time required to model such long periods. In conclusion, it is necessary to account for meteorological variability in air quality modelling and IAM since meteorology is a constraint that influences the effectiveness of emission reduction measures to some extent, though meteorology is not the primary cause for air pollution. This is generally addressed for in EU scale IAM studies (e.g. IIASA GAINS approach) but seems lacking at the regional/local scale, most probably due to the limited resources available.
For accurately assessing health effects of air pollution, detailed exposure estimates need to be available. Aggregating monitored data collected by different monitoring stations or concentrations measured at central monitoring stations or proximity measures do not seem to reflect the personal exposure. Estimating detailed personal exposure to air pollutants should be addressed more. Indeed, individual exposure studies should include parameters affecting their exposure (cultural, socioeconomic, ethnic, etc.). Although most health outcomes are not confined to a single pollutant, studies typically focus on the risks of single pollutants and do not consider the mixture of pollutants. There is a clear need to develop methods for evaluating and managing the effect of the air pollution with a multi-pollutant approach. However, it should also be remembered that particulate matter (PM) air pollution is already by itself a mixture of solid and liquid elements and not a single pollutant.
Health impact assessment shall consider the simultaneous exposure to multiple pollutants and particularly vulnerable groups of population. Usually, interaction among these different pollutants and the combined effect of these with pollutants that are naturally present in the environment are not included. Furthermore, epidemiology study carry out the potential HIA in the future are based on health outcomes measured in the past which is combined with a given exposure to a given pollutant, whereas more events may occur at any time, namely, due to changing air quality and to the characteristic of the population for the study period.
References
Amann et al., 2013. Compliance with air quality limit values for NO2 in the air quality management zones, TSAP Report #10
Carnevale C., Finzi G., Pisoni E., Volta M., Guariso G., Gianfreda R., Maffeis G., Thunis P., White L., Triacchini G., 2012. An integrated assessment tool to define effective air quality policies at regional scale. Environmental Modelling and Software 38, 306–315.
Keiswetter et al, 2013. Modelling compliance with NO2 and PM10 air quality limit values in the GAINS model, TSAP Report #9.
Zachary D.S. Drouet L., Leopold U. Aleluia Reis L., 2011 Trade-offs between energy cost and health impact in a regional coupled energy–air quality model: the LEAQ model. Environmental Research Letters 6, 1–9.
Potential Impact:
Impacts
The APPRAISAL results impact different areas:
Environmental and health impact. The IA framework and the guidance contribute to support the decision maker in selecting effective policies with assessment of rational methodologies and tools maximizing the impact on AQ and minimizing population exposure.
Economic Impacts. The IA framework includes the assessment of the economic impacts of the AQ policies. Such impacts are related to the optimized internal costs to achieve an AQ objective and at the same time to the minimized external costs due to population exposure.
Policy impacts. The IA framework and the guidelines support the use of scientific knowledge by policy makers and regulatory bodies in Member States. Moreover APPRAISAL contributed to the Air Quality Directive Review.
Social and societal impact. The support of the use of scientific knowledge by policy makers and regulatory bodies in Member States results in a greater harmonization and cohesion between the Member States AQ plans and directives, which will have a valuable social and societal impact.
Scientific impacts. The project contributes to the knowledge base on integrated assessment of the Air quality plans at regional and local scale. The systematic review of the existing IA methodologies and tools, the identification of weaknesses, the IA framework design, the validation of the guidance on specific test cases fix the state of the art of the Integrated Assessment methodologies for air quality planning, and identify the key areas to be addressed by research and innovation.
Main dissemination activities and the exploitation of results
The Consortium members within APPRAISAL are all committed to maximize the outcome and benefits of the project, disseminating in different ways the project results. Each partner is interested in the potential impacts and especially the knowledge transfer opportunities, which result from APPRAISAL. This is particularly true for stakeholder and policy makers, which benefit from APPRAISAL state-of-the-art methodologies to assess AQ and its impact on health and to plan effective AQ policies.
Tools to disseminate APPRAISAL results
The tools used to disseminate the APPRAISAL results are the following:
1. Website creation, administration and maintenance: A website (e.g. http://appraisal-fp7.terraria.com/site/index.php) has been created.
The project website contains the relevant project information. In particular, it details the project structure (work packages); the list of partners and stakeholders; the documentation (both the public one accessible to everybody, and the private one dedicated to the partners and protected by password); the relevant meetings and publications; the list of upcoming events. Also, it has been used to advertise the “1st” and “Final” Annual Conferences of the project, and it is continuously updated to contain all relevant outcomes produced during the project life.
2. APPRAISAL database: the APPRAISAL website hosts a database, to collect information on Integrated Assessment modeling (IAM) approaches for air quality at regional and urban scale. The database has been structured to address all the key issues related to IAM at regional scale. At first, partners and stakeholders of the project have filled in the database. Then, in a second phase, a webpage has been to facilitate the collection of the IAM experiences, and its handling. Through the dedicated webpage new IAM can be filled in (e.g. by regional authorities), but also be reviewed (this is a phase managed by the project partners, that can check the formal correctness of the questionnaires before publishing them on the website), and automatic queries can be done (by all interested users of the webpage) to extract and elaborate information from the database. At this stage, more than 50 questionnaires have been validated and are available in the DB, which has been used to analyze the current IAM approaches, understand their limitations, and propose a new framework for IAM at regional scale.
3. Project poster (informative panels) and leaflet. The Project factsheet, the project communication toolkit and the project leaflet and poster have been prepared in the first months of the project; this material can be found on the website (http://appraisal-fp7.terraria.com/site/index.php). The project brochure has been finalized, printed, and distributed during the APPRAISAL final conference. The Layman’s report (e-book format) is available and downloadable from the project website.
4. Two dedicated conferences have been organized, for targeted audiences (scientists, stakeholders and policy makers).
On November 19-20 2013, a conference was held at the Committee of the Regions (Brussels) together with the LIFE+ project ATMOSYS titled “Tools for the design and implementation of air quality programmes in EU regions”. On the first day the results of the APPRAISAL project were presented and discussed together with the view of the European Commission (DG-RTD, DG-ENV, JRC), European organizations (Committee of the Regions), member states (national representative of Portugal), regions (Emilia-Romagna, Cataluña, Styria) and cities (Brussels, Berlin).
At the end of this day a round table was held to discuss the gaps and challenges for integrated assessment methodologies among representatives of the different policy levels. It became clear that modelling plays an essential role in air quality planning. Recurring topics in the discussion were the need for (i) more reliable data to quantify emissions, comprehending the costs and efficiency of abatement measures including the non-technical measures; (ii) integrated systems that take into account all impacts including health, noise and political and societal acceptance, over the different scales ranging from European to local; (iii) a wide dissemination of knowledge on such systems to the different authorities and the general public; (iv) the exchange of information among the decision levels to improve the synergies.
On May 11 2015, the final project conference was held at the Committee of the Regions (Brussels). It was titled “Integrated Assessment Modelling for the design and implementation of Air quality plans in EU Regions”. The conference was a satellite event of the Green Week 2015.
The results of the APPRAISAL project were presented and discussed together with the view of the European Commission (DG-RTD, DG-ENV, JRC), the European Parliament (J. Girling, Member of the European Parliament and rapporteur of the National Emission Ceiling and Air directive review at Committee on the Environment, Public Health and Food Safety), UNECE-CLRTAP TFIAM, member states, regions and cities.
The conference objective was to provide an overview of the project’s key results, focusing on the:
- description of existing Integrated Assessment methodologies (IAM) at regional and local scales within the EU;
- proposal of a decision framework for the design of Air Quality plans;
- showcase of two applications (on Brussels city and on Porto Region) of the decision framework, through advanced Integrated Assessment Modelling tools.
In addition to this, the conference also provided the opportunity to:
- discuss how IAM is dealt with at the European scale, considering the NEC directive and regional/local plans;
- deliver key information on the SEFIRA FP7 companion project;
- focus on best practices adopted in EU regions to improve air quality.
The final round table focussed on the tools to define AQ policies required/needed by the directive. It was clear from the discussion that important issues are:
- more reliable data to quantify emissions, comprehending the costs and efficiency of abatement measures including the non-technical measures;
- integrating the different geographical scales and the decision levels;
- choosing the right measures at the right decision level (e.g.. low emission zones should be decided and applied at local scale, while EURO standards at EU level);
- implementing platforms for sharing best practices and encouraging dialogue between the EU level and the member states;
- integrated systems that take into account all impacts including health, noise and political and societal acceptance, over the different scales ranging from European to local;
- a wide dissemination of knowledge on such systems to the different authorities and the general public;
- caring for social acceptability of air quality measures.
5. Scientific papers publication on peer-review journals.
A special issue of the scientific journal “Environmental Science and Policy”, titled “Multidisciplinary research findings in support to the EU air quality planning: experiences from the APPRAISAL, SEFIRA and ACCENT-Plus FP7 projects” is under preparation. It will include 6 papers describing the Appraisal project activities. The publication is expected for October 2016. The preliminary titles of the papers are the following:
- “A Decision framework for Integrated Assessment Modeling at regional and local scale”;
- “Overview of current regional and local scale air quality modeling practices: assessment and planning tools in the EU”;
- “Added value of health in an Integrated Assessment Modeling (IAM) of air quality to optimize the design of action plans“;
- “Air Quality Integrated Assessment Modeling in the context of EU policy: a way forward”;
- “Uncertainty evaluation in air quality planning decisions”;
- “Applying integrated assessment methodologies to air quality plans: Two European cases”.
A volume, titled “Air quality integrated assessment: a European perspective (Eds. G. Guariso; M. Volta)” has been submitted for publication to Springer. First revision step already passed and final answer expected by October 2015.
6. Participation in EC concertation meetings, events and exhibitions and other interesting EC initiatives.
A joint APPRAISAL-NIAM (National Integrated Assessment Modeling network) meeting has been organized in Brescia on June 29th to collect from NIAM (project stakeholder) input for the IAM database structure.
A meeting with DG-RTD and AIR (Air Initiative for Regions) representatives have been organized on May 14 2013 to communicate them the project objectives and results and to define a list of joint actions for the Green Week 2013, the REGIO Open Days 2013 and the APPRAISAL 1st conference.
A meeting with DG-RTD and DG-ENV has been organized on May 14 2013 to present the project objectives and results.
The project contributed to the Green Week 2013 with a stand and to the debate “Science & evidence-based environmental policy making – the scientific input to the EU Air Policy Review”.
The Consortium has communicated, during the first project period, the main results of the IAM methodologies review to the DG-ENV to provide feedback on the Air Quality policies review process. This process also been also continued communicating the project outreaches, namely the design of the Integrated Assessment Modelling framework and the guidelines for Integrated Assessment Modelling, to the relevant Commission Services.
