Modeling and Forecasting Activities in Support of MAXMex Jerome Fast, Pacific Northwest National Lab

1
Modeling and Forecasting Activities in Support
of MAX-Mex Jerome Fast, Pacific Northwest
National Laboratory
2
Why MAX-Mex for Modeling?

• Measurements
• MAX-Mex, MCMA, MIRAGE-Mex, and INTEX-B
  investigators will collect extensive concurrent
  meteorology, chemistry, and particulate data
  using in-situ and remote sensing instrumentation
  deployed at surface sites and on research
  aircraft and satellites
• Particulate loading in Mexico City will likely
  produce strong signals on radiative forcing

polluted day
unpolluted day
3
Why MAX-Mex for Modeling?

• Measurements
• Coordination of research aircraft during a series
  of semi-Lagrangian experiments that will enable
  evaluation of predictions of particulate plume
  evolution - up to several days downwind
• surface (T0,T1, T2) and aircraft (G-1, King-Air)
  measurements during MAX-Mex are critical to
  obtain near-source data

DC-8
C-130
King-Air
G-1
4
Why MAX-Mex for Modeling?

• Models
• Evaluate performance of current state-of-the
  science models by quantifying the uncertainties
  in particulate mass, composition, size
  distribution, and optical properties
• Opportunities for collaborative research, such as
  comparing models with different aerosol process
  modules or comparing local/regional and global
  model predictions of particulate evolution
• Availability of global model output for use as
  boundary conditions of local and regional models
• Use MILAGRO data set as a test-bed to develop and
  evaluate new aerosol process modules

Munoz-Alpizar et al., JGR, 2002
5
Science Questions

• Aerosol Processing
• Does aerosol aging occur mostly over Mexico
  City so that the downwind plume is simply diluted
  by transport and mixing processes?
• Or are aerosol properties evolving significantly
  downwind? What are the processes responsible for
  evolving aerosol properties?
• How does biogenic emissions and other
  anthropogenic sources over central Mexico alter
  the evolution of particulates downwind?
• How do local recirculation processes, that mix
  aged air parcels with fresh emissions, affect
  aerosol properties?
• How can models be improved to better predict
  secondary organic aerosol (SOA) formation
  downwind of Mexico City?

Tie, 5th MM5/WRF Workshop, 2004
6
Science Questions

• Climate
• What is the impact of the Mexico City plume on
  local, regional, and global climate?
• Do global models adequately simulate the
  magnitude and distribution of aerosol radiative
  forcing downwind?
• Can information from Mexico City be used to
  better understand radiative forcing from other
  megacities?

Heating Rates from Aerosols Based on Radiative
Transfer Model
GCM Sulfate Column Burden from Mexico City (March
- May)
Barth and Church, JGR, 1999
Raga et al., Atmos. Environ., 2001
7
Local Meteorology

• We know much about the local meteorology from
  IMADA 1997 and other field campaigns, but
• Downwind regional transport and mixing processes
  largely unknown

IMADA 1997 radar wind profiler and sounding sites
Teotihuacan
Cuautitlan
3700 m
2200 m
3500 m
Chalco
UNAM
4700 m
1300 m
8
Downwind Transport Pathways

• Predict downwind chemical evolution
• Determine transport pathways during months of
  Feb., Mar., and Apr.
• Modeling meeting held at UNAM in Mexico City on
  15 March 2005

Madronich (NCAR), Tie (NCAR), Fast (PNNL), de Foy
(MIT)
chemical evolution as simulated by NCAR master
mechanism over 5 days 13 out of 29 days (15 Feb -
14 Mar 2004) had outflow of CO column towards
NE dispersion for 14 days simulated between 24
Feb - 18 Mar 1997 forward and back trajectories
(Apr -May 2003), NE and SW pathways radar wind
profiler or HYSPLIT (Feb - Mar 2002, 2003), large
fraction towards NE
9
Effects of Local Emissions

• How do power plant and volcanic emissions effect
  particulate evolution?
• How frequently do these emissions mix with fresh
  pollutants over the city and with aged pollutants
  further downwind?

Simulated SO2 11 LST 21 March 1997
Anthropogenic Emissions
Tula Power Plant
Popocatepetl
Mexico City
lt 3 km MSL layer
3-6 km MSL layer
Simulated SO2 19 LST 21 March 1997
3-6 km MSL layer
lt 3 km MSL layer
10
Effects of Biomass Burning

• How does the mixing of biomass burning plumes and
  the Mexico City plume affect downwind particulate
  evolution?
• What is the relative contribution of biomass
  burning and Mexico City emissions on regional
  radiative forcing?

MODIS Satellite Image 2 March 2004
marine stratus common
biomass burning sites (red)
11
Effects of Clouds

• How do clouds affect aerosol properties over
  central Mexico and do does the Mexico City plume
  significantly alter cloud evolution?
• What is the fraction of boundary layer
  particulates vented into the free atmosphere
  through clouds?

