air quality modeling group-2007-8 - МЖСПП · pdf fileair quality :modeling groupair...
TRANSCRIPT
Air Quality :Modeling GroupAir Quality :Modeling Group
R t d i kRecent and ongoing workGroup members, 31/8/2007
Ari Karppinen Mikhail Sofiev Juha NikmoJari Härkönen Kari RiikonenMarke Hongisto Leena KangasMi P hj l Ilkk V lkMia Pohjola Ilkka ValkamaJukka-Pekka Jalkanen Joana SoaresMervi Haakana Mari Kauhaniemi (Kuopio)J i Jä i Minna Rantamäki (25%)Jouni Jäppinen Minna Rantamäki (25%)
Noora Eresmaa (mat. leave) Leena Partanen (mat leave)Leena Partanen (mat. leave)
Personell: 14.25 + 2
FMI-Budget: 4.75 py+ projects: 9.5 py (300 k€ + 500 k€)Total: 0.8 M€/year (14.25 py + ~100k€ other costs)
http://www.fmi.fi/organisaatio/yhteys_34.htmlhttp://www.fmi.fi/research_air/air_2.html
Dispersion modelling
Development and evaluation ofair quality modelsair quality modelsCombination of meteorological
d l d di i d lmodels and dispersion modelsApplication of models and di i ti f i f tidissemination of information
Focus areas in modelling
1 Integrated modelling systems (from1. Integrated modelling systems (from emissions to impacts, from street canyon to global scale)global scale)
2. Combined utilisation of meteorological models and dispersion modelsmodels and dispersion models
3. Health effects of air pollution, especially d lli f th t ti f dmodelling of the concentrations of and
exposure to particulate matter
Modelling system - FMIWeather prediction models
Dispersion models -long-range, regional
Dispersion and effects models – urban, local
g y
ECMWFHIRLAMRCR
SILAM LRT, meso, radioactivity
UDM FMI b
CAR-FMI, roadsidey
OSPM (NERI), street canyon
UDM-FMI, urban
HILATAR
HIRLAMMBE
AROME,LAPS
LRT, meso ,chem.
ESCAPE, chemical id
BUOYANT, fires
FLUENT, CFD code Aerosol process models: UHMA (U Helsinki, FMI)
MONO32 (U Helsinki, Stadia)
accidents
EXPAND (FMI YTV)( , ) EXPAND (FMI, YTV) population exposure
MPP-FMI, Meteorological pre-processing model
A centre of expertise at the Kumpula campus: wild-land firesFMI Air Quality Earth Observation Climate and Global Change Kuopio Unit and
Ilmanlaadun seuranta-asemat 2003
FMI Air Quality, Earth Observation, Climate and Global Change, Kuopio Unit, and Univ. Helsinki
Joensuu
Sevettijärvi
Oulanka
Hailuoto
Kevo
Raja-JooseppiPallas
Hietajärvi
SodankyläEMEPWMO/GAWEGAPKansallinenYhdennetty seurantaErikoisasemaRadioaktiivisuusOtsoni
Ground-based observations of fire plumes at selected stations and in the TROICA
Ähtäri
Kotinen
Virolahti
Utö
Hietajärvi
Tikkakoski
Nurmijärvi
Helsinki
Evo
Ilomantsi
Jokioinen
Combineduse of data
Earth observation of wildfires: Fire
se ected stat o s a d t e O Ccampaign
Data assimilationand analysis
use o data
ECMWFHIRLAMAtlantic
HIRLAMNorth-Europe
HIRLAMLINUX
ECMWFHIRLAMAtlantic
HIRLAMNorth-Europe
HIRLAMLINUX
Top of the boundary layerPassive plume
Buoyantplume
wind
Gaussianmodel
K-theorymodel
Top of the boundary layerPassive plume
Buoyantplume
wind
Gaussianmodel
K-theorymodel
Earth observation of wildfires: Firespread and plume dispersion
analysis
.
. ..
p
Air flow
Burning region
Deposition
...
.. ...
.. . . .
Particles
.
. ..
p
Air flow
Burning region
Deposition
...
.. ...
.. . . .