7. Participation in relevant conferences: members of the Consortium participated to the events listed in section 4.2 Use and dissemination of foreground (Table A).
8. Collaboration with other EC projects. APPRAISAL had some exchanges with ongoing scientific research projects in support to the EU Air policy. It tried to establish working links with activities funded by the EU in connection to the Air Quality review process: e.g. in the field of social sciences.
More in detail, the following projects have been contacted/linked: SEFIRA, TRANSPHORM, ATMOSYS, ACCENT+, OPERA, MAPLIA (PT), FRESH (EEA), APHEKOM, ECLAIRE, together with the FAIRMODE initiative. Some of them contributed to the APPRAISAL database.
Stakeholder involvement
The project activities have been supported by the consultation with a number of decision makers. Table 2 summarizes the involvement of the stakeholders for the project WP activities.
A number of stakeholders have been directly involved in the following project activities (Table 2):
(1) designing the DB
(2) populating the DB
(3) designing the decision framework
(4) discussing the uncertainty issue
(5) Guidance document: evaluation Tier1
(6) Guidance document: evaluation Tier2 (APPRAISAL test cases)
(7) 1st APPRAISAL conference
(8) 2nd APPRAISAL conference
(9) disseminating the project results
Table 2 - Stakeholders involved.
Ongoing and future dissemination activities
Scientific publications
- A special issue of the journal “Environmental Science and Policy”, published by Elsevier, has been approved and is under preparation (the final publication will be ready in October 2016). The special issue (title: “Multidisciplinary research findings in support to the EU air quality planning: experiences from the APPRAISAL, SEFIRA and ACCENT-Plus FP7 projects”) is a collection of several papers that will present the main results from the APPRAISAL, SEFIRA (Socio Economic Implications for Individual Responses to air pollution policies in EU) and ACCENT-plus (Atmospheric Composition Change – the European Network) FP7 projects.
- The publication of aA book has been submitted to Springer and approved in the SpringerBrief Series. The book is under preparation, and will be published within next Spring. Title of the book is “Air Quality Integrated Assessment: a European perspective “, and it will contain, in comparison to the special issue on “Environmental Science and Policy”, a more detailed analysis of the APPRAISAL project results.
The project results will be presented at the following conferences:
- Zone Atelier Environnementale urbaine in Strasbourg (FR), relating to Air Pollution and Energy, November 2015.
- 21st Session of the Conference of the Parties to the United Nations Framework Convention on Climate Change (COP21/CMP11), Paris (FR), December 2015.
- Air Quality 2016, March 2016, Milan (IT).
Dissemination activities at local/regional/national levels
- The project results will be presented in a number of national meetings, as in the Meeting of the Air Quality Technical Committee coordinated by the Portuguese Agency for the Environment with representatives of the regions of Portugal, November 2015.
- A stakeholder mailing list has been compiled including (1) the local, regional and national decision makers that contributed to the project activities or participated to the project conferences, (2) the SMEs and (3) the research institutes that expressed their interest in the methodologies and tools tested during the project. The project results will be disseminated sending the APPRAISAL Layman’s Report and the Guidance document to such a stakeholder mailing list. The stakeholders will be informed of any further initiatives addressed to the local/regional decision makers.
- RIAT+ model and software has been identified as an example of the decision framework defined in the APPRAISAL project to support local/regional decision makers. JRC is providing the RIAT+ first-guess dataset for European regions. This dataset will be released by end 2015, and will contain data on emission, source-receptor models and emission reduction measures useful to start using RIAT+ with a simplified input configuration. This will allow the decision makers of the European regions to apply in a simple (even if preliminary) way an Integrated Assessment Modelling system to identify effective air quality improvement measures. The APPRAISAL stakeholders will be informed about the availability of the RIAT+ first-guess dataset by email.
Collaboration with other projects/institutes
- APPRAISAL has been cooperating with the SEFIRA project exploring new methodologies to include the social acceptability in the integrated assessment models at local and regional scale.
- The APPRAISAL results will contribute to the FAIRMODE programme and the e-reporting process designed by the EEA to collect data on Air Quality plans in Europe.
- The JRC, Ispra, is developing the SHERPA (Screening for High Emission Reduction Potentials on Air quality) tool, to evaluate in a fast way how “emission changes” influences concentration. SHERPA and RIAT+ will be used in a sequential and integrated way: (1) SHERPA to perform a first screening and (2) RIAT+ in cases in which a further/more detailed analysis (involving costs) is needed.
List of Websites:
The project website (http://appraisal-fp7.terraria.com/site/index.php) includes the project aims, results, deliverables and the EU Air Quality Plans and Projects Database.
Contacts:
Project Coordinator:
Marialuisa Volta (marialuisa.volta@unibs.it)
Department of Mechanical and Industrial Engineering
University of Brescia
Brescia, Italy
Communication Manager:
Philippe Thunis ( philippe.thunis@jrc.ec.europa.eu)
European Commission JRC – IES
Ispra (VA), Italy
Financial Manager:
Elena Murelli (murelli.elena@gmail.com)
University of Brescia
Brescia, Italy
Air quality in Europe is still facing a continued wide-spread of exceedances, particularly regarding PM, NOx and O3. In case of non-compliance the 2008 Air Quality Directive requests Member States (MS) to design local and regional plans and assess their impacts on air quality and human health. MS have therefore developed and applied a wide range of modelling methods to cope with these obligations. Today, with the revision of the EU air quality policy pending, there is a need to consolidate and assess the research results in the field of Air Quality and health Impact Integrated Assessment and make them accessible to policy makers.
In this context, the APPRAISAL project aims at:
1. Undertaking on overall review of the Integrated Assessment methodologies used in different countries at regional and local scale, from the simple (scenario analysis) to the more comprehensive (cost-benefit, cost-effectiveness analysis). This includes evaluating both top-down and bottom-up approaches to systematically analyse their strengths and weaknesses and to identify key areas to be addressed by further research.
2. Designing an Integrated Assessment Modelling framework, based on the information collected during the review process, for different policy-maker requirements, model capabilities and levels of data completeness.
3. Drawing guidelines on how to implement the defined Integrated Assessment System framework, based on identified strengths and weaknesses and best practice examples among the Integrated Assessment Systems in place within MS.
4. Communicating with key stakeholders and in particular to policy-makers the state-of-the-art scientific knowledge on emission abatement assessment.
APPRAISAL includes 15 highly experienced groups working on both air quality and health impacts assessment.
Partners come from all over Europe to guarantee the review process representativeness. A group of stakeholders are closely connected to the Consortium to ensure a direct line of communication with key policy makers. APPRAISAL contributes to improve knowledge on regional and local IA methodologies and supports the revision of EU air quality policies.
Project Context and Objectives:
Exceedances of air quality limit values in urban areas in Europe remain widespread, particularly for PM, NOX and O3. This is not simply a compliance issue but has significant implications for the health and wellbeing of European citizens.
The recent reports on the review of the Thematic Strategy on Air Pollution (Amann et al., 2013; Keiswetter et al., 2013) show the evolution trend of compliance from the base year 2010 to 2025 (assuming current legislation only), the improvement for the optimised A5 so-called ‘Central Policy Scenario’ by 2025 and the further compliance achieved in 2030, by implementing all technical measures (MTFR). The assessment of compliance with the annual mean NO2 limit value and the daily PM10 exceedances limit value are both shown in Figure 1 and Figure 2, respectively. In each case, the limit values used for assessing compliance are those of the current Ambient Air Quality Directive.
Figure 1. NO2 Annual mean compliance assessment (Amann et al. 2013).
Figure 2. PM10 compliance assessment (Amann et al. 2013).
Some important observations can be derived from these figures:
(i) The first observation in comparing the 2010 map with the 2025 CLE case is the clear move away from a general picture of non-compliance (2010) to few limited remaining areas of non-compliance. European wide measures (already mandated) will determine a significant improvement in compliance especially in the EU-15 Member States. What is also clear by comparing the 2025 CLE with the 2025 A5 (defined as ‘central policy scenario’) is the limited potential of further EU-wide measures to improve compliance; this is further underlined by comparing the 2025 A5 scenario with the 2030 MTFR scenario.
(ii) Introducing tougher European-wide measures to address residual non-compliance confined to 10% of the urban zones in Europe (the extent of NO2 non-compliance according to IIASA in the 2025 CLE scenario) would likely be significantly more costly than directly addressing the non-compliance areas with specifically designed measures based on bottom-up Integrated Assessment using regional/local data. This has significant implications for the role of regional/local ‘bottom up’ approaches to develop effective Air Quality Management Plans to efficiently achieving compliance.
(iii) In this regard, regional Integrated Assessment tools such as RIAT (Carnevale et al., 2012), LEAQ (Zachary et al., 2011), etc. with their ability to identify cost-optimised local strategies are already available to quantify the cost-effective split between further European wide measures and regional/local measures. They will inevitably need to find wider application and play an increasing role in these emerging ‘discrete islands of non-compliance’.
(iv) A further observation comes from comparing the 2025 CLE case with the 2025 A5 scenario. A5 is a high ambitious scenario (delivering 75% of the further health benefits of MTFR for the EU as a whole). At this high ambition EU level, a number of Member States are already forced to deploy all available pollution abatement measures (i.e. MTFR). Yet, from an AQ compliance perspective, this does not substantially change the picture from 2025 CLE. This points again to the key role of local targeted technical and non-technical measures in order to achieve compliance. As already noted, such measures (low emission zones, special fuels for captive fleets, captive fleet retrofitting etc.) can only be appropriately designed using ‘bottom-up’ tools.
These observations motivate the growing interest in IA models and tools for local and regional scale. Such interest is also due to the European and national directive requirements. As it is apparent from recitals 1 and especially 2 in the preamble to Air Quality Directive 2008/50/EC (AQD), European air quality legislation puts the main emphasis on protecting human health and the environment as a whole and stresses that “it is particularly important to combat emissions of pollutants at source and to identify and implement the most effective emission reduction measures at local, national and European level.” The basic principles have already been formulated in the former so-called Air Quality Framework directive (96/62/EC) and its daughter directives (1999/30/EC, 2000/69/EC, 2002/3/EC, 2004/1007/EC). The 2008/50/EC Directive encourages the use of air quality modelling, in combination with monitoring, as a scientifically relevant tool for a range of policy applications. Thus, “Air quality plans” according to AQD’s Art. 23 (formerly “Plans and Programmes”) are the strategic elements, to be developed with the aim to reliably meet ambient air quality standards considering not only the effect of emission reduction measure on ambient air quality, but considering aspects of cost-effectiveness as well. The importance of model based approaches for air management becomes apparent again in connection with Article 22 “Postponement of attainment deadlines and exemption from the obligation to apply certain limit values” commonly called “Notification for time extension”. For both, air quality plans and time extension, more elaborated requirements are formulated in Annex XV compared to former regulations. The 12 December 2011 (2011/850/EU) implementing decision already reflects this aspect, looking at the reporting obligations laid down there (Article 13, Annex II, Section H, I, J and especially K) and looking at the amount of information that has to be provided regularly. This important step forward will become real when e-reporting will enter in full operation mode, starting 1 January 2014.
