applications of models-3 in coastal areas of canada m. lepage, j.w. boulton, x. qiu and m. gauthier...
TRANSCRIPT
Applications of Models-3Applications of Models-3in Coastal Areas of in Coastal Areas of
CanadaCanada
M. Lepage, J.W. Boulton, X. Qiu and M. Gauthier
RWDI AIR Inc.
C. di CenzoEnvironment Canada, P&YR
3rd Annual CMAS Models-3 WorkshopOctober 18-20, 2004 Chapel Hill, NC
Eastern Domain(36-km)
Western Domain(12-km)
120°0'0"W130°0'0"W140°0'0"W150°0'0"W
110°0'0"W
110°0'0"W
100°0'0"W
100°0'0"W 90°0'0"W
90°0'0"W
80°0'0"W
80°0'0"W
70°0'0"W
70°0'0"W
60°0'0"W 50°0'0"W 40°0'0"W
30°0'0"N
40°0'0"N
40°0'0"N
50°0'0"N
50°0'0"N
60°0'0"N
30°0'0"N
120°0'0"W
130°0'0"W
140°0'0"W
60°0'0"W
50°0'0"W
40°0'0"W
60°0'0"N
Modelling DomainsModelling Domains
(4-km)(4-km)
Factors to be considered:Factors to be considered:
Meteorological fields (MM5, MC2…)Emission inventory and modelingCMAQ Chemical mechanism
(SAPRC-99, CB-IV & RADM2…)Grid resolution and domainsMonitoring data and evaluation
methods
Western Canada: Western Canada:
MC2, SMOKE, CMAQModeled events: Aug ’01, Dec ‘024 km grid in MC2 (from UBC)Emission Inventories: 2000 (GVRD),
1995 (BC), 1999 (US), BEIS312 km, 4 km grids in CMAQRADM2, SAPRC99
CMAQ Model ResultsCMAQ Model Results
Pitt Meadows
Modelled & Observed Time HistoriesModelled & Observed Time HistoriesOzone, Aug. 9 – 20, 2001Ozone, Aug. 9 – 20, 2001
ppb
Day
ModelledMeasured
Legend
ModelledMeasured
Legend Betternighttimeozone
ppb
Day
Nighttime ozone sensitiveto grid resolution… Day
ppb
Pitt Meadows
NO2
Day
Day
ModelledMeasured
Legend
ModelledMeasured
Legend
STATION PITTMEADO WS Concentrations in ppbv, except CO in ppm and PM2_5 in g/m 3
0 24 48 72 96 120 144 168 192 216 240 2640
20
40
60
80
100
O3_ST N4
PDO3_ST N4
time0 24 48 72 96 120 144 168 192 216 240 264
0
20
40
60
80
100
120
NO2_STN4
PDNO2_STN4
time
0 24 48 72 96 120 144 168 192 216 240 2640
20
40
60
80
100
120
NOx_STN4
PDNOx_STN4
time
0 24 48 72 96 120 144 168 192 216 240 2640
10
20
30
40
50
60
PM25_ST N4
PDPM25_ST N4
time
0 24 48 72 96 120 144 168 192 216 240 2640
2
4
6
8
10
SO2_STN4
PDSO2_STN4
time
0 24 48 72 96 120 144 168 192 216 240 2640
0.2
0.4
0.6
0.8
1
CO_ST N4
PDCO_ST N4
time
Under-prediction
Ozone
Good NO2
Predictions
Recent tests suggest better results with SAPRC99…
ppb
ppb NO2
Modelled & Observed Time HistoriesModelled & Observed Time HistoriesOzone, Aug. 9 – 20, 2001 (cont’d)Ozone, Aug. 9 – 20, 2001 (cont’d)
CMAQ Model ResultsCMAQ Model Results
Modeled & Observed Time Histories:Modeled & Observed Time Histories:PM2.5, Aug. 9 – 20, 2001PM2.5, Aug. 9 – 20, 2001
ModelledMeasured
Legend
ModelledMeasured
Legend
STATION PITTMEADO WS Concentrations in ppbv, except CO in ppm and PM2_5 in g/m 3
0 24 48 72 96 120 144 168 192 216 240 2640
20
40
60
80
100
O3_ST N4
PDO3_ST N4
time0 24 48 72 96 120 144 168 192 216 240 264
0
20
40
60
80
100
120
NO2_STN4
PDNO2_STN4
time
0 24 48 72 96 120 144 168 192 216 240 2640
20
40
60
80
100
120
NOx_STN4
PDNOx_STN4
time
0 24 48 72 96 120 144 168 192 216 240 2640
10
20
30
40
50
60
PM25_ST N4
PDPM25_ST N4
time
0 24 48 72 96 120 144 168 192 216 240 2640
2
4
6
8
10
SO2_STN4
PDSO2_STN4
time
0 24 48 72 96 120 144 168 192 216 240 2640
0.2
0.4
0.6
0.8
1
CO_ST N4
PDCO_ST N4
time
PM2.5 (µg/m3)
STATION ABBOTSFORD Concentrations in ppbv, except CO in ppm and PM2_5 in g/m 3
0 24 48 72 96 120 144 168 192 216 240 2640
20
40
60
80
100
O3_ST N5
PDO3_ST N5
time0 24 48 72 96 120 144 168 192 216 240 264
0
20
40
60
80
100
120
NO2_STN5
PDNO2_STN5
time
0 24 48 72 96 120 144 168 192 216 240 2640
20
40
60
