Using meso-scale models to make
decisions
Bernard Fisher
Environment Agency key messages from meso-scale models
Meeting Hertfordshire 14 December 2010
Evaluation of complex models is difficult
Inter-comparisons are a good way of understanding model behaviour
Operational evaluation is not enough
Uncertainty can be averaged in integrated sectorial or domain aggregated
assessments
Local uncertainties may be a problem particularly for a regulator
Sector changes v individual sources
Source attribution needs a model or AIRTRACK analysis
We have looked at footprints - regulatory and diagnostic value
Particularly for secondary pollutants such as ozone and PM which are difficult
to do any other way. One needs a chemical transport model
Capability of regulator’s staff. Be practical.
Multiple policy options or sensitivity analysis.
ADMS and AERMOD
around here 1000.00
900.00
800.00
1
0.50
700.00
600.00 0.30
0
0 100 200 300 400 500
500.00
0.10
Average dependence on distance
r along radial trajectory 400.00
0.05
300.00
Annual average NO2 concentration
from Heathrow in µg m-3 200.00
0.01
CMAQv4.6 at 1km resolution
100.00
0.00
0.00 100.00 200.00 300.00 400.00 500.00 600.00
1
0
0 1 00 2 00 3 00 4 00 5 00
Average dependence on distance r
along radial trajectory
Annual average PM10 concentration from power station in µg m-3
CMAQ v4.6 at 5km resolution
0.45
1
0.40
0.35
0
0 100 200 300 400 500 0.30
0.25
Average dependence on distance
0.20
r along radial trajectory
0.15
0.10
0.05
Annual average PM10 concentration from power station in µg m-3
CMAQ v4.7 at 5km resolution
1000.00
Similar behaviour is seen 900.00
1
when considering PM2.5
800.00
0.45
700.00
0.40
0
0 100 200 300 400 500
600.00 0.35
2
1000.00 0.30
500.00
0.25
900.00
1 400.00 0.20
800.00
0.15
0.45 300.00
0 700.00 0.10
0 100 200 300 400 500 0.40
200.00
0.05
600.00 0.35
100.00
0.30
500.00
0.25 0.00
0.00 100.00 200.00 300.00 400.00 500.00 600.00 700.00
400.00 0.20
300.00
0.15 CMAQ v4.6 PM2.5
0.10 interpolated at 5km
200.00
0.05
100.00
CMAQ v4.7
0.00
0.00 100.00 200.00 300.00 400.00 500.00 600.00 700.00 PM2.5
interpolated at 1km
1000.00
900.00
1 0.45
800.00
0.40
700.00
0 0.35
0 100 200 300 400 500
600.00
0.30
Average dependence on distance
500.00 0.25
r along radial trajectory
0.20
400.00
0.15
Annual average PM2.5 300.00
0.10
concentration from Fawley
Refinery for 2003 using 200.00
0.05
CMAQ v4.6 model in µgm-3.
100.00
0.00
0.00 100.00 200.00 300.00 400.00 500.00 600.00
1000.00
900.00
800.00
Annual average O3 -1.000E-005
concentration from Fawley 700.00
Refinery for 2003 using -1.000E-004
CMAQ v4.6 model in µgm-3.600.00
-1.000E-003
500.00
-1.000E-002
400.00
300.00 -1.000E-001
200.00
-2.000E-001
100.00
0.00
0.00 100.00 200.00 300.00 400.00 500.00 600.00
Example of the footprint from a stationary major point source e.g. power
station. This is needed for regulation (1) response to change, (2) source
apportionment (responsibility).
1
0.5
900.00
0 800.00
0 100 200 300 400 500
0.45
700.00
Average dependence 1 2
Radial trajectories
0.40
C (r , )d
on distance r from source 2 0 600.00 0.35
0.30
2 500.00
0.25
400.00 0.20
1
0.15
300.00
0 0.10
0 100 200 300 400 500
200.00
0.05
Average dependence on distance r 100.00
along radial trajectory 1 2 rC(r , )d
2 0 100.00 200.00 300.00 400.00 500.00 600.00
Footprint C(q,Q)
where q is source strength
Q background i.e. all the other sources
C(q,Q)
C(q,0.5Q)
Environmental
consequence
2020 2010
Scenario reducing overall emissions
We have not done 2020 yet with CMAQ. Some systematic approximations may suffice.
Conclusions
Messages for regulator
(1) Danger of calculation without understanding
(2) Uncertainty and sensitivity, robust decisions
(3) Informed user or have guidance available
(4) Role of assimilated/empirical models?
(5) Regulatory metric may not be as useful as diagnostics
(6) Justify both domain-wide assessment of policies and individual sources
(7) Some way to go - does not require to much extra effort
(8) Integrated assessment essential for decision