Neither the Directive nor the implementing decision indicates what methodologies are required to devise efficient air pollution abatement measures.
In this context, the APPRAISAL objectives are:
I. to perform an overall review of the methodologies, from simple (e.g. scenario approach) to more comprehensive ones (e.g. full cost-benefit analysis), used in different countries to address and assess the impact of local and regional air quality plans and their health implications.
The proposed methodology to achieve this objective comprises two steps:
a. designing a database (keywords, tables and fields) in which strengths and weaknesses of the different methodologies are classified and organised around 4 main areas:
● Synergies among National, regional and local approaches, including emission abatement policies,
● Air quality assessment, (e.g. modelling, measurement, source apportionment...)
● Health impact assessment approaches,
● Uncertainty and robustness, including Quality Assurance / Quality Control (QAQC).
b. collecting and classifying the available information according to the defined common format (data entry). The database is the main tool to perform the comparison of the existing Integrated Assessment Systems (IASs) on the basis of the defined common format (systematic review) and lead to the identification of existing limitations.
II. To identify the limitations of the currently available assessment methods as emerging from the review process.
III. To design an Integrated Assessment (IA) framework, based on the information collected during the review process, for different levels of data availability, policy-maker requirements and model capabilities and to provide guidelines on how to implement the defined IA framework and to test the application of such guidelines to real cases.
IV. To communicate with key stakeholders, and in particular to policy-makers state of the art scientific knowledge on emission abatement assessment.
V. To identify key areas to be addressed by research and innovation. This has been addressed by the review document, which reports strengths and weaknesses of the various methodologies used in the MS.
To achieve such objectives, the APPRAISAL activities have been structured into 5 Work Packages (WPs).
WP1 (Coordination and Management) guarantees the coordination and management of the project and will ensure that the project provides the expected deliverables according to the foreseen time-plan.
WP2 (Review and gaps identification in Air Quality and health assessment methodologies at regional and local scale) deals with the review of methodologies to assess the effects of emission abatement policy options and measures on the reduction of atmospheric pollutants concentrations and on human health (objective (I)). It takes into account the wide range of developed and applied methods by the EU Member States in the last decade. The review work is organized in four main tasks, covering the main important aspects of the assessment activities, namely (1) synergies among national, regional and local approaches, including emission abatement policies; (2) air quality assessment, including modelling, measuring and source apportionment; (3) health impact assessment approaches; and (4) uncertainty and robustness, including QAQC. WP2 also identifies the limitations of the current assessment methods (objective (II)) and the key areas to be addressed by research and innovation (objective (V)).
The revision phase (WP2) supports the design of an Integrated Assessment framework (project objectives (III)) at regional and local scales (WP3 - Designing IA systems interconnecting national, regional, local models and strategies). The design phase addresses in general what elements need to be considered in an IA framework and in which way these elements need to be connected to design efficient policies. The IA framework design phase also defines the data structure allowing a correct communication between the functional IA framework elements. The WP3 activities lead to identify key areas to be addressed by research.
In WP4 (Guidance on integrated air quality and health assessment systems) results of the review phase (WP2) and design phase (WP3) are condensed into a guidance document that can be used to assess and improve current integrated air quality and health impact assessment modelling practices in the EU (objective (III)).
The final WP4 document provides the stakeholders with answers to questions related to the choice, the setup of an Integrated Assessment Modelling tool and the evaluation of its output, i.e.:
● What is an IA framework? When and what can it be used for? What are its essential components? What are the different types of IAS and their strengths/weaknesses? Which IAS design should be chosen to answer a given problem?
● Which inputs does an IAS require? What are recommended values or procedures to obtain these?
● What can I get from an IAS, how accurate are these results? How can I use these results for designing policies?
To underpin the guidance document, guidelines have been evaluated in practice on real test-cases.
Within WP5 (Dissemination and policy support) dissemination, communication and capacity building activities occurred during the life cycle of the project through tasks such as the establishment of a network community, the provision of publishable deliverables and the setting up of conferences. The communication to key stakeholders of the state of the art scientific knowledge on air quality assessment, as well as the support to the AQ policy revision acted as drivers throughout the entire project structure and duration (objective (IV)).
Project Results:
According to the stated objectives and activities, the project results are:
1. The online database collecting the AQ Plans currently used at regional and local scale in EU MS
An overall review of Integrated Assessment (IA) methodologies currently used in EU Member States to assess regional and local scale air quality and its impacts has been performed.
The review is structured in the online “EU Air Quality Plans and Projects Database”. It focuses on
five topics:
- how emission abatement strategies at the national (Member State) level are considered/integrated in the definition of Regional (zone) and/or Local (agglomeration) Plans and Programmes (RLPPs) to improve the air quality, and vice-versa?
- what are the current modelling techniques used to assess the air pollutant concentrations and the impacts of RLPPs (e.g. air quality models used to produce what-if scenario-analysis or full cost-benefit, cost-effectiveness integrated assessments)?
- what is the usefulness of complementary methodologies which could be used either upstream or within the integrated assessment process, particularly source apportionment?
- are we using the state-of-the-art tools, methods and knowledge to assess the impacts of air pollution on human health?
- what are the available and current-practice methods to assess modelling uncertainty and robustness?
The structured on-line database was fully open to APPRAISAL partners and stakeholders to collect and classify methodologies and systems from member states current practices and from research funded projects. The APPRAISAL consortium contacted the National contact points in European Member States and stakeholders involved in the development of air quality plans, but also the users applying models in the frame of research projects.
Air quality experts and regional/local environmental agencies were invited to fill a questionnaire, detailing the methodologies they use to build their Air Quality Plan (AQP) or research project. More than 60 air questionnaires were finally completed. In particular, almost 60% were AQPs and 30% were research projects (10% other). Almost 70% concern AQ Planning, more than 20% concern AQ assessment and only 1% deal with health assessment. About the number of questionnaire completed in each Country: only Germany exceeds 10, 9 questionnaires are from Portugal and UK, 8 for Belgium and all the other Countries are below this threshold.
TOPIC 1 - Synergies among national, regional and local approaches, including emission abatement technologies
This topic is suitable to identify how emission abatement strategies at the national level are considered and integrated in the definition of regional and local air quality plans and programs.
In this respect, the main stakeholders were interviewed and more than half indicated that these synergies are taken into account.
The analysis on the type of measure included in local AQ plans shows that a greater number of non-technical measures are considered; looking at the National Strategies, the focus appears to be on transport.
Examining the responses to the questions in Topic 1, a common difficulty appears to be how balancing and reconciling emission inventories, especially between EU, national, regional and local scales. In addition, there is a recognition that policy needs properly consider the responsibility hierarchy required in the implementation phase; the policy written without taking into account the risks connected to these factors should become “undeliverable”, both in terms of attainability of targets and in terms of an appropriate executive function or body.
TOPIC 2 - Air quality assessment and planning, including modelling and measurement
This topic investigates the air quality assessment and planning processes, with special attention to how results are obtained at different scales and how these are combined.
Figure 3: IA methodologies used by MS in the scope of air quality plans (left) and by research projects (right).
Currently, air quality integrated assessment and planning are mainly done through scenario analysis, while optimization methods are still more used in research field. Furthermore, at present, many different models are applied for air quality assessment without any standard modelling tools. Moreover, assessing the air quality, the resolution of the emission inventory and other inputs are adapted to the geographical zone under study: this indicates that, in the most of cases, scale represents an issue for using the measurement data in model evaluation.
Another crucial point concerns the challenge posed by modelling air quality at a local scale and especially the integration of these local scale results with results at larger scales.
TOPIC 3 - Health Impact Assessment (HIA)
“HIA is a mean of assessing the health impacts of policies, plans and projects using quantitative, qualitative and participatory techniques” (http://www.who.int/hia/en/). HIA addresses different methodological issues, which must be clarified for proper interpretation of results by policymakers. Different methodologies exist for developing a HIA. The aims and objectives of the assessment, data availability, resources and time-frames have an influence for the HIA methodologies used in particular assessment. With HIA, policymakers will be able to make a more balanced or scientifically weighted decision for the different policy options or scenarios. HIA is usually performed in a relative way by which different options, and interventions can be weighted against each other.
The questionnaire results pointed out that the most common approach used is the predictive approach, with time-series focused on both short-time and long-term exposure. The most frequent air pollutants included in the health impact assessments are the “traditional” pollutants, such as particles (PM10 and PM2.5) nitrogen oxides (NOx) and ozone (O3).
For accurately assessing health effects of air pollution, detailed exposure estimates need to be available. Aggregating monitored data collected by different monitoring stations, or concentrations measured at central monitoring stations do not seem to reflect the personal exposure in many cases. Analyzing epidemiological studies, it is also evident that, to fully assess the health impacts, we must take a multiple pollutant exposure approach and consider also that air pollution exposure has both physical and psychological effects. This latter dimension is less documented and is more difficult to measure. Subjective indicators constitute an appropriate alternative.
TOPIC 4 - Uncertainty and robustness
The topic analyzes model validation and uncertainty estimation describing their limitations. In particular, the key point concerns the use of model for regulatory purposes, and the different uncertainty approaches in Air Quality (AQ) Assessment, Health Impact Assessment (HIA) and Integrated Assessment Modelling (IAM), also with respect to the EU legislation requirements.
The main outcome from the analysis of the questionnaires indicates that model evaluation and uncertainty estimation is more regularly performed in air quality modelling, while it is not often applied in other components, like in HIA applications. The needs that emerged from the questionnaires relate to the quality and quantity of input and validation data, to the improvement of modelling tools and the use of better modelling practices. Many respondents reported the need for the establishment of an evaluation protocol to standardize and harmonize validation and uncertainty estimation methods in EU countries.
TOPIC 5 - Source Apportionment Methodologies
This topic analyzes the studies on source apportionment in Europe and their increasing number in recent years, closely related to the development of improved tools with better functionalities and performance. However, the lack of a European network of aerosol pollutants monitoring sites is becoming a limiting factor for a further growth and consolidation of source apportionment techniques. The partial information on European sources emission factors and fingerprints is limiting sources identification.
Moreover, the definition of European methodological protocols to guarantee a minimum level of quality and to make results from different studies comparable is required.
From the methodological point of view, both Receptor models and Eulerian models appear as the most dynamic areas in the source apportionment field.