Particles
Modelevaluation
Modelling of fire spread, plume rise, atmosphericdispersion and source apportionment
Controlled forest burningexperiment
NOx emissions from marine trafficJukka-Pekka
R l ti it i t• Real-time monitoring system for ship emissions• Position reports from p
transponder messages• NOx estimate based on
available technical dataavailable technical data• Current speed vs. design
speed • Possibility to track
emissions ship by ship• Combine emission data with• Combine emission data with
3D atmospheric models to model long range transport
Mari, Mervi
Modelling NOx and NO2 concentrations in different urban environments
Vehicle Type Distribution
Models:• CAR-FMI (open area line source model)• OSPM (street canyon model)
Vehicle Type DistributionVehicle Type2
Vehicle Type3
Hourly Distribution of Traffic
0.060.070.080.090.10
ffici
ent
OSPM (street canyon model)
Conditions:• 4 open area line source cases
Vehicle Type1
0.000.010.020.030.040.05
OSP
M c
oefp
• 6 street canyons cases• 2 years• 3 heights in street canyons
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24timeWeekdays Saturdays Sundays
3 heights in street canyons• NOx and NO2 emission factors
Daily and Monthly Distribution of Traffic
1.2
1.3
s0.7
0.8
0.9
1.0
1.1
1.2
OSP
M c
oeff
icie
nts
Example of the street canyon in the OSPM
0.5
0.6
1 2 3 4 5 6 7 8 9 10 11 12time
O
Mon Tue Wed-Fri Sat Sun
Mari
Publication (in progress for Atm. Env.)Evaluation and application of a modelling system forEvaluation and application of a modelling system for predicting the concentrations of PM2.5 in an urban areaKauhaniemi M1, Karppinen A.2, Härkönen J.2, Kousa A.3, Alaviippola B.3, Koskentalo T.3, Aarnio P.3, Elolähde, T.3 and Kukkonen J.2 (1 = FMI Kuopio, 2 = FMI Helsinki, 3 = YTV)
VALLILAweekdays (in 2002)
combust12.7 %
KALLIOweekdays (in 2002)
non-combust
8.5 %
cold start
combust4.7 %
non-combust22.8 %
cold start3.5 %
long-range transport61.0 %
cold start1.3 %
long-range transport85.5 %
VALLILA
40
45
50
55
2:1 1:1
KALLIO
40
45
50
55
2:1 1:1
15
20
25
30
35
Obs
erve
d (µ
g/m
3 )
1:2
15
20
25
30
35
Obs
erve
d (µ
g/m
3 )
1:2
y = 0.94x - 0.630
5
10
0 5 10 15 20 25 30 35 40 45 50 55
Predicted (µg/m3)
R2 = 0.54N = 360
y = 0.95x + 1.150
5
10
0 5 10 15 20 25 30 35 40 45 50 55
Predicted (µg/m3)
R2 = 0.60N = 358
Joana
Traffic iF for BZ for different urban environmentsstreet canyon (OSPM) Helsinki downtown area (PC-CAR) and the whole Helsinkistreet canyon (OSPM), Helsinki downtown area (PC CAR) and the whole Helsinki Metropolitan Area (URB-CAR).
• PM2.5 iF spatial distribution from a single source: Salmisaari cogeneration PP
Ilkka Valkama/FMI
• Benzene regional dispersion modelling• Aerosol dynamics implementation
SILAM
Kari
Salmisaari power plant: iF of PM2.5
Intake fraction =
Concentration x Breathing rate x ActivityEmission
Leena
Intake fraction (iF)fraction (iF) studyfor PMfor PM2.5emissions from thefrom the cogenerating power plantpower plant Salmisaari in HelsinkiHelsinki
PM2.5 iF for 2005 calculated by EXPAND
Leena
Intake fraction (iF) studyfor traffic emitted benzene in Helsinki Metropolitanfor traffic emitted benzene in Helsinki Metropolitan area
1 6
1
1.2
1.4
1.6
0.4
0.6
0.8
µg m
-3 ModelledMeasured
0
0.2
Kallio Tikkurila Lintuvaara
Comparison of modelled and measured annual average benzene concentrations Benzene iF for 2005 calculated by EXPAND
Leena
Local scale dispersion calculations for fine particle emissions from domestic combustion and road trafficfine particle emissions from domestic combustion and road traffic
Leena, Mia
Dispersion studies over a noise wall
southern wind, ~2 m/s
Dispersion 30.3.2007
300
S
150
200
250
9 (u
g/m
3)
N
0
50
100PM9
stationary measurements
-2 10 20 30 40 50 60 70
Distance from the wall
stationary measurements on the sidewalk
Measurements will be compared with model resultsMeasurements will be compared with model results
Kari,Ilkka,Juha
NBC Modelling & Simulation
OVERALL OBJECTIVE• Develop improved model chains for simulation of NBC-scenarios for
planning and decision support
• SPECIFIC GOALS FOR THIS PROJECT• SPECIFIC GOALS FOR THIS PROJECT• Clarify uncertainties in NBC M&S