MODIS Satellite Image 31 March 2005
Mexico City
clouds often form over the mountains surrounding
Mexico City during the dry season
12
MILAGRO Modeling Group
MILAGRO and INTEX-B joint planning Meeting, Oct
24-26, Boulder, CO Co-chairs of Forecasting and
Modeling Working Group Jerome Fast (PNNL) and
Ernesto Caetano (UNAM)

• Ernesto Caetano, Victor Magana, UNAM, MM5
• Benjamin de Foy University of California - San
  Diego, MM5FLEXPARTCAMx
• Bill Skamarock NCAR / MMM, WRF
• Xuexi Tie NCAR / ACD, WRF-chem
• Andreas Stohl Norwegian Institute for Air
  Research, FLEXPART
• Greg Carmichael, Youhua Tang, Marcelo Mena
  University of Iowa, STEM
• Peter Hess, Louisa Emmons NCAR / ACD, MOZART
• Daniel Jacob Harvard University, GEOS-chem
• Brad Pierce NASA Langley, RAQMS
• Other investigators that may contribute
• Georg Grell NOAA, operational version WRF-chem
• Alma Hodzic, NCAR / ACD, CHIMERE (using MM5-UCSD
  as input)

13
Forecasting Models

• Likely Models
• MM5 Local - Regional, meteorology, Dx 24, 8 km
• MM5CAMx Local - Regional, meteorology,
  chemistry, trajectories, dispersion, Dx 36,
  12, 3 km
• WRF Local - Regional, meteorology and tracers,
  Dx 9, 3 km
• WRF-chem Local - Regional, meteorology,
  chemistry, Dx 6 km
• STEM Regional - Continental, meteorology,
  tracers, chemistry, particulates, Dx ? km
• FLEXPART Global, passive scalar using GFS
  forecasts Dx 1 degree
• MOZART Global, chemistry, Dx 2 degree
• GEOS-Chem Global, chemistry, Dx 50 km
• RAQMS Global, chemistry, ozone assimilation, Dx
  2 x 2.5 degrees
• Possible models
• WRF-chem Continental, meteorology, chemistry,
  particulates, Dx 40 km
• CHIMERE Local - Regional, chemistry and
  particulates

scale
14
Forecasting Operations

• Veracruz Operations
• support G-1, C-130, King-Air, J-31
• use MILAGRO forecast models and other operational
  models to produce forecast of weather and
  location of Mexico City plume
• Houston Operations
• support DC-8
• Mexico City
• support surface measurements
• Video conferencing
• daily weather briefing 10-11 AM
• coordinate aircraft for next day
• forecasts focus on next day, but also produce 3-5
  day forecasts

15
Pre-Field Campaign Modeling

• PNNLs version of WRF-chem community model, Dx
  18, 6, and 2 km
• Chemistry / Particulate Simulations
• series of 2-day case study simulations during
  1997 IMADA field campaign to examine evolution of
  trace gases, particulates, and radiative forcing
  over central Mexico
• Tracer Simulations
• Series of 2-day simulations during 1997 IMADA
  field campaign period to examine pollutant
  pathways and model sensitivity to PBL
  formulations
• Series of month-long simulations to examine
  transport pathways and multi-day accumulation of
  pollutants over central Mexico

16
Aircraft Scenarios 1 March 1997
good day for T0-T1-T2 Lagrangian scenario, 11 LT
surface ozone 11 LT
7 6 5 4 3 2 1
D C B A
T2
D
T1
7 6 5 4 3 2 1
T0
C
7 6 5 4 3 2 1
ozone 1600 m AGL 11 LT
King-Air Overflights
B
G-1 Lagrangian
7 6 5 4 3 2 1
G-1 Source Characterization
A
west
east
17
Aircraft Scenarios 1 March 1997
good day for T0-T1-T2 Lagrangian scenario, 16 LT
surface ozone 16 LT
7 6 5 4 3 2 1
D C B A
T2
D
T1
7 6 5 4 3 2 1
T0
C
7 6 5 4 3 2 1
ozone 1600 m AGL 16 LT
B
7 6 5 4 3 2 1
A
18
Aircraft Scenarios 13 March 1997
Not a good day for T0-T1-T2 Lagrangian scenario,
11 LT
surface ozone 11 LT
7 6 5 4 3 2 1
D C B A
T2
D
T1
7 6 5 4 3 2 1
T0
C
7 6 5 4 3 2 1
ozone 1600 m AGL 11 LT
B
7 6 5 4 3 2 1
A
west
east
19
Aircraft Scenarios 13 March 1997
Not a good day for T0-T1-T2 Lagrangian scenario,
16 LT
surface ozone 16 LT
7 6 5 4 3 2 1
D C B A
T2
D
T1
7 6 5 4 3 2 1
T0
C
7 6 5 4 3 2 1
ozone 1600 m AGL 16 LT
B
7 6 5 4 3 2 1
A
west
east
20
Lagrangian Aircraft Flights
26th
synoptic conditions during March can be favorable
for long-range transport of Mexico City plume to
the NE over several consecutive days balloon
trajectory will aid Lagrangian aircraft sampling
strategy
26th
27th
12 LT 27 February
12 LT 28 February
28th
27th
28th
1st
12 LT 1 March
12 LT 2 March
21
Local Aerosol Processing

• Testing Conceptual Models
• Then
• Model predictions employed to develop a
  conceptual model of diurnal evolution of
  pollutants in the valley
• Same-day recirculation of pollutants possible
• Little multi-day accumulation of pollutants
• Now
• 2006 campaigns will provide data to test these
  hypotheses
• Same-day recirculation will transport aged air
  parcels back over the city

22
Summary

• MAX-Mex collaborative activities with the other
  MILAGRO scientists is a great opportunity for ASP
  to improve our understanding of aerosol evolution
  (particularly organic and elemental carbon) and
  radiative forcing associated with anthropogenic
  emissions
• MILAGRO data set will be useful in quantifying
  uncertainties in the predictions of particulate
  mass, composition, size distribution, and aerosol
  optical properties from current 3-D models and
  developing improved aerosol process modules