The most important source categories to target in order to abate exceedances of air quality limits are traffic, (e.g. nitrogen oxides), agriculture (e.g. ammonia) and biomass burning during the cold season.
According to these results, future studies on source apportionment in Europe should focus on the relevance of gas-to-particle conversion and photochemical processes as the main contributors to recalcitrant pollutants like PM and ozone.
2. The analysis of the limitations of the currently available assessment methods
The systematic review of the AQ plans and projects collected in the APPRAISAL database allowed to identify the limitations of the current methodologies.
● A recurring theme is the challenge that is being posed by local scale modelling, especially with respect to the integration of these local scale results within larger scale ones. This is certainly true also for the integrated assessment tools, which currently are still lacking when considering local scale integrated assessment. Further research should be devoted to this topic.
● When speaking about weaknesses in IAM, the main issue (reported in 70% of the studies) is related to the emission inventories. Even if standard emission inventory (such as EMEP) are available at the European level, these are still lacking when moving to local scale inventories and the consistency between the emissions used at different scales is a major concern.
● For air quality measurements the question was raised on how model results can be combined with observed values in scenario calculations.
● In the overall IAM framework, source apportionment methodologies can bring added values. In fact, in recent years the number of studies on source apportionment in Europe has steadily increased. This is closely related to the continuous development and improvement of tools with respect to functionality and performance. Nevertheless, the lack of an appropriate European network of urban monitoring sites with detailed chemical and physical characterization of aerosols is becoming a limiting factor for a further growth and consolidation of SA techniques. This is true for both receptor models and Eulerian models. From the methodological point of view, a combination of both receptor models and Eulerian models possibly complemented with other techniques (mainly Lagrangian models) appears as the most dynamic area in the further development of source apportionment tools.
● The IAM optimization approach fully responds to the AQ Directives. The emission reduction measures are selected by an optimization algorithm that assesses their impact on air quality, health exposure and implementation costs. Such optimization algorithms require thousands of air quality assessments; in these cases, AQ systems cannot directly be used because of the high computing time demand, so they are used to provide a limited number of simulations that are then processed to identify ‘simple’ emissions-AQ links (source-receptor relationships/models) that are able to capture the specific features of a region.
● There is a need to establish an evaluation protocol in order to standardise and harmonise validation and uncertainty estimation methods in EU countries. Within an IAM framework, model evaluation and uncertainty estimation are more regularly performed in air quality modelling, while they are rarely applied in other IAM components such as, for example, in the health impact assessment. Operational and diagnostic evaluations are the methods preferred both in the case of modelling for the purpose of air quality planning as well as for research projects. For the purpose of air quality plans, expert judgement is also frequently used. Uncertainty propagation methodologies are used, although not so often, to quantify confidence levels of air quality model results.
3. The design of an IA framework interconnecting national, regional, local models
An Integrated Assessment System framework, at the regional and local scale, has been designed starting from the results of the revision phase. It is based on the EEA DPSIR scheme. The building blocks and the connections among them are represented in Figure 4.
Figure 4: the DPSIR scheme adapted to IAM at regional/local scale.
The meaning of each block is as follows (quoting from EEA glossary):
- DRIVERS: this block describes the “action resulting from or influenced by human/natural activity or intervention”. Here we refer to variables (often called “activity levels”) describing traffic, industries, residential heating, etc...
- PRESSURES (Emissions): this block describes the “discharge of pollutants into the atmosphere from stationary sources such as smokestacks, and from surface areas of commercial or industrial facilities and mobile sources, for example, motor vehicles, locomotives and aircrafts.” PRESSURES depend on DRIVERS, and are computed as function of the activity levels and the quantity of pollution emitted per activity.
- STATE (Air Quality): this block describes the “condition of different environmental compartments and systems“. Here we refer to STATE as the concentrations of air pollutants resulting from the PRESSURES defined in the previous block. In IAM implementations, STATE can sometimes be directly measured, but more often it is computed using some kind of air quality model.
- IMPACT: this block describes “any alteration of environmental conditions or creation of a new set of environmental conditions, adverse or beneficial, caused or induced by the action or set of actions under consideration”. In the proposed framework, we refer to IMPACT on human health, vegetation, ecosystem, etc... derived by a modification of the STATE. Again the calculation of the IMPACT may be based on some measure, but normally requires a set of models (e.g. health impacts are often evaluated using dose-response functions).
- RESPONSES: this block describes the “attempts to prevent, compensate, ameliorate or adapt to changes in the state of the environment”. In our framework, this block describes all the measures that could be applied, at a regional/local scale, to improve the STATE and reduce IMPACTS. In particular, 3 different group of responses can be considered:
o Energy/Structural measures, impacting on the DRIVERS block and including both energy efficiency and population behaviour measures.
o End of Pipe measures, impacting on EMISSIONS and related to the introduction of new technologies affecting the level of emission for each activity.
o Technical measures, impacting directly on STATES and including technological improvements allowing for direct concentration reductions (e.g. ecological tarmac/concrete).
Each of these blocks is analysed in terms of input, functionality, output, synergies among different scales and uncertainties related to the block.
Drivers
The basic function of the DRIVERS block is to model the development of key activities (i.e. road traffic, residential combustion, centralized energy production/industry, agriculture) over time. For instance, for traffic, it means to analyze and/or project the fuel consumption due to traffic, the driven km by cars, the current and future fuel, etc. This information is then used as an input to compute emissions (PRESSURES).
More in general, the input parameters for drivers are factors that represent causes of emission-wise essential activities. Important input parameters include population, general economic activities (e.g. in the form of GDP), more specific activity factors (e.g. sector specific production intensities, transport demand, energy demand, etc.) and technology change factors (e.g. vehicle stock structure, energy efficiency of buildings, etc.) that may be driven by international, national or local requirements or goals (RESPONSES block) or “natural”, non-forced development.
Pressure
Air pollutant emissions act as pressures on the environment. Thus, the PRESSURES block corresponds to the computation of the quantity of pollutants emitted into the atmosphere from stationary sources (such as stacks of large industrial facilities, e.g. power plants), surface areas (residential, commercial or industrial facilities), and mobile sources (for example locomotives, aircrafts, ships, cars, etc.).
The emission of a pollutant at a source (emitter) can be measured (as in large point sources) or estimated. In the last case, it is generally calculated as the product of the activity of this emitter (DRIVERS) and an emission factor, which is the quantity of pollutant emitted per unit of activity.
State
In the case of air quality, the STATE describes the ambient concentrations of targeted pollutants. In the case of traffic, it means to consider the air concentrations due to the emissions from cars. The air quality state can be described as gridded concentrations over the studied area, or as local concentrations on receptor sites, depending on the objectives and on the available tools. Also, in addition to the spatial dimension, the air quality state has a temporal dimension, considering that a pollutant can be monitored/ modelled with a temporal resolution of hours/days, etc. Once concentrations are estimated in space and time with different available approaches, one can calculate various air quality indicators, such as aggregation of the initial air quality data e.g. to provide the number of daily exceedances of particulate matter concentration on a cell, the annual mean of nitrogen oxides aggregated over a domain, etc. In this document we will focus on concentrations as a state indicator, even if the content would be basically the same for deposition of atmospheric pollutants.
Impact
IMPACT describes the consequences of any alterations or modifications of environmental conditions related to the STATE of air quality, being either beneficial or adverse. Among the various impacts, we could distinguish between impacts on human health, on environment (vegetation and ecosystems), on social and economic aspects, on climate or on visibility. Moreover some impact could be derived from another, such as economic valuation of human health or of ecosystems. This means, for instance, to evaluate a monetary equivalent of the increased morbidity/mortality related to traffic emissions. In this context, we specifically focus on impacts on human health. In general, the following input is needed to compute IMPACTS:
- air pollution concentrations;
- population data;
- dose/concentration-response functions.
Responses
The RESPONSES block represents the decision framework, that is to say the set of techniques/approaches that can be used to take decisions on actions (mostly, emission reduction measures) to be implemented.
The main components of a decision framework are defined considering: control variables (these represent, for instance, the emission reduction measures that can be applied by the regional/local authority, as car-sharing, congestion charge, etc.); objectives (these represent what a Decision Maker would like to improve); constraints (these can be of different types, as legislative, economic, physical, etc. They can be mathematically formalized, if using a formal approach to take decisions, or they can be taken into account when making decisions, but without explicitly modelling them).
The designed Integrated Assessment framework classifies (in broad terms) two possible decision pathways:
● scenario analysis (Figure 5, a). This is currently the most commonly used approach to design “Plans and Programmes” at regional/local scales. Expert judgment or Source Apportionment are used to identify potential emission reduction measures, the effects of which are then tested by running a regional/local air quality model. This approach has the advantage of simplicity but does not guarantee that cost-effective measures are selected. An evaluation of costs and impacts can only be performed “ex-post”.
● optimization (Figure 5, b). This approach identifies cost-effective measures to improve air quality by solving a mathematical programming problem. During such a process, abatement costs and impacts are continuously compared until the least costly set of measures is found, so that a given air quality target is met (or vice versa, minimizing the Air Quality target given a maximum available budget). This approach guarantees the selection of the set of policies that are cost-effective for a given domain and, in principle, can incorporate all costs and impacts in the optimization procedure.
Figure 5: the DPSIR scheme adapted to IAM at regional/local scale. The red arrow in the Figure represents the “feedback on cost-effectiveness”, provided by the optimization approach.
4. The taxonomy of Air Quality plans
The IAM framework can be used to classify air quality plans according to the complexity of the available data and the methodologies used. An overview of these different levels and their main characteristics for the different blocks is given in Table 1.
Table 1 Different levels of detail for the different DPSIR blocks.
The analysis of individual AQPs can be summarized using a radar chart. For each block of the framework, five levels of complexity have been defined: N/A – information not available, Level 0 – the block is not considered in the AQP, Level 1 – low level of complexity in the implementation, Level 2 – medium level of complexity, and Level 3 – high level of complexity.
The radar chart in Figure 6 represents the “average graph” computed considering all plans available in the database.
Figure 6: A radar graph representing the average complexity level of AQ plans.
5. AQ policies uncertainty assessment
Uncertainty and Sensitivity analyses are key issues in the definition and evaluation of emission control strategies. An AQ plan should include information about the source of uncertainty related to each part of the IA modelling system used to define it. Such assessment is a very complex task due to the fact that an IAM is implemented by a set of integrated models, whose input are usually uncertain and whose output can become the input of the next stages. This fact makes the analytical and partial derivative related methods for both uncertainty and sensitivity analysis quite impossible to be used on the overall system, and sometimes also for a portion of it.
In the next subsections, a brief discussion about the specific uncertainty sources for each block of the IAM framework is presented.