- Compare results from national model chains for reference scenarios- Emergency response (very fast) ---- studies (“unlimited” time for calc.)- Identify key weaknesses
M d lli G id li• Modelling Guidelines• Best Practice• Possible start of model developmentPossible start of model development
Ilkka
Particle-cloud formation
- in SILAM (above)- compared to ”traditional” cylinder
Ilkka Valkama/FMI
ycloud (left)
Kari,Ilkka,Juha
EXPLOSION OF AN IMPROVISED NUCLEAR DEVICE (FI)
Helsinki RN-ScenarioApproximate evolution of the cloud( j ti th d l l f ll h i ht )(projection on the ground level from all heights)
Ilkka
SILAM Nuclear Fallout Modulecloud parametres- cloud parametres
- particle formation- activity distribution
CLOUD dimensions
140
80
100
120
ht (k
m)
Cloud radius(obs)SILAM
0
20
40
60
heig
h SILAM
42.097.0 WRadius =
00.1 10 1000 100000
yield (kT)Ilkka Valkama/FMI
Ilkka
Particle-cloud formation
- in SILAM (above)- compared to ”traditional” cylinder cloud (left)
Ilkka Valkama/FMI
Jari
Further development of MPP-FMI utilizing p gsatellite data (ESA)
Available data:
• MODIS: T-profile product every 6th hour (in cloud-f diti )
Available data:
free conditions)• RS-data twice a day• Ground surface measurements at least once a hour
Jari
Polar orbiting satellites:Aqua Terra
HELSINKITESTBED:MODIS
Test PlanAqua, TerraRS - RASS
- SODAR
Test Plan
700700 km
SO- CEILOMETER- Meteorological
0
k
700 km and air qualitymonitors
m0-7 km
100 km100 km
Jokioinen Helsinki
Mia
VIEME-project:• Computational fluid• Computational fluid
dynamics study of dispersion of particulate matter pfrom a highway over a noise wall
• PM will be simulated as an inert substance
• 2D, standard k-ε
90 m behind the noise wallnoise wall
Mia
COST-732
• “QUALITY ASSURANCE AND IMPROVEMENT OF MICRO-SCALEMICRO SCALE METEOROLOGICAL MODELS”
• 23 members• 23 members• Test case of 120
containers on a field• To be developed:
Guideline on how to model dispersion from thedispersion from the experiences obtained
Mia
SRIMPART: Source-receptor and inverse modelling to quantify urban particulate emissionsquantify urban particulate emissions
• Aims:• Aims:• To improve emission estimates with emphasis on
emissions from domestic wood burningemissions from domestic wood burning.• To improve emission estimates for wood burning
applied in the EMEP model and in other internationalapplied in the EMEP model and in other international modelling systems
• To intercompare and assess uncertainty in p yindependent methodologies for assessing emission rates of particulates from urban sources
Mia
• AIM: provide an overview of the latest research findings on air quality and health to support European sustainable development action plans and strategies
• WP 4: “Report on current state-of-the-art research in urban air quality and health”
CAIR4HEALTH• The Air Quality Key Questions have been decided on, and now
answers are being drafted for them:H U b Ai P ll t t LEVELS CHANCING i th l t• How are Urban Air Pollutant LEVELS CHANCING in the long term - past 10 years and future trends, including expected effects of climate change? (Editor: UH, writers UH, FMI, AUTH)A th t ll t t li it l tt i bl th ff ti• Are the current pollutant limit values attainable – the effectiveness of REDUCTION MEASURES? (Editor: UH, writers UH, FMI, AUTH)Wh t th t hi t f INTEGRATED• What are the current achievements of INTEGRATED ASSESSMENT TOOLS for estimating the health impacts of air pollution? (Editor: AUTH, writers AUTH, TNO, FMI)Wh t i k b t th CHEMICAL COMPOSITION AND SIZE• What is known about the CHEMICAL COMPOSITION AND SIZE DISTRIBUTION OF URBAN PM and their HEALTH EFFECTS? (Editor: FMI, writers: UH, FMI)
CLOUD dimensions
70 2248.03104 WH =
30
40
50
60
ght (
km)
Cloud-top(obs)
310.4 WHtop
305.0917.3)20( WktWH =≤
0
10
20
30
0 1 10 1000 100000
hei
SILAM2-param
1774.0362.6)20( WktWH =>
0.1 10 1000 100000
yield (kT)CLOUD dimensions
40
2296.04932 WH = 2025303540
ht (k
m)
493.2 WHbottom =
3433.0166.2)20( WktWH =≤ 05
101520
heig
h
Cloud-bottom(obs)SILAM21594.0439.4)20( WktWH =>
00.1 10 1000 100000
yield (kT)
2-param
Ilkka Valkama/FMI
Activity-particle size distribution
22)(l (1)(l (1 ⎤⎡ ⎞⎛⎤⎡ ⎞⎛
Radioactivity-particle size distribution (bi-modal log-normal distribution)
23 )(ln(21)(ln(
21
)2()1(
)2()( ⎥
⎥⎦
⎤
⎢⎢⎣
⎡⎟⎟⎠
