Application of CMAQ with RAMS to air quality

1
CMAS workshop 02/10/23
Application of CMAQ with RAMS to air quality
in the Osaka area in Japan
Seiji Sugata, Shinji Wakamatsu (NIES)
Toshimasa Ohara (Shizuoka Univ.) Itsushi Uno
(Kyushu Univ.) (e-mail
sugatas_at_nies.go.jp)
Two main contents Use of RAMS for CMAQ
Application to the Osaka area in Japan
(preliminary result)
2
Difference of MM5 and RAMS

MM5 RAMS Technically points
Grid system Arakawa-E
Arakawa-C Map projection Lambert
Polar-stereo Vertical coordinate
sigma-p sigma-z Scientific
Hard to say in one sentence
Each strong and weak
points Users Many
Many
Most popular regional met. models License
Free
Necessary
3
A wide choice of meteorological models
for input to CMAQ ? increasing flexibility of
CMAQ application

CMAQ-RAMS interface history - Original
idea by Byun and Uno 1999 Interface for
CMAQ 1999 and RAMS3b developed
by Byun and Sugata 2000 updated for RAMS4.3
(? Sugata) 2002 updating for CMAQ 2001
updating for CMAQ 2002
(MCIP2)
available
preparing
4
Data flow in the interface
postprocessing routine
Main routines of RAMS (version 4.3)
REVU (RAMS Evaluation and Visualization
Utility) Modified dumpout.f
Met. data in GTOOL format
RAMS2IOAPI
Met. data in I/O API format
CMAQ CTM
modified MCIP
5
Air quality observation campaign in Osaka
Aircraft
Tokyo
Surface
Osaka
Osaka Bay
One month for March, 2001 surface 0m,
60m concentrations of SO2, NOx, Ox, PMs,
etc. and high volume
sampling for aerosols aircraft
19,20,21,23 March, 2001
concentrations of SO2, NOx, Ox, PMs, etc.
and high volume sampling for
aerosols
6
Application of CMAQ/RAMS to the OSAKA area
Calculation Period Mar.15-24,2001
(Aircraft measurement Mar. 19, 20, 21, 23)
Resolution 5km mesh (Polarstereo)
21 layers from 25m up to 19km, interval starts
from 50m Domain Greater Osaka area (next
slide) no-nesting RAMS
300km square CMAQ 200km
square Emission 1998 data by the Ministry of
the environment (10km mesh) Models RAMS4.3
assimilation by ECMWF 0.5 degree mesh 4
times per day CMAQ 1999 ver.
7
Domain for CMAQ (200km square)
Domain for RAMS (300km square)
Met. data
Nudging by ECMWF one hour at edges two days
inside
8
Emissions (NO and SO2)
9
CMAQ vs. Obs. (Osaka) ? 62m obs. ? surface
obs. Solid lines CMAQ
NO2
O3
NOx
SO2
10
NO2 (time series and scatter diagrams)
11
NOx (time series and scatter diagrams)
12
NO (time series and scatter diagrams)
13
O3 (time series and scatter diagrams)
14
SO2 (time series and scatter diagrams)
15
Vertical profile of concentration of SO2 along
the flight path (Red CMAQ)
16
Comparison of wind velocity and direction
between RAMS and observation at five
observational sites, Osaka and neighboring
four sites.
17
Wind velocity and direction RAMS vs.
Obs.(Black RAMS Red Observation)
Direction
Velocity
Osaka
Kobe (west of Osaka)
18
Wind velocity and direction (RAMS vs.
Obs.)Black RAMS Red Obs.
Direction
Velocity
Toyonaka (north of Osaka)
Sakai (south of Osaka)
19
Scatter diagram for wind velocity at Osaka
Obs.
RAMS
20
Summary of correlation coefficients
Gas concentrations (Obs.-CMAQ)
NO2 Ox
SO2 0.33
0.51 0.13
(Day/Night) 0.16 / 0.34 0.36 / 0.11
0.16 / 0.05
Met. Field (Obs.-RAMS)
Osaka Kobe Toyonaka
Sakai Ikomayama Velocity 0.61
0.37 0.61
0.53 0.59 (Day/Night) 0.68 / 0.17 0.51
/ 0.14 0.61 / 0.28 0.64 / 0.13 0.61 /
0.44 Direction 0.37 0.33
0.08 0.60
0.49 (Day/Night) 0.33 / 0.38 0.45 / 0.26
0.11 / 0.12 0.50 / 0.54 0.52 / 0.46
? concentrations ltsimilar coefficientsgt directions
21

• Summary
• - Developing MCIP-RAMS interfaces.
• - Simulated air quality in the Osaka area in
  March 2001
• using CMAQ and RAMS
• Compared gas concentrations between CMAQ
• and observations (surface and aircraft).
• Compared wind velocity and direction
• between RAMS and observations.
• Simulations reproduce essentially orders of
  observational
• concentrations
• but still show quantitative disagreement.
• RAMS wind fields
• strong wind for night calm air
• better in velocity and worse in direction

22
Next

• Development of the interface for MCIP2
• Ready for distribution for those who are
  interested in
• Interested in learning MM5
• To get more realistic RAMS results
• One or two more nesting domains
• to consider long-range transport
• Others