Introduction to

1
Introduction to The REMSAD Modeling System
Presented By Gerard Mansell ENVIRON
International Corporation
January 2002
2
What is REMSAD?
Regulatory Modeling System for Aerosols and
Deposition

• Designed to be
• Fast screening tool for control strategies for PM
• Continental-scale tool for deposition of toxics
  and acidic precipitation

3
REMSAD as a PM Model

• Applicable over regional scales
• Includes both Primary and Secondary particulates
• Responds to inventory-level control measures
• Detailed representation of spatial and temporal
  distributions
• Treats meteorological influences on transport and
  removal directly

4
Technical Overview

• Based on UAM-V
• Recast in sigma coordinate
• Extended vertically to tropopause
• Less detailed photochemistry (?CB-IV)
• Processes Treated
• Transport
• Deposition (wet and dry)
• Chemistry (gaseous, aqueous, aerosols)
• Inputs
• Emissions
• Meteorology
• Airquality
• Chemical rates
• Landuse

5
Technical Overview
(continued)

• Outputs
• Ammonium Sulfate Particulates
• Acidic Sulfate Deposition Flux
• Ammonium Nitrate Particulates
• Nitric Acid Deposition Flux
• Secondary Organic Aerosol
• Primary Coarse Particulates
• Primary Fine Particulates
• Total Fine Particulates
• Visibility Measures (Extinction, Deciview, Visual
  Range)

6
Advantages and Limitations

• Existing databases
• Simplified treatment allows cost effective
  solutions
• Key features components consistent with
  state-of-the-science
• Previous results peer review demonstrated
  capabilities of producing credible results

7
Advantages and Limitations
(continued)

• Simplified O3 chemistry may not be adequate to
  simulate O3 concentrations or interactions of O3
  with other pollutants (PM)
• No SOA treatment

8
Processes

• Transport
• Advection
• Smolarkiewicz on Arakawa-C grid
• Other schemes -- Petrov-Galerkin,
  Taylor-Galerkin, Van Leer
• Turbulent Diffusion
• K-Theory or First-order Closure
• Kx, Ky, Kz - Eddy Diffusivity Coefficients
• Horizontal -- Deformation of Horiz. Wwinds
  (Smagorinsky)
• Vertical -- Diagnosed form wind temp (McNider
  Pielke) or interpolated from MM5

9
Processes
(continued)

• Transport
• Cloud Dynamics
• Diagnosed as in MM5 (Anthes et al., 1987)
• Deep Convective Clouds
• Vertical redistribution due to deep convection
• Calculated only if rainfall
• Within Cloud -- tendency for species proportional
  to cloud cover and vertical gradients within and
  below
• Mass Conservation/Continuity -- tendencies within
  cloud balanced with subcloud layer, vertical
  exchange between cloud layer and subcloud layers

10
Processes
(continued)

• Stratiform Clouds
• Vertical redistribution performed if maximum
  cloud coverage gt 10 and
• Cloud depth gt 0.1s and
• Bottom of cloud layer lt 3000 m AGL
• Explicit K-theory (Lin et al., 1994)
• Shallow Cumulus Clouds (not considered)

11
Processes
(continued)

• Deposition (dry)
• Dry Deposition based on RADM Algorithm
  (Wesley,1989)
• Flux to ground F - C Vd
• Deposition velocity (gases)
• Ra Aerodynamic resistance (PBL similarity model
  Louis, 1979)
• Rb Boundary (quasi-laminar) layer resistance
• Rs Surface resistance (dependent on leaf
  stomata and cuticles, lower canopy resistance,
  soil litter, and water)
• Deposition velocity (particles)

12
Processes
(continued)

• Gaseous Deposition (wet)
• Parameterized based on
• Henry Laws solubility coefficients, rainfall
  rates, cloud depth and temperature
• Hales Sutter (1979) original work and developed
  ionization constants for SO2
• ATDM extended to (NO, NO2, SO2, NH3, VOC, HNO3)
  and 7 toxic species
• New solubility constants taken from literature,
  now temperature dependent
• Drop size and velocity parameterized based on
  rainfall rate - Scott (1978)
• Each model layer successively adjusted for
  deposition
• Typical cloud pH of 4.5 has been used for
  solubility calculations

13
Processes
(continued)

• Particle Deposition (wet)
• Based on relationships developed by Scott (1978)
• Rainfall rate and cloud type related to fraction
  of sulfate within rainwater reaching ground
• Extended to treat any aerosol -- function of
  sulfate rate
• Key assumptions
• Aerosols lt 1 ?m in dia - Nucleation of cloud
  droplets around aerosols
• Aerosols gt 1 ?m in dia - removed only through
  impaction with falling drops
• Within-Cloud washout rate
• Time in which active hydrometeor growth occurs
• cloud depth, temperature, rainfall rate, and
  cloud type
• Below cloud base to ground washout rate (liquid
  only)
• function of rainfall rate

14
Processes
(continued)

• Gas Phase Chemistry
• Parameterized
• Uses OH lookup tables OHf(O3,NOx,sunlight)
• O3 input
• Treats NOx, SO2, NH3, VOC, HNO3

