Canadian Operational Air Quality Forecasting Systems: Status, Recent Progress, and Challenges 2 nd CAMS General Assembly, Warsaw, Poland , 16-18 May 2017 Authors: Radenko Pavlovic 1 , Didier Davignon 1 , Sylvain Ménard 1 , Rodrigo Munoz-Alpizar 1 , Hugo Landry 1 , Samuel Gilbert 1 , Paul-André Beaulieu 1 , Michael D. Moran 2 , Jack Chen 3 , Verica Savic-Jovcic 2 , Paul Makar 2 , Craig Stroud 2 and Cynthia Whaley 2 1 Air Quality Modeling Applications Section, Environment and Climate Change Canada, Montreal, Quebec, Canada 2 Air Quality Research Division, Environment and Climate Change Canada, Toronto, Ontario, Canada 3 Air Quality Research Division, Environment and Climate Change Canada, Montreal, Quebec, Canada
21
Embed
Canadian Operational Air Quality Forecasting Systems: Status, … · 2017-05-22 · Canadian Operational Air Quality Forecasting Systems: Status, Recent Progress, and Challenges 2
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Canadian Operational Air Quality Forecasting Systems: Status, Recent
Progress, and Challenges
2nd CAMS General Assembly, Warsaw, Poland , 16-18 May 2017
Authors: Radenko Pavlovic1, Didier Davignon1, Sylvain Ménard1, Rodrigo Munoz-Alpizar1, Hugo Landry1, Samuel Gilbert1, Paul-André Beaulieu1, Michael D. Moran2, Jack Chen3, Verica Savic-Jovcic2, Paul Makar2, Craig Stroud2 and Cynthia Whaley2
1Air Quality Modeling Applications Section, Environment and Climate Change Canada, Montreal, Quebec, Canada2Air Quality Research Division, Environment and Climate Change Canada, Toronto, Ontario, Canada 3Air Quality Research Division, Environment and Climate Change Canada, Montreal, Quebec, Canada
Page 2 – May-22-17
Canadian AQ Forecast Overview
An 18-year-old program that has evolved from an O3-only forecast program in eastern Canada in 1999, to O3 and PM2.5 in 2004 and to a Canada-wide O3, NO2, PM2.5 forecast program in 2007.
As of 2007, forecasts are communicated in most areas as an Air Quality Health Index (AQHI)
– 2-bin sectional representation of PM size distribution (i.e., 0-2.5 μm and 2.5-10 μm) with 8 chemical PM components
– Full process representation of oxidant and aerosol chemistry: gas-, aqueous- &
heterogeneous chemistrymechanisms
aerosol dynamics dry and wet deposition
(including in- and below-cloud scavenging)
GEM-MACH Grid
GEM-LAM10 Grid
Page 6 – May-22-17
RAQDPS Components: UMOS-AQOperational statistical AQ post-processing
At the Canadian Meteorological Centre (CMC) UpdateableModel Output Statistics (UMOS) is applied as a statisticalpost-processing (SPP)system for weather- and AQ-relatedpredictands.
– SPP can compensate for models’ inherent systematic errors– SPP takes into account scales and phenomena not yet resolved
by dynamical models– SPP can be a very helpful tool for operational ECCC AQ
forecasters
At the Canadian Meteorological Centre (CMC) Updateable ModelOutput Statistics (UMOS) is applied as a statistical post-processing(SPP) system for weather- and AQ-related predictands.
– SPP can compensate for models’ inherent systematic errors– SPP takes into account scales and phenomena not yet resolved by dynamical models– SPP can be a very helpful tool for operational ECCC AQ forecasters
Useful tool for urban AQ forecasting.UMOS-AQ sharpens the model forecast, as it
FireWork in 2016The 2016 wildfire season was marked by the Fort McMurray wildfires that occurred in Alberta’s oil sands production area. The fire burnt 590,000 hectares through the city of Fort McMurray, and more than 80,000 people were evacuated.
GOES satellite image showing the area affected by wildfire smoke, analysed by operational Canadian Meteorological Centre forecasters, valid for May 18th 2016 13:30UTC
Forest fire emissions contribution to the total column PM2.5valid at 2014-07-25 13UTC, forecasted by the 2014-07-24 12UTC run. The region affected by dense smoke was forecasted 24h in advance.
OBSERVED FORECASTED
In support to the federal emergency response organisation, in May 2016, the CMC performed different wildfire pollution analyses on a daily basis
Wildfire Pollution Modeling- extent of population exposure to pollution
Percentage of Canadian population exposed to forecasted wildfire pollution above given thresholds during wildfire season (May-Sept)
%
2014 (May-Sept): The percentage of hours with forecasted PM2.5 wildfire emission contribution to total surface PM2.5 concentrations above 1µg/m3
For most Canadians, wildland fire smoke contributes a small but significant fraction of air pollution during the summer
More information about FireWork can be found at: Pavlovic et al., 2016, J. Air & Waste Manage. Assoc., 66, 819-841, DOI: 10.1080/10962247.2016.1158214.