80
100
120
NOx_STN5
PDNOx_STN5
time0 24 48 72 96 120 144 168 192 216 240 264
0
10
20
30
40
50
60
PM25_ST N5
PDPM25_ST N5
time
0 24 48 72 96 120 144 168 192 216 240 2640
2
4
6
8
10
SO2_STN5
PDSO2_STN5
time
0 24 48 72 96 120 144 168 192 216 240 2640
0.2
0.4
0.6
0.8
1
CO_ST N5
PDCO_ST N5
time
No2 (ppb)
Precursor diurnalPattern propagatedto secondary PM
p-nitrate (ug/m3)
Too strong diurnal circle
Eastern Canada: Eastern Canada:
MM5, SMOKE, CMAQModeled events: Jul ’99, Aug ‘01Emission Inventories: 1995 (EC),
1999 (US), BEIS336 km, 12 km, 4 km grids in CMAQCB4, SAPRC99
Model Performance: OzoneModel Performance: Ozone Time Histories for Nearby, 4-km Resolution Urban (Time Histories for Nearby, 4-km Resolution Urban (redred) )
and Suburban (and Suburban (greengreen) Monitoring Sites, Hamilton, Ontario) Monitoring Sites, Hamilton, Ontario
Ozone
Predicted (blue) nighttime ozone aligns better
with the suburban monitoring site
Daytime obs ozone the same at both sites
PM2.5
Sulfate dominated aerosol: precursor diurnal cycles not propagated.
Model Performance: PMModel Performance: PM2.52.5
Time Histories for Nearby, 4-km Resolution Urban (Time Histories for Nearby, 4-km Resolution Urban (redred) ) and Suburban (and Suburban (greengreen) Monitoring Sites, Hamilton, Ontario) Monitoring Sites, Hamilton, Ontario
CMAQ PerformanceCMAQ Performance
Hourly and daily Ozone predictions are overall reasonably good in western and eastern Canada, (sometimes poor at night)
Kz(min) could be a contributing factor? Grid resolution is important for ozone
predictions, especially at nighttime (NOx): higher resolution is better
Meteorological inputs can be a major factor Chemical mechanism also plays a role
CMAQ Performance (cont’d)CMAQ Performance (cont’d) Daily mean PM predictions are reasonably
good in western and eastern Canada Hourly values are generally poor, but
somewhat better for Eastern Canada SOA module could be a contributing factor? Better biogenic emissions and temporal factors
are needed, especially in western Canada due to biogenic SOA.
Meteorological inputs can be a major factor Chemical mechanism also plays a role
MM5 or MC2 can produce the necessary MM5 or MC2 can produce the necessary meteorological inputs. However, there is meteorological inputs. However, there is uncertainty in PBL predictions relating touncertainty in PBL predictions relating to::
Model grid resolution Geophysical fields Initial and boundary conditions Microphysics and convective
parameterization schemes PBL and surface schemes ???
MC2 Sensitivity StudyMC2 Sensitivity Study
Sensitivity tests on 4 km and 2 km grid resolutions (13 runs) combining:
24km GEM output as initial fields ISBA surface scheme Kain-Fristch deep convective scheme Kong-Yau microphysics scheme Improved geophysics resolution Increased vertical grid resolution
P. Temperature Profiles - MC2 Sensitivity Runs LGY, 2001081912
0100200300400500600700800900
1000
280 282 284 286 288 290 292
Theta (K)
z (m
-ag
l)
fcrest_4km_base
isba_2km_base
isba_2km_chgall
isba_2km_kfcky
isba_4km_base
isba_4km_chgall
isba_4km_chgmoist
isba_4km_chgtopo
isba_4km_chgvg
isba_4km_chgz0
isba_4km_kfcky45lyr
isba_4km_kfcky
OBS
OLD MC2
MC2 Sensitivity Study:MC2 Sensitivity Study:Preliminary ResultsPreliminary Results
CONCLUSIONS CONCLUSIONS
Reasonably good performance for 1-hr and 24-hr ozone and 24-hour PM2.5
Better ozone diurnal pattern on 4 km grid than on 12 km grid
Working on improvements to meteorological fields
Problems with PM and SOA chemistry – now working with SAPRC99