Main uncertainties of DRIVERS block components are:
● Road traffic: Traffic models and/or detailed road segment specific traffic information are relatively commonly available. Technological parameters are relatively well known at least at national level. Parameters required for reliable non-exhaust emission assessment (e.g. road surface type and condition) can be a considerable source of uncertainty.
● Non-road traffic and machinery: For some forms of non-road activities, e.g. sea vessels, trains and airplanes, activities and spatial patterns are often relatively well known. For many other forms of machinery, in contrast, the activity data can be much more uncertain.
● Residential combustion: Residential wood combustion activities and technology information are often uncertain because a lot of the wood fuel is used from private stock directly, and house-hold level heating system stock is poorly known. Furthermore, spatial assessment (i.e. gridding) of residential combustion activities is often uncertain because of the lack of building registers for residential heating appliances.
The uncertainties associated to emissions inventories (PRESSURES) are directly related to accuracy. This accuracy can be split into two main contributions:
- structural inaccuracy, which is due to the structure of the inventory. The structural accuracy estimates the inventory structure ability to calculate as precisely as possible the real emissions. This uncertainty can be split into 3 contributions
o inaccuracy due to aggregations: the emissions are calculated on defined spatial and time scales and for some of them these scales are different from the real emissions ones. This can be due to lack of information on the emission processes or on the variability of the real emissions;
o incompleteness, which means that an emission inventory may be inaccurate due to the absence of emission sources because of a limited understanding of the emission processes;
o inaccurate mathematical formulation and calculation errors: the mathematical formulation used is generally inaccurate (simplified), for example by considering that the relation between emission and activity is supposed linear, which is generally not true.
- inaccuracy on the input data (i.e. activity data, emission factors).
The uncertainties on the input data are mainly due to the lack of information on the different parameters used to estimate the emissions of an inventory.
When the AQ STATE is evaluated through measurements only, uncertainties are related to the measurements themselves, to the geo-statistical methods used to interpolate point measurements and to the representativeness of measurement sites to characterize the area under study. Intrinsic uncertainties of AQ modelling are mainly related to errors in the physical formulation of the model, and to uncertainties in the input data. An operational validation of the AQ model by comparison with measurements is required, opening the question of the representativeness of the chosen measurement sites in relation to the model scale. Evaluating the indefiniteness of prospective study is more challenging and would require the use of diagnostic evaluation (e.g. sensitivity tests) or probabilistic evaluation (e.g. errors propagation). Furthermore, as mentioned earlier, for prospective IAM, estimating the AQ state over a relatively short temporal period (up to one year) introduces uncertainties on the representativeness of the AQ state itself.
Health IMPACT analysis relies on two main processes, namely exposure assessment and epidemiological analysis relating exposure to the health outcome. These two processes include a number of basic steps, finally leading to the quantification of the expected atmospheric pollution induced health burden in the target population, most commonly expressed in terms of years of life lost attributable to the exposure to the atmospheric pollutants under study. Assumptions and uncertainties related to each process may significantly influence the result of the analysis. The main sources of uncertainty in HIA studies can be summarised as follows:
● Uncertainties in estimating the impact for each health outcome. This uncertainty is mainly related to the health-outcome frequencies data. Mortality may be considered generally accurate, but frequency measures of morbidity and data on health-care systems contain uncertainties. Furthermore, in contrast to directly countable events listed in national health statistics (eg, deaths or injuries due to traffic accidents), it is not possible to directly identify the victims of mixtures with cumulative toxicity, such as smoking or air pollutants. Also, the health outcomes may not be specifically linked to air pollution due to synergistic effects with other factors.
● Uncertainties in exposure assessment. Poor exposure assessment is an important source of uncertainty in HIA and can result from errors and biases in either air quality models or in exposure models. The different sources of error and uncertainties in the exposure models result from variability not modelled or incorrectly modelled, inaccurate inputs, errors in coding, simplifications of physical, chemical and biological processes to form the conceptual models, and flaws in the conceptual model. Emission and meteorological input data accuracy and physical/chemistry assumptions and parameterisations in the air quality model largely affect the reliability of its results on the spatial distribution of ambient pollutant concentrations. Furthermore, statistical methods (e.g. kriging) used to produce higher resolved air pollution fields starting from air quality model results and other inputs (local observations, emissions etc) may also introduce uncertainties at specific locations far away from the observations.
● Uncertainties related to the concentration-response functions, estimated by epidemiological models: Some of the formal approaches for uncertainty analysis in epidemiological concentration-response models include Bayesian analysis, Monte Carlo analysis and model intercomparison.
● Uncertainties concerning the temporal scale of effects, i.e. the latency times from exposure to adverse event. This is an uncertainty mainly associated with long-term exposure studies, as acute effects follow exposure by a few days.
It is important that RESPONSE decision approach focuses on robust strategies, that is to say on “policies that do not significantly change due to changes in the uncertain model elements”. This issue is linked to the need of defining a set of indexes and a methodology to measure the sensitivity of the decision problem solutions. It is in fact worth underlining that, while for air quality models the sensitivity can be measured by referring in one way or the other to field data, for IAMs this is not possible, since an absolute “optimal” policy is not known and most of the times does not even exist. The traditional concept of model accuracy must thus be replaced by notions such as risk of a certain decision or regret of choosing one policy instead of another.
6. Guidance document to support the IAM design
The guidance document, focuses on the different topics that need to be addressed to set up an AQ Integrated Assessment Modelling (IAM) system or methodology.
The document has been structured around the building blocks of the EEA DPSIR scheme. It does not provide a single set of recommendations but a range of solutions, so that each of the building blocks can be elaborated to a different level of detail, adapted to the data and tools that are available in practice or what is required to solve the specific air quality problem that is targeted by the IAM (‘fit for purpose’). The different topics of the document are addressed through a number of questions. Some of these questions are specific to the building block while others relate to topics that reappear in all blocks such as how the uncertainty or the interaction between scales should be tackled.
The guidance document has been evaluated on existing AQPs with the aim to identify which guidance was currently lacking and should be further extended .
Eight case studies have been analysed in detail and are listed below:
• AQP for Antwerp
• AQP for Athens
• AQP for the Northern Region of Portugal
• Preliminary AQP for Emilia Romagna
• Research project for the Alsace Region
• AQP for the Warsaw Agglomeration
• Research project for the Warsaw Agglomeration
• AQP for Helsinki
The scheme to explore these ‘practical applications’ has been always the same: the AQP was interpreted in terms of the DPSIR blocks, and a level of detail was assigned to the Drivers, Pressures, State, Impact, Response, according to the way they have been analysed. Then, possible improvements to the DPSIR blocks used were derived together with (possibly) missing or underestimated topics in the guidance document.
In all the considered cases (Figure 7) the drivers and, to a somewhat lesser degree, the pressures and state are well elaborated, while in none of the studies the health impact has been considered with the highest level of detail and also the choice of abatement measures has mostly been based on expert judgment.
Figure 7: Summary of complexity levels for the eight studies considered.
Based on the analysis for the AQP studies the following observations can be made:
● Much effort is put into quantifying the drivers and pressures (emissions) in all the studies that were considered. In all examples, the drivers are treated at the highest level of complexity and only in one of the examples the emissions are based on a top-down and thus ‘low level’ methodology. Regardless of this already high level of detail, the drivers and the emissions are still seen by some studies as the topic where most of the additional guidance would be welcome; and more in particular the guidance document should be extended with practical references to additional data sources with an emphasis on: (i) ‘real life’ emission factors for traffic, domestic heating (wood/coal burning appliances) and agriculture; (ii) projections and future estimates and (iii) practical examples on combining emission inventories at different scales.
● The state (concentrations) is in all the examples handled with one or more models. For some of the AQPs that lack detailed local scale concentration assessments, it is also acknowledged that this is a point for improvement. The document needs to be extended with guidance on using data assimilation and a better reference should be made to FAIRMODE initiative for the evaluation of model results.
● Uncertainty, both in the emissions and concentrations, is not addressed except in a single research study, which was included in this analysis. Based on the feedback given from the 8 cases, there is however need for additional guidance on how to perform such analysis.
● Even if health impact is considered in most of the examples, this is never done based on a detailed temporal and spatial resolution for the exposure and population data (in three of the plans health impact is not addressed at all). Also for this topic more practical guidance would be welcome with references to real case applications and to where input data (dose-response, ..) can be found. As this is a subject unfamiliar to most of those involved in developing air quality plans, some of the topics addressed in the guidance document should be better explained.
● Only the two studies using the RIAT+ modelling system rely on a multi-objective optimisation method to identify the optimal mix of abatement measures. It can however be noted that for both applications, even if they imply the highest level of complexity for the response block, some of the identified improvements relate to the database, e.g. the costs and emission reduction efficiencies for the abatement measures or the weights attributed to single objectives when doing a multi objective optimisation. It should be clear that – as in any modelling activity – the results of the optimisation process heavily rely on the quality of the input data that are used in this process.
7. The design and the assessment of AQ plans for Porto and Brussels regions
The final step of the APPRAISAL project was the application of the IAM tool RIAT+ in two test cases: one in the Brussels Capital Region in Belgium and in the region of Porto in the North of Portugal. The experience obtained through these two test cases was used to further improve the guidance document.
The RIAT+ system, developed during the OPERA project (www.operatool.eu LIFE09 ENV/IT/092, Best LIFE Environment projects Award 2014), is an IAM tool designed to help regional decision makers to select optimal air pollution reduction policies that will improve the air quality at minimum costs. To achieve this, the system incorporates explicitly the specific features of the area of interest with regional input data-sets for the:
● precursor emissions of local and surrounding sources;
● abatement measures (technical and non-technical) described per activity sector and technology with information on application rates, emission removal efficiency factor and cost;
● the effect of meteorology and prevailing chemical regimes through the use of site specific source-receptor functions.
The system runs as a stand-alone desktop application and can be downloaded (free license) from the project web-site (http://www.operatool.eu/download/). RIAT+ has been already applied in Emilia-Romagna Region (IT) and in Alsace (FR) during the OPERA project and in Lombardy Region to evaluate AQP measures.
The tool allows the two possible decision pathways already mentioned in the DPSIR scheme: scenario analysis and optimisation. The main outputs from RIAT+ are a summary of emission reductions on the domain, a table of the application rates for the different measures, maps of a set of relevant air quality indexes (AQIs) and, for the optimization pathway, the Pareto Curve providing the efficient solutions of a specific AQI ranked by costs.
A source receptor (S/R) model is used, internally, to link emissions (pressure) to an air quality indicator, that represent the state. A S/R model can be as simple as a linear relationship, or as complex as a chemical transport model (CTM). To limit the computational time, RIAT+ normally uses a nonlinear relations identified by means of Artificial Neural Networks (ANNs), tuned to replicate the results of a limited set of simulations performed by the users with deterministic air quality model calibrated for the specific site.
