Uma Shankar1 and Prakash Bhave2

1
Box Model Tests of Two Mass Transfer Methods for
Volatile Aerosol Species in CMAQ

• Uma Shankar1 and Prakash Bhave2
• Sixth Annual CMAS Conference
• October 1-3, 2007
• 1 UNC Institute for the Environment
• 2 Atmospheric Modeling Division, NOAA (in
  partnership with EPA-NERL)

2
Overview

• Treatment of Coarse PM in CMAQ
• Mass Transfer Theory
• Approach Box Model Development
• Results
• Fine-particle Equilibrium
• Fully Dynamic Approaches (4 schemes)
• Next Steps

3
Coarse-Mode Chemistry in CMAQ

• Prior to CMAQv4.5
• Coarse mode is inert.
• Fine mode species equilibrate instantaneously w/
  inorganic gases

SVOCs
COARSE MODE
2 FINE MODES
4
Coarse-Mode Chemistry in CMAQ

• CMAQv4.6 (current) treatment
• Coarse mode is inert.
• New species shown in RED.
• Fine mode species equilibrate instantaneously w/
  inorganic gases

SVOCs
COARSE MODE
2 FINE MODES
5
Coarse-Mode Chemistry in CMAQ

• Next CMAQ release
• Coarse mode will interact with inorganic
  gases
• New species and interactions are shown in RED

SVOCs
COARSE MODE
2 FINE MODES
6
Time Scale for Mass Transfer
Coarse PM takes 10h to reach equilibrium with
surrounding gases, so instantaneous equilibrium
approach is not applicable.
Dynamic approach needed for gas-particle mass
transfer
7
Mass Transfer Rate, J
Composition-dependent term concentration at the
particles surface (cs) is determined by
gas/particle equilibriumpositive gradient ?
condensationnegative gradient ? evaporation
Most implementations of dynamic mass transfer to
date have been done in sectional models (e.g.,
PMCAMx, CMAQ-MADRID). One exception Modal
Aerosol Module in Polyphemus (Sartelet et al.,
2006).
8
Approach

• Adapt aerosol code from CMAQ v4.6 to develop a
  stand-alone box model for aerosol microphysics
• Extend the box model to treat gas-particle
  transfer with all 3 modes dynamically
• Add some simplifying assumptions to maintain
  computational efficiency
• Resulting module will be implemented in next
  release of CMAQ.

9
Approach

• Adapt aerosol code from CMAQ v4.6 to develop a
  stand-alone box model for aerosol microphysics
• Extend the box model to treat gas-particle
  transfer with all 3 modes dynamically
• Add some simplifying assumptions to maintain
  computational efficiency
• Resulting module will be implemented in next
  release of CMAQ.
• Test case. Mimics the transport of a marine air
  mass into a polluted urban area such as Los
  Angeles

10
Box-Model Test Conditions

• Developed by Pandis et al.
• 38-hour scenario to test different
  gas-to-particle mass transfer schemes over a
  range of RH, particle acidity, and pollution
  concentrations.
• Used previously in development/testing of
  sectional aerosol models in CMAQ-MADRID and
  PM-CAMx

Large plumes of NH3 provide a realistic challenge
for dynamic-transfer module.
Reference Pilinis et al., Aerosol Sci. Technol.,
32482-502 (2000).
11
Box-Model Test Conditions

• Initial conditions
• NH3 0.3 µg m-3
• HNO3 4.0 µg m-3
• Marine particle distribution
• Convert to tri-modal distribution, for
  compatibility with CMAQ

Reference J. Lu and F.M. Bowman, Aerosol Sci.
Technol., 38391-399 (2004).
12
Box-Model Test Results

• First, compare the fine particle equilibrium
  approach of CMAQ v4.6 with a reference model a
  multi-component aerosol dynamics module (MADM)
  run with 10 sections
• Focus of comparisons is total PM concentrations
  of inorganic species predicted by different
  models as a function of time.

13
Box-Model Test Results
Reference curve is from a state-of-the-science
multi-component aerosol dynamics module (MADM)
run with 10 sections. Sulfate matches very well,
because SO42- a non-volatile condensing species.
14
Box-Model Test Results
CMAQv4.6 NH4 also matches reference very
well. Jim Kelly discovered an error in reference
case past hour 30 and thus we excluded these data
from subsequent comparisons.
15
Box-Model Test Results

• In CMAQv4.6, nitrate is underpredicted throughout
  the simulation because
• During first 16 hours, coarse-mode NaNO3 is not
  formed.
• After NH3 is emitted on Hour 16, NH4NO3 formation
  is restricted to the fine modes.