FireWork web page: https://weather.gc.ca/firework/index_e.html
FireWork forecast and product explanations: Canadian Wildland Fire & Smoke Newsletter, 2016, p18-29: https://sites.ualberta.ca/~wcwfs/CWFSN/newsletters/CWFSN_Fall_2016.pdf
- Surface energy, water and momentum balance depends on land use type
- Impacts on Urban heat island, Air Quality, Public health and comfort
- Heterogeneous environments have significant variability at the urban scale
Example of variability in the urban fabric of a large city (New York) (Source: NRCNA, 2012)
• Developed at Meteo-France (Masson 2000)
• Imported into GEM and GEM-Surf during the CRTI projects (Lemonsu, 2005-2007)
• Further improved and tested since by Leroyer
• Multi-layer version also available (Husain)
“Street canyon
Urban AQ ModellingThe Town Energy Balance (TEB): An urban surface parametrisation
Page 14 – May-22-17
Urban AQ ModellingExperimental, Integrated Urban Very High-Resolution Numerical Weather Prediction System with TEB (Town Energy Balance)
• The High Resolution Deterministic Prediction System with 2.5km resolution has been running in experimental mode at the Canadian Meteorological Centre since 2011.
AQ modelling at 2.5kmRAQDPS at 2.5km was run during the 2015 PanAm games
Experimental, Integrated Urban Very High-Resolution Numerical Weather Prediction System(AQ modelling done at 10km and 2.5km)
(Courtesy of S. Leroyer)(Courtesy of C. Stroud)
(Con't)
Page 15 – May-22-17
Experimental AQ modeling at 2.5 km over western Canada• RAQDPS at 2.5 km- since 2013• FireWork - starting summer 2017
– Already tested over summer 2013
• FireWork was implemented at 2.5 km resolution examining ammonia emissions and chemistry in the vicinity of the Canadian oil sands.
• RDPS 10km weather forecast used to drive HRDPS 2.5 km weather forecast and GEM-MACH10km chemistry forecast –these provide initial and boundary conditions for the 2.5 km GEM-MACH forecast.
Courtesy of : C. Whaley, P.A. Makar, M. Shephard, L. Zhang, J. Zhang, Q. Zheng, A. Akingunola, G. Wentworth, J. Murphy
RAQDPS -10km
RAQDPS -2.5km
Aug-Sept 2013 OS simulations with Firework fire emissions (Average surface NH3 concentrations)
Average fire contribution to surface NH3
Page 16 – May-22-17
Upcoming RAQDPS Improvements
• Extend RAQDPS forecasts from 2 to 3 days
• Upgrade anthropogenic emissions– Current emission inventories: 2010 Can, 2011 USA, 1999
Mexico– New emission inventories: 2013 Can, 2017 (projected) USA,
2008 Mexico
• Upgraded version of the Air Quality objective analysis– Improved statistics– 3D analysis – not limited to surface level
Planned for Fall 2017
Page 17 – May-22-17
Upcoming FireWork Improvements• Replacement of FEPS by CFFEPS – Canadian Forest Fire Emissions
Prediction System– Developed by Canadian Forest Service - Natural Resources Canada– Fire spread and growth modelled using forecasted meteorology – Elliptical fire growth model– Three emission types: flaming / smoldering / residual– Updated emission factors (PM, NO, NMHC, NH3, CO, CO2, CH4 etc.)
Plume injection height based on fire energy thermodynamics
black bodyradiation
Qplume
QradiationQevaporation
Qfuel heating
Fincomplete combustion
dQ/dt
Qsurface heating
Qfire
Energy Balance
Preliminary results
Saskatoon
Page 18 – May-22-17
Upcoming Urban AQ Modelling• Experimental urban AQ Modelling, with incorporated TEB, is planned for
2018
• 2 windows (E/W Canada) are considered
• Objective is to include 3 metropolitan areas Toronto, Montreal and Vancouver where more than a third of the Canadian population (~ 13 million) lives
Urban FractionVancouver
TorontoMontreal
Page 19 – May-22-17
Challenges and Possible Future Directions• Changing emissions: Use projected inventories instead
of retrospective base-year inventories?
• Chemical data assimilation to initialize GEM-MACH?– Some testing already done. Planned for long-term operational
delivery.
• Updated or new AQ process representations
• Improved chemical lateral boundary conditions from global GEM-MACH
Page 20 – May-22-17
Challenges and Possible Future Directions• Benefits of higher-resolution deterministic AQ systems?
- Urban-scale systems and services emerging rapidly
• Statistical Post-Processing of AQ forecasts– How to address extreme values, including extreme AQ events such