The selection of which simulations to use is an important aspect in setting up the ANN. They must be representative of the range of emissions/concentrations that can be encountered during the optimization procedure. The definition of this training data set is typically referred to as the ‘Design of Experiment’: this establishes the configurations for the CTM simulations.
Both the test cases are presented below with the same scheme: a quick introduction on the area and the proposed abatement measures, followed by the application of RIAT+, with a focus on the technology database used, the chemical transport modelling runs, the Design of Experiment and the identification of the source-receptor models and, in the end, the results obtained.
Bruxelles case study
The Brussels Capital Region (BCR) has an area of 161 km2 and is home to more than 1.1 million people. The region consists of 19 municipalities, one of which is the Brussels Municipality, the capital of Belgium (Figure 8).
Figure 8: Location of the BCR (red zone) in Belgium.
The proposed abatement measures were provided by BIM (Brussels Environment, http://www.ibgebim.be). They are a list of 13 measures consisting of 9 traffic measures and 4 domestic heating measures, that have been approved by Brussels authorities. For these abatement measures, BIM provided order-of-magnitude estimations of the costs and emission reductions.
The RIAT+ database with abatement technologies that are available for the macro-sectors of interest – non-industrial combustion (2) and transport (7) – was derived from GAINS Europe (http://gains.iiasa.ac.at/gains/EUN/index.login?logout=1) and the OPERA project database (www.operatool.eu).
For air quality modelling of the BCR, the AURORA chemical transport model (see: http://pandora.meng.auth.gr/mds/showlong.php?id=167) was used with a domain of 49 x 49 grid cells at 1 km resolution (Figure 9) with base emissions for the year 2009 and a hourly time step. For the vertical discretization, 20 layers were used for a domain extending up to 5 km.
Figure 9: Model grid used for the CTM calculations.
The results model setup were validated by comparison to the observed values at the measurement stations inside the model domain. For the model validation, the FAIRMODE methodology (http://fairmode.jrc.ec.europa.eu/) was adopted.
For the Design of Experiment phase, three levels of emission reduction were distinguished: base case (B), high emission reductions (H) and low emission reductions (L). In order to determine the emission reduction scenarios for the ANN training, the three levels B, H, L were combined to produce 14 emission scenarios. These scenarios were applied to the emissions both in and outside the policy application domain (PAD), which, in this case, is the BCR.
In this study, the AQIs considered are:
● PM10: yearly average of PM10 concentrations
● NO2: yearly average of NO2 concentrations.
RIAT+ was then run with this S/R model to look for optimal policies beyond the 2020 Current LEgislation. As the emission changes, that can be obtained with the selected set of measures, are limited, unsurprisingly, the concentration changes are also limited. Figure 10 shows an example of NO2 concentration changes due to the emission abatement measures.
Figure 10: Yearly average NO2 concentration changes (µg/m3) for all traffic and all non-industrial heating measures as well as for the combination of these two in 2020 compared to the reference (CLE 2020). The number in parentheses is the maximum concentration change.
Porto case study
The Great Porto Area is a Portuguese NUTS3 subregion involving 11 municipalities. It covers a total area of 1024 km2 with a total population of more than 1.2 million inhabitants (Figure 11).
Figure 11: Location of the Great Porto Area in Portugal and in the Northern Region of Portugal.
The GAINS database of reduction technologies, which contains a large data set collected for Portugal by IIASA (http://www.iiasa.ac.at) was used. It contains values for the years 2010, 2015, 2020 and 2025, including costs and emissions effects. The reference scenario «TSAP» of March 2013 was considered in the study.
The TAPM model (http://pandora.meng.auth.gr/mds/showlong.php?id=120) was applied to the Great Porto Area (150x150 km) for one entire reference year (2012) with a 2x2 km2 spatial resolution using disaggregated emissions from the Portuguese 2009 emission inventory, which is the most recent available. Extending previous evaluations of the same model for the Great Porto Area, the results of the TAPM simulations were compared to the measured values at the monitoring stations inside the model domain. As in the Brussels case, the FAIRMODE methodology was used for the validation (http://fairmode.jrc.ec.europa.eu/).
The Design of the Experiment was influenced by the computational time constraints, so just 10 emission reduction scenarios were simulated. Processing the TAPM simulation results, low and high concentration levels were obtained. Each TAPM simulation is a full year with a hourly time step. The target considered in this application was the PM10 annual mean.
RIAT+ was applied in the optimization mode and Figure 12 shows the efficient solutions obtained for the Great Porto domain. On the horizontal axis of the figure, there are industrial costs (i.e. those related to the implementation of end-of-pipe measures), considered over CLE (i.e. those mandatory under the current legislation) and expressed in M€. On the vertical axis, there is the spatial average of PM10 annual mean (the AQI value for this particular case) estimated over the entire study area.
Figure 12: Pareto curve for the optimization of PM10 yearly mean concentration.
The Pareto Curve shows that a PM10 mean concentration of 28.8 μg/m3 can be reached by adopting emission reduction technologies costing around 7.6 Million € per year, but shows at the same time, that a mean concentration below 28 μg/m3 cannot be attained even investing a very large sum of money in local reductions. To further improve air quality, a much larger application domain must be considered for reduction measures.
For the point of the Pareto curve mentioned above, Figure 13 shows the spatial distribution of average PM10 annual concentration values. The largest reduction of PM10 emissions and concentration levels are expected over the Porto municipality where the population has the highest density.
Figure 13: Mean PM10 concentration resulting from RIAT+ application (point C of the Pareto curve).
Main comments to the test cases
The main aim of these applications was to confront the practical application of RIAT+ with the guidance document.
The following main conclusions can be drawn:
- In practice, the list of options for abatement measures is restricted not only by what is technically and economically feasible, but possibly even more by political and social acceptance. IAM tools should therefore be extended to allow their users to take into account the implications of political and social acceptance at an early stage of the decision process.
- The applications demonstrate that tools exist which can be practically applied in an integrated assessment of air quality that does not only consider compliance of concentration to limit values but also efficiently take into account internal and external costs (e.g. health impact) of different available abatement options.
- The biggest task when implementing such a comprehensive IAM is - as is also the case in regular air quality modelling applications – to obtain high quality input data i.e.. information on local emissions and the cost and effectiveness of possible abatement measures. When such data are lacking, one can still rely on existing European inventories and databases with data on abatement measures such as EMEP and GAINS well keeping in mind the assumed validity of such data for the region of interest and the implications for the results obtained using the IAM.
- If an IAM system uses source-receptor relationships (artificial neural networks, linear regression, ...) to relate emission changes to concentration changes, such relationships should be carefully tested to ensure that they not only correctly replicate the concentration values obtained through more complex modelling tools (e.g. CTMs) but also capture the dynamics i.e. the concentration changes calculated by the model for which they are a surrogate.
In the Brussels case, a lot of time was put into estimating precise measures while the impact on air quality of these measures is rather limited due to the dimension of the area selected. A first screening step such as a simple scenario to check the importance of the impacts should be done before using a complex methodology as the latter has limited added value in such cases.
In the Porto case a list of available technologies from an existing database was used and the main sectors were selected and identified. Nevertheless, a more detailed list of locally available measures needs to be decided and discussed with stakeholders and policy makers. With the optimization approach, it was possible to have a first idea of the optimal investment costs and benefits to achieve a given PM10 air quality objective, and to identify the complete range of their feasible values.
8. The identification of the key research areas
The key areas to be addressed by research have been identified and organized following the DPSIR scheme building blocks.
Drivers (Activities)
Considerable deficiencies were identified for the DRIVERS block, for all activity sectors contributing to local scale emissions, not only for power plants, road traffic and residential combustion, but also for agriculture, non-road traffic and machinery.
An important future research line should be devoted to the integration of activity inventories at different spatial scales. A further key issue for future research is related to activity evolution. On the one side, one would certainly like to improve the estimation of how local economic sectors will develop and adapt in the future, taking into account both internal factors, such as economic downturns, and external ones, as climate changes. This means considering new land use policies (activity location) as part of the IAM problem. On the other side, since a perfect prediction of activity evolution is out of question, new methods to deal with uncertain predictions (ensemble modelling, risk aversion,...) have to be developed and possibly become standard.
Pressures (Emissions)
In the IAM database collected by APPRAISAL, 70% of the respondents identified emission values as the main weakness of their modelling approach. Quantifying the effectiveness of specific abatement measures within a zone presumes that the emission inventory is disaggregated with sufficient details both spatially and per categories to properly consider the emission abatement measures. This level of detail is unfortunately lacking in inventories, leading to unreliable estimates of the effect of measures; i.e. the official national and European (EMEP) emission inventories only contain detailed (per category) emission totals for the member state as a whole, or alternatively, gridded data with only SNAP 1 detail. So this issue can cause, as already said, inconsistencies when integrating data coming from different scales.
More in detail, key areas to be investigated by research and innovation concern various topics:
- emission inventory harmonization;
- emission scenario projection, emission factors estimation.
State (Concentration levels)
Key areas to be addressed by research and innovation, in the STATE module, are the following:
- integration of ground-based and remote-sensing monitoring methods, to precisely assess the “current” AQ situation;
- better understanding of composition of the various PM fractions;
- regional and local scale CTM implementation and validation;
- source-receptor model identification approaches.
Impacts (Human health)
In such field the key areas to be addressed by research and innovation are:
- health impact assessment integration in IAM;
- exposure-response functions;
- acute effects and short-term impact on mortality;
- mortality and morbidity factors due to long term NO2 and O3 exposure.
Responses (Methodologies to design measures)
Most of the APPRAISAL recommendations on key areas for future research and innovation are related to the decision process (or RESPONSES block within the DPSIR cycle), mainly considering:
- Development of more detailed information on abatement measures.
- Inclusion of socio-economic aspects in the analysis.
- Better account for “ Energy Efficiency / non-technical measures”. These measures are now limited to scenario analysis because of the difficulty to estimate their associated costs. Still progress is required to include these measures in optimized IAMs.
- Multi-scale interactions: as already done for Chemical Transport Models, there is the need to develop IAMs nesting capabilities (one-way/two-ways) to include EU/national constraints in regional analyses, and at the same time providing feedbacks from the regional to the EU/national scale.
- Provide guidance for developing and using IAM approaches and favour harmonization.
- Better integrate Air Quality and Climate Change policies in the IAM framework. In a “resource limited” world, efficiency (to get win-win solutions from Air Quality and Climate Change actions) is of extreme importance and requires guidelines to better integrate climate change policies (normally established at national or even international levels) into air quality plans (developed at regional/local level).
- Develop dynamically evolving impact assessments. Current approaches aim at solutions to be reached at a given time horizon and are therefore “static”. But the system is non-stationary (see the effect of the current economic crisis) and it becomes important for decision makers to know where to invest in priority with potential to adapt decisions with time.