16
Box-Model Test Results

• In CMAQv4.6, Cl- is constant because
• Initial mass of Cl- is entirely in coarse mode
• There is no coarse-mode chemistry
• In reference case
• In first 12 hours, Cl- in coarse PM is gradually
  replaced by NO3-.
• On Hour 16, large NH3 plume leads to NH4Cl
  formation.

17
Box-Model Test Results

• Next, we implemented a dynamic mass transfer
  scheme with a uniform 10 s time step.
• Fluxes of volatile acids and NH3 are calculated
  independently of each other uncoupled
  transfer
• Call ISORROPIA in reverse mode w/ particle-phase
  concentrations as input. Output is the
  equilibrium concentration, Cs, at particle
  surface.
• Focus on Hours 0 16, when marine aerosol is
  reacting gradually with HNO3, before encountering
  large NH3 emissions.
• Does the model capture the replacement of Cl- by
  NO3?

18
Box-Model Test Results
In dynamic model, loss of Cl- from coarse mode is
captured quite accurately!
In dynamic model, NaNO3 reaches the correct
endpoint, but temporal evolution needs further
study.
What happens in dynamic model after Hour 16?
19
Box-Model Test Results
After encountering the NH3 plume on Hour 16,
dynamic model becomes unstable. Abrupt
transition of coarse mode from acidic to
alkaline, causes rapid NH3 evaporation, and the
system never recovers...
So we investigated the use of special mass
transfer schemes when particle composition
approaches neutral pH
20
Treatment Near pH-Neutrality

• 3 approaches in literature (all sectional models)
• Sun Wexler, Atmos. Environ. 1998Coupled
  Transport Transfer acids and bases in
  equimolar quantities such that H remains stable
  near pH-neutrality.
• Pilinis et al., Aerosol Sci. Technol.
  2000Restrain the transfer of all volatile gases
  to allow only small changes in acidity during
  each time step.
• Jacobson, Aerosol Sci. Technol. 2005Uncoupled
  dynamic transfer of acids followed by
  instantaneous equilibrium transfer of NH3.

21
Treatment Near pH-Neutrality

• 3 approaches in literature (all sectional models)
• Sun Wexler, Atmos. Environ. 1998Coupled
  Transport Transfer acids and bases in
  equimolar quantities such that H remains stable
  near pH-neutrality.
• Pilinis et al., Aerosol Sci. Technol.
  2000Restrain the transfer of all volatile gases
  to allow only small changes in acidity during
  each time step.
• Jacobson, Aerosol Sci. Technol. 2005Uncoupled
  dynamic transfer of acids followed by
  instantaneous equilibrium transfer of NH3.
• Implement and test each scheme in box model.

22
Box-Model Test Results
If acids and base are both condensing or both
evaporating, coupled transfer when near
pH-neutral Oscillatory behavior persists but
trend improves substantially.
23
Box-Model Test Results
Jacobson-like scheme Best agreement with
reference case
24
Box-Model Test Results
Jacobson-like scheme Oscillations appear more
pronounced due to scale of the plot. Under-
prediction after hr 16 matches overprediction in
NH4
25
Next Steps

• Implement and test the Pilinis et al. mass
  transfer scheme in our modal model
• Develop a computationally-efficient solution for
  modal model
• Hybrid scheme (fine particles at equilibrium w/
  gas phase, dynamic transfer of coarse particle
  mass)
• Tabulate Cs on coarse mode or treat as an
  irreversible heterogeneous reaction (e.g., Hodzic
  et al., 2006)
• Benchmark our results
• against sectional implementation by Pilinis et
  al.
• against modal implementation by Sartelet et al.
• Compare size-resolved output to multiple
  reference cases
• Apply our fully-dynamic and computationally-effici
  ent schemes in CMAQ simulations
• Incorporate into next years CMAQ release

26
Acknowledgements

• Bill Benjey (EPA-ORD)
• Frank Binkowski (UNC)
• Frank Bowman (UND)
• Adel Hanna (UNC)
• Jim Kelly (EPA-ORD)
• Bonyoung Koo (ENVIRON)
• Spyros Pandis (CMU)
• Christian Seigneur (AER)
• Shaocai Yu (STC)

Disclaimer The research presented here was
performed under the Memorandum of Understanding
between the U.S. Environmental Protection Agency
(EPA) and the U.S. Department of Commerce's
National Oceanic and Atmospheric Administration
(NOAA) and under agreement number DW13921548.
This work constitutes a contribution to the NOAA
Air Quality Program. Although it has been
reviewed by EPA and NOAA and approved for
publication, it does not necessarily reflect
their policies or views.