15
Processes
(continued)

• Gas Phase Chemistry
• ?CB-IV
• Reduced organic speciation
• Inorganics and radicals same as CB-IV
• 3 organics species VOC, ISOP, CARB
• Responds to NOx and VOC controls
• Approximates full CB-IV
• Predicts precursors to secondary particulates

16
Processes
(concluded)

• Aqueous Phase Chemistry
• In-cloud sulfate formation
• Reactions of SO2 w/ H2O2, O2 and O3
• Uses representation of Martin (1994) to estimate
  rate of sulfate formation
• Assumed pH of 4.5
• Parameterized chemistry version uses algorithms
  from RTM-II

17
Processes
(concluded)

• Aerosol Chemistry
• Nitrate from equilibrium between nitrate,
  sulfate, and ammonia
• MARS-A equilibrium routines from CMAQ
• Earlier versions use parameterized chemistry to
  estimate equilibrium

18
REMSAD System
19
Flow of Major ATDM Processes
20
Nesting GridExamples
21
Vertical Grid Structurein Typical
REMSADApplication
?

?????????
0.0
pt
0.1
0.3
0.5
0.7
0.78
0.84
0.93
0.98
1.00
22
Examples ofVertical Nesting
23
Model Inputs

• Airquality (Initial Boundary concentrations)
• ppm for gases µg/m3 for particulates)
• CAMx or UAM preprocessor
• Winds (u v components m/s2)
• supplied for coarse grid optionally for nested
  grids
• w component calculated internally
• Temperature ( K )
• three dimensional temps at cell centers
• Surface air temperature
• supplied for coarse grid only

24
Model Inputs
(continued)

• Humidity (water vapor mixing ratio kg/kg)
• supplied for coarse grid only
• used for cloud diagnosis sulfate chemistry
  deposition
• Cloud Liquid Water (mixing ratio kg/kg)
• supplied for coarse grid only
• used to compute reaction rates for aqueous-phase
  chemistry wet deposition rates
• Rain Liquid Water (mixing ratio kg/kg)
• supplied for coarse grid only
• used to compute reaction rates for aqueous-phase
  chemistry

25
Model Inputs
(continued)

• Surface pressure (mbar)
• supplied for coarse grid optionally for nested
  grids
• defines vertical heights and 3D pressure
• used to calculate reaction rates, photolysis
  rates, deposition, plumerise
• Vertical Turbulent Exchange Coefficients (m2/s)
• supplied for coarse grid optionally for nested
  grids
• determines rate of vertical diffusion
• non-zero minima (0.1-1 m2/s)
• Precipitation (rainfall rate in/hr)
• supplied for coarse grid only
• used in cloud diagnosis, wet deposition

26
Model Inputs
(continued)

• Surface characteristics
• supplied for coarse grid optionally for nested
  grids
• defines landuse fraction for each of 11 land use
  categories
• used to calculate deposition rates
• Terrain (m above sea level)
• supplied for coarse grid only
• surface elevation
• Not needed (use dummy values of zero)
• CHEMPARAM
• Defines species to be simulated and lower bounds
• particle size and mass distributions
• varies by model version

27
Model Inputs
(concluded)

• Photolysis Rates
• tabulated rates for five key reactions
• photolysis of NO2, HCHO, O3 and ALD2
• function of height zenith angle under constant
  albedo (0.08), ozone column (0.318) and haze
  (0.2)
• Emission inventory
• Gridded low-level area sources
• supplied for coarse grid and all nested grids
• Elevated point sources
• NO, NO2, POA, PEC, GSO4, PNO3, VOC, SOA, SO2, CO,
  NH3, CARB, ISOP, PMFINE, PMCOARS
• EPS2 or SMOKE processor

28
Model Outputs

• Average files
• average layer 1 concentrations
• µg/m3 for all species
• user specified species
• Instant files
• three dimensional instantaneous concentrations
• ppm for all species
• includes all species
• Deposition files
• wet and dry surface depositon
• g/m3 for all species
• user specified species
• Diagnostic outputs

29
REMSAD PostprocessingSystem Flowchart
30
Visibility Measures
bext (?) bscat (?) babs (?)
bscat babs (usually)

• Extinction
• Deciview
• Visual Range

DV 10 ln (bext/10), where Bext is in
Mm-1
- ln (0.02) bext
3.912 bext
vr

31
WRAP REMSAD Modeling Domain
32
Computer Resource Requirements

• WRAP modeling domain (120 x 84 x 12)
• Disk storage (per simulation day)
• Inputs
• Emissions 45 Mb
• Meteorology 88 Mb
• Outputs
• Average 26 Mb
• Deposition 52 Mb
• Instant 315 Mb
• CPU (per simulation day)
• 50 min. on Linux box w/ 1.2 GHz AMD Athelon 1
  Gb memory

33
Computer Resource Requirements
(Continued)
34
Computer Resource Requirements
(Concluded)
35
(No Transcript)
36
(No Transcript)
37
(No Transcript)
38
(No Transcript)