WP5: Air Quality Forecasting Bas Mijling Ronald van der A Arjan Lampe AMFIC Progress Meeting ● Barcelona ● 24 June 2009
Jan 21, 2016
WP5: Air Quality Forecasting
Bas Mijling
Ronald van der A
Arjan LampeAMFIC Progress Meeting ● Barcelona ● 24 June 2009
AMFIC Work Package 5
5.1 Implementation of a chemical transport model on a 0.25 degree resolution
5.2 Model improvements, especially the incorporation of aerosols in the model
5.3 A better emission database will be incorporated in the model. If possible, this database will be improved based on retrieval-model comparisons
5.4 Implementation of an operational chemical weather forecast service
Overview
1. Implementation and use of Chimere model
2. Validation; High nocturnal values
3. Future steps
1Implementation Chimere model
Overview Chimere
output
CHIMERE
ECMWFmeteo
MELCHIOR scheme44 species120 reactions
INTEX-B2006 inventory
GLGC database1 x 1 km²
emissions
monthly averagesgas phase: LMDZ-INCA2aerosol: GOCART
boundaryconditions
end field of previoussimulation
initial field
landuse
chemistry
Resolution 0.25 x 0.25 degree8 horizontal layersHourly concentrations of44 gases and aerosols
by Qiang Zhangand David Streets
INTEX-B Emissions
Use of Chimere Results
• Forecast of air quality in China for large urban areas– www.amfic.eu/bulletin
• Background values for street level model CERC– www.beijingairquality.cn/
• Background values for regional model VITO– www.vito.be/amfic
• Air quality forecast during Olympic Games– www.knmi.nl/beijing2008
www.amfic.eu/bulletin
www.amfic.eu statistics
Jan 2009 Feb 2009 Mar 2009 Apr 2009 May 2009 Jun 2009*
Totaal 3256 1021 6002 3743 9072 3044
articles 51 69 148 173 120 49
articles.php 15 15 64 51 42 19
bulletin 169 155 1075 1073 1420 616
forecast.php 15 14 54 44 49 9
grounddata.php 8 6 17 19 13 3
index.php 18 22 102 83 62 25
internal 20 61 69 31 389 374
links.php 21 21 69 71 28 19
members.php 20 21 92 62 61 11
overpass_data 1150 1127 1533 303
satellite.php 2760 458 2716 682 5038 1563
validationdata.php 14 21 71 74 50 17
*) up to 16 June
2Validation, Nocturnal Values
Measurements by BBC
PM10 MODEL vs BBC Measurements
0
50
100
150
200
250
300
350
400
2008
0709
2008
0711
2008
0713
2008
0715
2008
0717
2008
0719
2008
0721
2008
0723
2008
0725
2008
0727
2008
0729
2008
0731
2008
0802
2008
0804
2008
0806
2008
0808
2008
0810
2008
0812
2008
0814
2008
0816
2008
0818
2008
0820
2008
0822
2008
0824
2008
0826
PM10 Chimere 13h LT
PM10 BBC
Olympic Games
http://news.bbc.co.uk/2/hi/in_pictures/7506925.stmDaily PM10 measurements at 13h in central Beijing
Comparison with Beijing
Diurnal variation
Measurements (W.-x. Wang et al., 2008Air Qual Atmos Health)
Chimere (0.25°×0.25°)
Measured August 7 to September 30, 2007at Chinese Research Academy of Environmental Sciences
Diurnal profile PM10
Beijing (Aug-Sep 2007) Hong Kong (Aug-Sep 2007)
Time series Hong Kong
PM10 at 15h local time PM10 at 03h local time
19 Aug
26 Aug 6-7 Sep
• Nocturnal boundary layer height too low over megacities (e.g. BLH=20 m for Beijing)
• Trace gases not well mixed due to low vertical diffusion coefficient (e.g. Kz=0.1 m2/s for Beijing)
Caused by a lack of Urban Heat Island Effect: too low sensible heat flux over megacity.
Taking BLH=750 m and Kz=1 m2/s for Beijing decreases nocturnal PM10 with 25-30%
3Future Steps
• Improvements of model– Nocturnal boundary layer– Aerosols– …
• Improvement of emissions– Higher resolution (VITO) – Assimilation of satellite data– Better diurnal cycle by sector– …
Future Steps
Desaggregation of emissions
Original INTEX-B (0.5x0.5)Desaggregated by VITO to
AMFIC grid (0.25x0.25)
Conclusions
• Chimere model is successfully implemented• Daily operational forecasts are available on
website http://www.amfic.eu/bulletin
• Results validate well, but high nocturnal values due to unrealistic boundary layer for megacity.
• More ground sources needed for further validation!
• Model and emissions are to be improved
Thank you!