9. The contribution to the Air Quality review
One of the main results of the project is the contribution to the AQ review process.
In general the Air Quality Directive 2008/50 and more recently the Commission Implementing Decision 2011/850 do not specify what methodology is required to devise efficient measures to improve the quality of the air. The contents of the template provided for reporting however indicate that a scenario approach supported by source apportionment can be useful addressing the following:
1. Source apportionment: Which are the main emission sources responsible for the pollution, distinguishing local and regional (transboundary) contributions? With which accuracy is the emission source base case known?
2. Air quality assessment for the current situation: In which zones (location, type) are exceedances of the limit values of a pollutant observed and how large is the population that is exposed?
3. Air quality assessment for future years or emission scenarios:
● What is the baseline level i.e. the concentration to be expected in the year when the limit value comes into force without any measures beyond those already agreed or implied by existing legislation.
● Which measures are currently in place beyond those required by current legislation and what is their effect on the air quality?
● Which additional measures are planned and what is their effect on the air quality?
With respect to emissions, the Directive 2008/50/EC requires an air quality plan reporting the origin of pollution (Annex XV) by providing a list of the main emission sources responsible for pollution (map) and the total quantity of emissions from these sources (tonnes/year). The Commission Implementing Decision of 12 December 2011 requires the AQP to report on the emission scenario and the total emission for both the baseline and for the projection as well as the reduction in annual emissions due to the applied measures.
In fact, the Directive acknowledges the importance “to identify and implement the most effective emission reduction measures at local, national and Community level” (article 2). This presumes that:
1. the emission inventory used for the AQP is sufficiently detailed to allow mapping measures to the specific emissions managed at the different administrative levels that have to be considered.
2. the costs of emission reduction technologies are available.
3. a suitable optimization approach to select effective policies can be implemented.
Emission inventories and projections as needed for the assessment and planning at the local scale are currently developed ad-hoc. It is recommendable to take an initiative to harmonize the criteria and the procedures for developing such local emission inventories. Further fixing and specifying these procedures might improve emission data necessarily needed for air quality modelling and in consequence will improve modelling results for this part.
Moreover, the effectiveness of any type of remediation measure strongly depends on the reliability of the pollution source identification and quantification process. Hence, the use of methodologies with minimum biases and uncertainties certainly contributes to focusing valuable resources and time on the most contributing sources in the area of interest.
Article 25 of the Directive deals with the problem of transboundary air pollution. To be effective an air quality plan should appropriately take into account the contribution of sources outside the zone considered in the plan. This is especially true for long lived and secondary pollutants and where the zone that is modelled is small as in local and street level models. In those cases larger scale modelling is needed to properly incorporate the effect of the boundary conditions or at least a sensitivity analysis should be required to quantify the importance of the boundary conditions. If results at different scales are combined, the consistency of the inputs used should be checked and care should be taken to account for differences between the models.
On the other hand the problem of transboundary air pollution can be read as the issue to assess the impact of regional-local emissions, in other words, to quantify the effective potential of regional-local policies in a specific domain. Methodologies should be formalized and developed to fill this gap.
Integrated Assessment Modelling (IAM) should support the air quality authorities in selecting efficient mitigation strategies by providing tools for assessing and solving air quality planning problems at different spatial scales. AQ modelling is the IAM component explicitly mentioned in EU legislation. The Directive recognizes that modelling can be used in combination with measurements to obtain a better representation of the spatial distribution: “Where possible modelling techniques should be applied to enable point data to be interpreted in terms of geographical distribution of concentration” (Article 6). As population density is not necessarily homogeneous within a zone, the air quality plan report could be improved by replacing the single values for the concentration and population within the zone in the report by a map showing the spatial distributions for the concentration and the population.
As the role of modelling in understanding the influence of physical and chemical processes on the dispersion and transformation of pollutants is increasingly being recognised, and MS are already using models in their current assessment techniques it is recommended to further promote the use of modeling tools in the scope of the nowadays AQD revision. Moreover, there is no alternative to modelling for assessing the effectiveness of emission reduction measures in future years. Thus, modelling should become an essential part of air quality planning and any such modelling based report should include a complete description of the model and inputs used as well as an evaluation to quantify the reliability of the AQ assessment.
Today many different modelling tools exist that are being used for AQ assessment and planning so that there is currently no obvious standard model that could be imposed as a ‘preferred’ model for each of the different scales and pollutants considered. Preferably however the model as a whole or at least its subcomponents should have undergone a scientific peer review or a report should exist in which the model has been submitted to a diagnostic analysis.
The need to incorporate uncertainty estimation in air quality modelling is also recognised by policy makers and is required by the AQD, which specifies modelling quality objectives. As the directive does not provide guidelines on how to carry out model evaluation to achieve the quality requirements imposed, the development of relevant guidelines is necessary for modellers and authorities. Several attempts have been made for the establishment of uncertainty assessment guidelines within a number of projects, including AIR4EU and FAIRMODE. The Guidance Document that was elaborated within FAIRMODE is the current reference point for model users and regulators to ensure that their air quality model meets the quality criteria required by EU legislation.
Another important issue for a proper model application when developing an AQP, which is currently not covered by the AQD, concerns the representativeness of the simulation period for air quality planning. Currently an AQP usually relies on the model results for a single meteorological year. This year is often the year for which the exceedances were observed that triggered the demand for an AQP in the first place but which is not necessarily a year that is representative for the time horizon of the AQP. An alternative could be to select a single year (or years) based on criteria (e.g. typical meteorological year, critical meteorological conditions year) that assure it represents the air quality for the time horizon of the AQP. Another option could be to use a longer multiyear meteorological period for the AQP modelling work. The latter could however well be impractical considering the increase in computer time required to model such long periods. In conclusion, it is necessary to account for meteorological variability in air quality modelling and IAM since meteorology is a constraint that influences the effectiveness of emission reduction measures to some extent, though meteorology is not the primary cause for air pollution. This is generally addressed for in EU scale IAM studies (e.g. IIASA GAINS approach) but seems lacking at the regional/local scale, most probably due to the limited resources available.
For accurately assessing health effects of air pollution, detailed exposure estimates need to be available. Aggregating monitored data collected by different monitoring stations or concentrations measured at central monitoring stations or proximity measures do not seem to reflect the personal exposure. Estimating detailed personal exposure to air pollutants should be addressed more. Indeed, individual exposure studies should include parameters affecting their exposure (cultural, socioeconomic, ethnic, etc.). Although most health outcomes are not confined to a single pollutant, studies typically focus on the risks of single pollutants and do not consider the mixture of pollutants. There is a clear need to develop methods for evaluating and managing the effect of the air pollution with a multi-pollutant approach. However, it should also be remembered that particulate matter (PM) air pollution is already by itself a mixture of solid and liquid elements and not a single pollutant.
Health impact assessment shall consider the simultaneous exposure to multiple pollutants and particularly vulnerable groups of population. Usually, interaction among these different pollutants and the combined effect of these with pollutants that are naturally present in the environment are not included. Furthermore, epidemiology study carry out the potential HIA in the future are based on health outcomes measured in the past which is combined with a given exposure to a given pollutant, whereas more events may occur at any time, namely, due to changing air quality and to the characteristic of the population for the study period.
References
Amann et al., 2013. Compliance with air quality limit values for NO2 in the air quality management zones, TSAP Report #10
Carnevale C., Finzi G., Pisoni E., Volta M., Guariso G., Gianfreda R., Maffeis G., Thunis P., White L., Triacchini G., 2012. An integrated assessment tool to define effective air quality policies at regional scale. Environmental Modelling and Software 38, 306–315.
Keiswetter et al, 2013. Modelling compliance with NO2 and PM10 air quality limit values in the GAINS model, TSAP Report #9.
Zachary D.S. Drouet L., Leopold U. Aleluia Reis L., 2011 Trade-offs between energy cost and health impact in a regional coupled energy–air quality model: the LEAQ model. Environmental Research Letters 6, 1–9.
Potential Impact:
Impacts
The APPRAISAL results impact different areas:
Environmental and health impact. The IA framework and the guidance contribute to support the decision maker in selecting effective policies with assessment of rational methodologies and tools maximizing the impact on AQ and minimizing population exposure.
Economic Impacts. The IA framework includes the assessment of the economic impacts of the AQ policies. Such impacts are related to the optimized internal costs to achieve an AQ objective and at the same time to the minimized external costs due to population exposure.
Policy impacts. The IA framework and the guidelines support the use of scientific knowledge by policy makers and regulatory bodies in Member States. Moreover APPRAISAL contributed to the Air Quality Directive Review.
Social and societal impact. The support of the use of scientific knowledge by policy makers and regulatory bodies in Member States results in a greater harmonization and cohesion between the Member States AQ plans and directives, which will have a valuable social and societal impact.
Scientific impacts. The project contributes to the knowledge base on integrated assessment of the Air quality plans at regional and local scale. The systematic review of the existing IA methodologies and tools, the identification of weaknesses, the IA framework design, the validation of the guidance on specific test cases fix the state of the art of the Integrated Assessment methodologies for air quality planning, and identify the key areas to be addressed by research and innovation.
Main dissemination activities and the exploitation of results
The Consortium members within APPRAISAL are all committed to maximize the outcome and benefits of the project, disseminating in different ways the project results. Each partner is interested in the potential impacts and especially the knowledge transfer opportunities, which result from APPRAISAL. This is particularly true for stakeholder and policy makers, which benefit from APPRAISAL state-of-the-art methodologies to assess AQ and its impact on health and to plan effective AQ policies.
Tools to disseminate APPRAISAL results
The tools used to disseminate the APPRAISAL results are the following:
1. Website creation, administration and maintenance: A website (e.g. http://appraisal-fp7.terraria.com/site/index.php) has been created.
The project website contains the relevant project information. In particular, it details the project structure (work packages); the list of partners and stakeholders; the documentation (both the public one accessible to everybody, and the private one dedicated to the partners and protected by password); the relevant meetings and publications; the list of upcoming events. Also, it has been used to advertise the “1st” and “Final” Annual Conferences of the project, and it is continuously updated to contain all relevant outcomes produced during the project life.
2. APPRAISAL database: the APPRAISAL website hosts a database, to collect information on Integrated Assessment modeling (IAM) approaches for air quality at regional and urban scale. The database has been structured to address all the key issues related to IAM at regional scale. At first, partners and stakeholders of the project have filled in the database. Then, in a second phase, a webpage has been to facilitate the collection of the IAM experiences, and its handling. Through the dedicated webpage new IAM can be filled in (e.g. by regional authorities), but also be reviewed (this is a phase managed by the project partners, that can check the formal correctness of the questionnaires before publishing them on the website), and automatic queries can be done (by all interested users of the webpage) to extract and elaborate information from the database. At this stage, more than 50 questionnaires have been validated and are available in the DB, which has been used to analyze the current IAM approaches, understand their limitations, and propose a new framework for IAM at regional scale.