⎞⎜⎜⎝
⎛ −−
⎥⎥⎦
⎤
⎢⎢⎣
⎡⎟⎟⎠
⎞⎜⎜⎝
⎛ −−
−+= βα
βα
βπβπ
r
tv
r
tv e
rA
fer
AfrA
Log-normal distributions for volatile (left) and volatile (right) radionuclides.Ilkka Valkama/FMI
Available (/developed in Ubicasting): dense obs. N t k ( t & AQ) &
Now : offline met-data
Network (met & AQ) & high resolution met-modeling =>
Dispersion forecasts for toxic industrial chemicals and other hazardous releases (terrorism etc) GOAL :
Realtime dispersion & forecasts (1- 6 h)p ( )
New met-model-> OPERATIVEdispersion modellingdispersion modelling
Tailored AQ dispersion tool
• real time tool for assessing
ValitsekaasuLeviämis-
for assessing accidental releases
+1h
+2h
+1h
+2h
kaasu
kaasu 1
kaasu 2
ennuste
•Based on real-time met-
+3h+3h kaasu 3
kaasu 4
measurements
Säätilanne
Anturi1
Anturi2
Common goal: IL, YTV , Vaisala
C bi IL & YTV V i l t & k h• Combine IL & YTV, Vaisala measuremetns & knowhow :- Not only information on stations: also real-time AQ-
situation for the region & 1- 2 day forecastsg y- Prequisite: also the dispersion models need to be modified
– to be able to utilize all available (high-resolution) data
current AQ
Warnings
AQ & exposure -forecast
Warnings
Practical examples
Meteorological data:•FMI 3 meas.stations•MetPP-FMI/HIRLAM NO spatial variation in the met-pfields
bi iUbicastingMeteorological data•Helsinki Testbed (>100 st.)•Hirlam/Arome (3-9 km resolu.)•LAPS (1 km resol.)( )Met-resolution up to 1 km
Web & wapReal time information only at the YTV-stations
Ubicastingg
Web, wap, sms, mmsmmsReal time information for user defineduser defined (/GPS) )locations & periods
OSCAR:Need for an Integrated approach
Graphical User Interface/Control Module
PC OSCAR ASSESSMENT SYSTEM
Graphical User Interface/Control Module
PC OSCAR ASSESSMENT SYSTEM
Visual Basic/Visual.NET
Emissions Database Met Database AQ DatabaseEmissions pre-processor Met pre-processor AQ Data pre-processor
Input pre-
Visual Basic/Visual.NET
Emissions Database Met Database AQ DatabaseEmissions pre-processor Met pre-processor AQ Data pre-processor
Input pre-
Level I ModelsCorrelations of C U Traffic
Level II ModelsEg DMRB, CAR Int PEARL
Level III ModelsEg CAR-FMI, OSPM GAMMA
Numerical Models ADREA-HF MIMO
External ModelsEg Noise
Input preprocessor
Level I ModelsCorrelations of C U Traffic
Level II ModelsEg DMRB, CAR Int PEARL
Level III ModelsEg CAR-FMI, OSPM GAMMA
Numerical Models ADREA-HF MIMO
External ModelsEg Noise
Input preprocessor
C, U, Traffic Int, PEARL OSPM, GAMMA HF, MIMOEg Noise
Spatial/temporal distributions – CO, NO2, PM10, PM2.5, Assessment parameters
C, U, Traffic Int, PEARL OSPM, GAMMA HF, MIMOEg Noise
Spatial/temporal distributions – CO, NO2, PM10, PM2.5, Assessment parameters
Raw data:- Off line- Excel- Access
Visualisation:- Embedded - GIS-VIS5D, PAVE
Scenario Analysis:- Transport- Emissions
AQ Assessment:- Limit values- Annual means
Raw data:- Off line- Excel- Access
Visualisation:- Embedded - GIS-VIS5D, PAVE
Scenario Analysis:- Transport- Emissions
AQ Assessment:- Limit values- Annual means
WEB OSCAR ASSESSMENT SYSTEMWEB OSCAR ASSESSMENT SYSTEM
CAR-FMI : further development• EU/OSCAR (->2005)EU/OSCAR ( >2005)
• An ”international” version of CAR-FMI was integrated in to the OSCAR system
• Main issue: • users can easily use their own emission data
• ONGOING work:• Stand-alone international version
• Co-ordinate systemsy• Several options for (emission) input format• Routines to complement for missing meteorological p g g
parameters• Links to ArcInfo(GIS) / now only MapInfo
SummaryDevelopment and testing of disperion models ongoingDevelopment and testing of disperion models ongoing
in all scales : form microscale (CFD) to regional scale
Lot of effort is put on ”generalizing” the existingLot of effort is put on generalizing the existing models for wider user community -> from research models to ”easy-to-use” tools & services (strong co-
ti ith th /i tit t / i )operation with other groups/institutes/companies)
Linking dispersion models with new meteorological models and health risk assessment models increasingly important part of the work : final aimincreasingly important part of the work : final aim always a complete chain from political decisions to health effects