United Slates
          Environmental Protection
          Agency
             ()fficc lit Air Quality
             Planmnc and Standards
             Research Tnancle Park. NC 27711
1-PA-454/R-94-015
April 1994
          Air
& EPA
DEVELOPMENT AND TESTING
OF A DRY DEPOSITION
ALGORITHM (Revised)

-------
                                EPA-454/R-94-015
--O
ri
-4
O
      Development  and Testing
       of a  Dry Deposition
       Algorithm (Revised)
           U.S. Environmental Protection Agency

          Office of Air Quality Planning and Standards

               Technical Support Division

             Research Triangle Park, NC  27711


                             U.S. tf
-------
This report has been reviewed by the Office of Air Quality
Planning and Standards, U.S. Environmental Protection Agency, and
has been approved for publication.  Any mention of trade names
and commercial products is not intended to constitute endorsement
or recommendation for use.
                        EPA-454/R-94-015

-------
                        ACKNOWLEDGEMENTS
     This report has been prepared by Sigma Research Corporation
and funded by the U.S. Environmental Protection Agency under
Contract No. 68-D90067, with Jawad S. Touma as Work Assignment
Manager.

-------
                             PREFACE

     The ability to accurately estimate deposition due to
atmospheric releases is important to the modeling community.
Limitations of the Industrial Source Complex (ISC2) models  (dated
92273) for estimating deposition of small particles have been
known for some time.  An improved algorithm for estimating
deposition for a wide range of pollutants has been developed and
tested and is described in this report.

     The Environmental Protection Agency must conduct a formal
and public review before the Agency can recommend for routine use
this new algorithm in regulatory analyses.  This report is being
released to establish a basis for reviews of the capabilities of
this methodology in routine dispersion modeling analyses.  This
report is one part of a larger set of information on the ISC2
models that must be considered before any formal changes can be
adopted.

     The report "Development and Testing of a Dry Deposition
Algorithm", EPA-454/R-92-017, was issued in May 1993.  EPA
discovered that a typographical error in a journal paper used as
a basis for one of the models (ADOM 2) had propagated to the
report.  The corrected equation has been re-analyzed.  In
addition, a revised form of the Stokes number equation, based on
Slinn (1982), was used in the re-analysis.  Because the
performance of the corrected model was significantly different
than that originally reported,  this new report is being issued as
an entire replacement document.

-------
                                    Table of Contents


 Section                                                                           Page

 1.     Introduction	:	  1-1

 2.     Dry Deposition  	  2-1
       2.1    Dry Deposition of Particulate Matter	2-5
              2.1.1         ISC Method  	  2-7
              2.1.2         GARB Model	  2-13
              2.1.3         ADOM Model - Particles	  2-21
              2.1.4         UAM-V Model	  2-26
       22    Dry Deposition of Gaseous Pollutants	  2-26
              22.1         RADM Model	  2-27
              222         ADOM Model - Gases	  2-28
              223         NOAA/ARL Models  	  2-31

 3.     Plume Depletion Techniques	  3-1
       3.1 Source Depletion  	  3-1
       3.2 Surface Depletion	  3-2
       33 K-theory Approach	  3-2
       3.4 Modified Source Depletion	  3-4

 4.     Calculation of Meteorological Variables  	  4-1
       4.1    Unstable/Neutral Conditions	  4-1
       42    Stable Conditions	  4-5

5.     Model Evaluation Protocol  	  5-1
       5.1    Evaluation Approach	  5-1
       52    Stratification of Deposition Velocity Datasets  		  5-2
       53     Comparison of Predictions and Observations	  5-5
       5,4    Scoring Model Performance Using.Composite Measures.		5-7

6.     Model Evaluation Data Bases 	  6-1
       6.1    Particle Data Sets	  6-1
       62     Gas Data Sets  	  6-2

-------
                               Table of Contents - Continued
7.     Results of Model Evaluation	7-1
       7.1    Particle Deposition Models  	7-1
             7.1.1   Full Data Sets	7-4
             7.12   Stratification by Particle Diameter	7-7
             7.13   Stratification by Roughness Length	  7-14
             7.1.4   Stratification by Leaf Area Index	  7-17
             7.1.5   Stratification by Day vs Night	  7-19
             7.1.6   Stratification by Friction Velocity	  7-20
             7.1.7   Stratification by Temperature	  7-22
             7.1.8   Estimation of CPM from Tables	  7-25
       72    Discussion of Model Performance 	  7-25
             7.2.1   Uniform Size Distribution  	  7-25
             722   Sulfate Particle Distribution	  7-29
             722   Model Performance When Zeros are Included	  7-29
             7.2.4   Cumulative Distribution Results	  7-3-4
             72 J   Selection of Best Performing Deposition Model	•....  7-34