3. Project poster (informative panels) and leaflet. The Project factsheet, the project communication toolkit and the project leaflet and poster have been prepared in the first months of the project; this material can be found on the website (http://appraisal-fp7.terraria.com/site/index.php). The project brochure has been finalized, printed, and distributed during the APPRAISAL final conference. The Layman’s report (e-book format) is available and downloadable from the project website.
4. Two dedicated conferences have been organized, for targeted audiences (scientists, stakeholders and policy makers).
On November 19-20 2013, a conference was held at the Committee of the Regions (Brussels) together with the LIFE+ project ATMOSYS titled “Tools for the design and implementation of air quality programmes in EU regions”. On the first day the results of the APPRAISAL project were presented and discussed together with the view of the European Commission (DG-RTD, DG-ENV, JRC), European organizations (Committee of the Regions), member states (national representative of Portugal), regions (Emilia-Romagna, Cataluña, Styria) and cities (Brussels, Berlin).
At the end of this day a round table was held to discuss the gaps and challenges for integrated assessment methodologies among representatives of the different policy levels. It became clear that modelling plays an essential role in air quality planning. Recurring topics in the discussion were the need for (i) more reliable data to quantify emissions, comprehending the costs and efficiency of abatement measures including the non-technical measures; (ii) integrated systems that take into account all impacts including health, noise and political and societal acceptance, over the different scales ranging from European to local; (iii) a wide dissemination of knowledge on such systems to the different authorities and the general public; (iv) the exchange of information among the decision levels to improve the synergies.
On May 11 2015, the final project conference was held at the Committee of the Regions (Brussels). It was titled “Integrated Assessment Modelling for the design and implementation of Air quality plans in EU Regions”. The conference was a satellite event of the Green Week 2015.
The results of the APPRAISAL project were presented and discussed together with the view of the European Commission (DG-RTD, DG-ENV, JRC), the European Parliament (J. Girling, Member of the European Parliament and rapporteur of the National Emission Ceiling and Air directive review at Committee on the Environment, Public Health and Food Safety), UNECE-CLRTAP TFIAM, member states, regions and cities.
The conference objective was to provide an overview of the project’s key results, focusing on the:
- description of existing Integrated Assessment methodologies (IAM) at regional and local scales within the EU;
- proposal of a decision framework for the design of Air Quality plans;
- showcase of two applications (on Brussels city and on Porto Region) of the decision framework, through advanced Integrated Assessment Modelling tools.
In addition to this, the conference also provided the opportunity to:
- discuss how IAM is dealt with at the European scale, considering the NEC directive and regional/local plans;
- deliver key information on the SEFIRA FP7 companion project;
- focus on best practices adopted in EU regions to improve air quality.
The final round table focussed on the tools to define AQ policies required/needed by the directive. It was clear from the discussion that important issues are:
- more reliable data to quantify emissions, comprehending the costs and efficiency of abatement measures including the non-technical measures;
- integrating the different geographical scales and the decision levels;
- choosing the right measures at the right decision level (e.g.. low emission zones should be decided and applied at local scale, while EURO standards at EU level);
- implementing platforms for sharing best practices and encouraging dialogue between the EU level and the member states;
- integrated systems that take into account all impacts including health, noise and political and societal acceptance, over the different scales ranging from European to local;
- a wide dissemination of knowledge on such systems to the different authorities and the general public;
- caring for social acceptability of air quality measures.
5. Scientific papers publication on peer-review journals.
A special issue of the scientific journal “Environmental Science and Policy”, titled “Multidisciplinary research findings in support to the EU air quality planning: experiences from the APPRAISAL, SEFIRA and ACCENT-Plus FP7 projects” is under preparation. It will include 6 papers describing the Appraisal project activities. The publication is expected for October 2016. The preliminary titles of the papers are the following:
- “A Decision framework for Integrated Assessment Modeling at regional and local scale”;
- “Overview of current regional and local scale air quality modeling practices: assessment and planning tools in the EU”;
- “Added value of health in an Integrated Assessment Modeling (IAM) of air quality to optimize the design of action plans“;
- “Air Quality Integrated Assessment Modeling in the context of EU policy: a way forward”;
- “Uncertainty evaluation in air quality planning decisions”;
- “Applying integrated assessment methodologies to air quality plans: Two European cases”.
A volume, titled “Air quality integrated assessment: a European perspective (Eds. G. Guariso; M. Volta)” has been submitted for publication to Springer. First revision step already passed and final answer expected by October 2015.
6. Participation in EC concertation meetings, events and exhibitions and other interesting EC initiatives.
A joint APPRAISAL-NIAM (National Integrated Assessment Modeling network) meeting has been organized in Brescia on June 29th to collect from NIAM (project stakeholder) input for the IAM database structure.
A meeting with DG-RTD and AIR (Air Initiative for Regions) representatives have been organized on May 14 2013 to communicate them the project objectives and results and to define a list of joint actions for the Green Week 2013, the REGIO Open Days 2013 and the APPRAISAL 1st conference.
A meeting with DG-RTD and DG-ENV has been organized on May 14 2013 to present the project objectives and results.
The project contributed to the Green Week 2013 with a stand and to the debate “Science & evidence-based environmental policy making – the scientific input to the EU Air Policy Review”.
The Consortium has communicated, during the first project period, the main results of the IAM methodologies review to the DG-ENV to provide feedback on the Air Quality policies review process. This process also been also continued communicating the project outreaches, namely the design of the Integrated Assessment Modelling framework and the guidelines for Integrated Assessment Modelling, to the relevant Commission Services.
7. Participation in relevant conferences: members of the Consortium participated to the events listed in section 4.2 Use and dissemination of foreground (Table A).
8. Collaboration with other EC projects. APPRAISAL had some exchanges with ongoing scientific research projects in support to the EU Air policy. It tried to establish working links with activities funded by the EU in connection to the Air Quality review process: e.g. in the field of social sciences.
More in detail, the following projects have been contacted/linked: SEFIRA, TRANSPHORM, ATMOSYS, ACCENT+, OPERA, MAPLIA (PT), FRESH (EEA), APHEKOM, ECLAIRE, together with the FAIRMODE initiative. Some of them contributed to the APPRAISAL database.
Stakeholder involvement
The project activities have been supported by the consultation with a number of decision makers. Table 2 summarizes the involvement of the stakeholders for the project WP activities.
A number of stakeholders have been directly involved in the following project activities (Table 2):
(1) designing the DB
(2) populating the DB
(3) designing the decision framework
(4) discussing the uncertainty issue
(5) Guidance document: evaluation Tier1
(6) Guidance document: evaluation Tier2 (APPRAISAL test cases)
(7) 1st APPRAISAL conference
(8) 2nd APPRAISAL conference
(9) disseminating the project results
Table 2 - Stakeholders involved.
Ongoing and future dissemination activities
Scientific publications
- A special issue of the journal “Environmental Science and Policy”, published by Elsevier, has been approved and is under preparation (the final publication will be ready in October 2016). The special issue (title: “Multidisciplinary research findings in support to the EU air quality planning: experiences from the APPRAISAL, SEFIRA and ACCENT-Plus FP7 projects”) is a collection of several papers that will present the main results from the APPRAISAL, SEFIRA (Socio Economic Implications for Individual Responses to air pollution policies in EU) and ACCENT-plus (Atmospheric Composition Change – the European Network) FP7 projects.
- The publication of aA book has been submitted to Springer and approved in the SpringerBrief Series. The book is under preparation, and will be published within next Spring. Title of the book is “Air Quality Integrated Assessment: a European perspective “, and it will contain, in comparison to the special issue on “Environmental Science and Policy”, a more detailed analysis of the APPRAISAL project results.
The project results will be presented at the following conferences:
- Zone Atelier Environnementale urbaine in Strasbourg (FR), relating to Air Pollution and Energy, November 2015.
- 21st Session of the Conference of the Parties to the United Nations Framework Convention on Climate Change (COP21/CMP11), Paris (FR), December 2015.
- Air Quality 2016, March 2016, Milan (IT).
Dissemination activities at local/regional/national levels
- The project results will be presented in a number of national meetings, as in the Meeting of the Air Quality Technical Committee coordinated by the Portuguese Agency for the Environment with representatives of the regions of Portugal, November 2015.
- A stakeholder mailing list has been compiled including (1) the local, regional and national decision makers that contributed to the project activities or participated to the project conferences, (2) the SMEs and (3) the research institutes that expressed their interest in the methodologies and tools tested during the project. The project results will be disseminated sending the APPRAISAL Layman’s Report and the Guidance document to such a stakeholder mailing list. The stakeholders will be informed of any further initiatives addressed to the local/regional decision makers.
- RIAT+ model and software has been identified as an example of the decision framework defined in the APPRAISAL project to support local/regional decision makers. JRC is providing the RIAT+ first-guess dataset for European regions. This dataset will be released by end 2015, and will contain data on emission, source-receptor models and emission reduction measures useful to start using RIAT+ with a simplified input configuration. This will allow the decision makers of the European regions to apply in a simple (even if preliminary) way an Integrated Assessment Modelling system to identify effective air quality improvement measures. The APPRAISAL stakeholders will be informed about the availability of the RIAT+ first-guess dataset by email.
Collaboration with other projects/institutes
- APPRAISAL has been cooperating with the SEFIRA project exploring new methodologies to include the social acceptability in the integrated assessment models at local and regional scale.
- The APPRAISAL results will contribute to the FAIRMODE programme and the e-reporting process designed by the EEA to collect data on Air Quality plans in Europe.
- The JRC, Ispra, is developing the SHERPA (Screening for High Emission Reduction Potentials on Air quality) tool, to evaluate in a fast way how “emission changes” influences concentration. SHERPA and RIAT+ will be used in a sequential and integrated way: (1) SHERPA to perform a first screening and (2) RIAT+ in cases in which a further/more detailed analysis (involving costs) is needed.
List of Websites:
The project website (http://appraisal-fp7.terraria.com/site/index.php) includes the project aims, results, deliverables and the EU Air Quality Plans and Projects Database.
Contacts:
Project Coordinator:
Marialuisa Volta (marialuisa.volta@unibs.it)
Department of Mechanical and Industrial Engineering
University of Brescia
Brescia, Italy
Communication Manager:
Philippe Thunis ( philippe.thunis@jrc.ec.europa.eu)
European Commission JRC – IES
Ispra (VA), Italy
Financial Manager:
Elena Murelli (murelli.elena@gmail.com)
University of Brescia
Brescia, Italy
