Hyper-local scale air quality modelling Fernando Martn & Jos - - PowerPoint PPT Presentation

Hyper-local scale air quality modelling

Fernando Martín & José Luis Santiago CIEMAT

What means hyperlocal-scale modelling?

• “Hyper-Local scale modelling” refers to air quality modelling at

high spatial resolution (typically order of meter scale), usually focused on urban environments.

• These type of applications include

– Computational Fluid Dynamics (CFD) models – More simplified tools like street box models or parameterized Gaussian plume models.

Natural Background Regional Background Urban Background

Urban Zone Rural Zone

Local Sources

History

• Recently, some FAIRMODE participants expressed the need for more guidance

and exchange of best practices in “hyper-local scale” AQ modelling.

• Steering Group decided to investigate to what extent FAIRMODE could provide

support.

• 2014-2017. FAIRMODE exercises related to spatial representativeness, in which

some micro-local scale modelling was applied.

• Some sessions in former FAIRMODE meetings,

– Athens, 2017, first contact among few interested people on CFD modelling, – Tallin, 2018, special session on CFD modelling with more than 25 people attending. CFD gained attention for practical application linked to the AQD. FAIRMODE is good for exchange ideas and best practices, provide guidance… – Warsaw, 2019, CFD is quite interesting topic but competence building and benchmarking are yet necessary

• HARMO (Bologna, 2017). Meeting of the FAIRMODE people (about 5 groups

involved)

• Previous guidance and model comparisons:

– COST ACTION 732 (microscale meteorological models), 2005-2009 – COST ES1006 (local-scale emergency) 2011-2015

Priority topics: NCP survey

Questionnaire

• A short questionnaire was sent to a short group of people who

previously has shown some interest in very high resolution modelling and it is well-known they are working on it.

• Objective: a first step to clarify the relevance of the topic in the

context of FAIRMODE with the aim to support the discussions during the technical meeting in Madrid.

• Questions:
• 1. About relevance and purpose.
• 2. Tools
• 3. Expected support from FAIRMODE
• 4. Participation in this group

Results

9 responses:

• VITO (Bino Maiheu), BELGIUM
• IVU Umwelt GmgH (Florian Pfafflin), GERMANY
• RIVM (Joost Wesseling), THE NETHERLANDS
• City of Stockholm (Kristina Eneroth), SWEDEN
• COWI AB (Marie Haeger Eugensson), SWEDEN
• Aarhus University (Matthias Ketzel & Ulas Im), DENMARK
• Government of Styria (Dietmar Oettl), AUSTRIA
• University of Aveiro (Vera Rodrigues & Alexandra Monteiro),

PORTUGAL

• CIEMAT (Fernando Martin & Jose Luis Santiago), SPAIN

• Results. Question 1.
• How relevant is hyperlocal scale modelling for you in the framework of the Air

Quality Directive?

– YES for all. Someone said “very relevant” – Someone said: “it is possibly even more relevant than all the other "large-scale" modelling as limit values are (mostly) exceeded at the local level (hot-spots) and all the effort in large- scale modelling often serves only to provide background concentrations for local applications”

• What is the purpose of this type of modelling applications (assessment,

planning, population exposure, spatial representativeness of AQ stations…)?

–

• Assessment. 8

–

• Planning. 6

– Population exposure. 6 – Spatial representativeness. 5 – Support for compliance to EU regulations. 1 – Impact of emissions on air quality. 1 – Source apportionment. 1 – Validation of sensors. 1

• Results. Question 2.
• What kind of modelling tools (CFD model, parametric

models, street box models, others) are you using for the hyperlocal scale modelling?

– Complex models: 8 groups

• CFD models (Open FOAM, MISKAM, FLUENT, STAR CCM,

VADIS, GRAL, …) 8

• Lagrangian models 2

– Simplified models: 6 groups

• Parametric models – street box models (OSPM,…) 4
• Gaussian models (AERMOD, URBAIR, OML Multi) - Line-

based models (OML Highway) 4

• Others (Dutch OSP model) 1

• Results. Question 3 (1/3)
• What kind of support would you expect from FAIRMODE for

the hyperlocal scale modelling? – Benchmarking. Model Quality Objectives. Comparison.

• Model validation. Adaptation of FAIRMODE tools (Delta)
• Methodologies for compute AQ indicators (i.e. annual averages).
• Spatial representativeness, planning

– Collecting good practices. Recommendations. Guidance.

• How to use high resolution models especially CFD. Boundary

conditions, background concentrations

• Model Validation
• Methodologies for compute AQ indicators. Reporting exceedances
• Population exposure, spatial representativeness, planning

– Harmonisation

• Methodologies for compute AQ indicators
• Reporting exceedances

• Results. Question 3 (2/3)
• What kind of support would you expect from FAIRMODE for the

hyperlocal scale modelling? – Discussion

• Planning
• Model uncertainties
• Population exposure
• Air quality assessment. Reporting exceedances

– Providing/improving input data. Checking input data quality.

• Urban morphology of the European cities;
• High resolution emission data (methodologies for computing them)
• Hourly meteorological data (COPERNICUS)

– Hyperlocal scale composite mapping

• Results. Question 3 (3/3)
• To what extent is the guidance already provided by

FAIRMODE (e.g. Modelling Quality Objective,…) insufficient for this type of applications?

– No one answered directly this question but someone pointed it is good to review and adapt some FAIRMODE guides and tools (for example DELTA tool) – Others suggests to follow some existing guides from COST732 or national guides (Sweden)

• Results. Question 4.
• Would you contribute to an activity on hyperlocal

scale modelling within FAIRMODE (provided that the previous questions point to a real need)?

– Yes 2 – Yes, depending on activity 1 – Yes, limited by funding or personal resources 5 – Yes, but not sure 1 – No 0

Other remarks

• Some people suggest to rethink the term “Hyper-Local”

choosing other more known term (microscale, micro local scale, street scale, etc…)

• Do we agree that an specific group on Hyper Local Scale

modelling must be created in FAIRMODE?

• If YES, what can we do?

Discussion

Topics

• Modelling. Input data. Boundary

conditions

• Model uncertainties & validation
• AQ indicators estimates
• AQ assessement
• Reporting exceedances
• AQ planning
• Population exposure estimate
• Spatial representativeness
• Mapping
• Validation of sensors

FAIRMODE contributions

• Discussion
• Good practices
• Model Quality Objectives
• Bechmarking. Comparison
• Recomendations. Guidance
• Providing tools and data
• Harmonisation
• Competence building

FAIRMODE CONTRIBUTIONS Input data Boundary Conditions validation & uncertainties AQ indicator estimates AQ assessment AQ Planning Population Exposure Spatial representativeness Reporting exceedances Mapping Validation

• f sensors

Discussion

X X X X X X

Good practices Recommendations Guidance

X X X X X X

MQO- Benchmarking

X X X X X

Providing tools & data

X X X

Harmonisation

X X

Competence building

TOPICS

Discussion on work proposal

Current Good Practices

Modelling Validation AQ indicators Specific applications

Discussion Conclussions What is needed to do?

1 2

MQO Benchmarking Comparison Recomendations Guidance Harmonisation Tools Input data base (urban morphology, emission data,) Boundary conditions

3