8.     Summary and Conclusions	8-1

9.     References	.'	9-1

Appendix A  Estimation of ISC Deposition Velocity	A-l
Appendix B  Supplemental Graphics	B-l
Appendix C  Observational Particle Deposition  Velocity Data Sets 	C-l
Appendix D  Predicted Deposition Velocities vs Particle Diameter	D-l
Appendix E  Implementation of the Modified Source Depletion Method in ISC2  	E-1
                                            VI

-------
                                        List of Figures
 Figure 2-1.    Summary of observed SO2 deposition velocities	2-3

 Figure 2-2.    Observed deposition velocities (vj as a function of particle size for  	2-4
               U g/on3 density particles.

 Figure 2-3.    Predicted deposition velocities for u. » 100 cm/s and particle   	2-6
               densities of 1, 4, and 11.5 g/cm3.

 Figure 2-4.    Illustration of the reflection coefficient scheme used in ISC  	2-3
               for reflection coefficients of 0.0, 0.5, and 1.0.

 Figure 2-5.    Reflection coefficient as a function of the gravitational settling	 2-12
               velocity.

 Figure 2-6.    Deposition velocity as a function of particle diameter as predicted  	 2-18
               by the Sehmel/CARB model for three different values of ambient
               temperature (0°, 60°, 100° F).

 Figure 2-7.    Deposition velocity as a function of particle diameter as predicted  	 2-19
               by the Sehmel/CARB model for surface roughness lengths of 0.001,
               0.01, 3, and 10 cm.

 Figure 5-1.    Schematic illustration of division of deposition velocity data into subsets.  ... 5-4

 Figure 7-1.    A summary of absolute value of the fractional bias using all data 	7-S

 Figure 7-2.    A summary of absolute value of the fractional bias averaged over all  	7-9
               12 subsets.

 Figure 7-3a.    Co-plot of fractional bias of the standard deviation for each of	  7-10
               the deposition models.

Figure 7-3b.    Co-plot of fractional bias of the 11 largest deposition velocities   	  7-11
              for each of the deposition models.
                                             vu

-------
                                 List of Figures - Continued

Figure 7-3c.   Co-plot of fractional bias of the 17 smallest deposition velocities  	  7-12
              for each of the deposition models.

Figure 7-3d.   Co-plot of fractional bias of the 11 largest and 17 smallest  	  7-13
              deposition velocities for each of the deposition models.

Figure 7-4.    A summary of the total CFB of the six types of data subsets	  7-24

Figure 7-5.    A ranking of the models by CPM	  7-27

Figure 7-6.    A summary of the MCM for each unique model pair	  7-28

Figure 7-7.    A ranking of the models by CPM	  7-31

Figure 7-8.    A summary of the MCM for each unique mode pair	  7-32

Figure 7-9.    A ranking of the models by CPM	  7-33

Figure B-la.  Scatter plot of observed deposition velocity (cm/s) versus model  	B-l
              predicted deposition velocity (cm/s) for the complete small
              particle data  set.

Figure B-lb.  Scatter plot of observed deposition velocity (cm/s) versus model  	B-2
              predicted deposition velocity (cm/s) for the complete small
              particle data  set.

Figure B-lc  Scatter plot of observed deposition velocity (cm/s) versus model  	B-3
              predicted deposition velocity (cm/s) for the complete small
              particle data  set.

Figure B-ld.  Scatter plot of observed deposition velocity (cm/s) versus model  	B-4
              predicted deposition velocity (cm/s) for the complete small
              particle data  set.
                                            vui

-------
                                 List of Figures - Continued
Figure B-le.  Scatter plot of observed deposition velocity (cm/s) versus model   ..-..'	B-5
              predicted deposition velocity (cm/s) for the complete small
              particle data set

Figure B-lf.   Scatter plot of observed deposition velocity (cm/s) versus model   	B-6
              predicted deposition velocity (cm/s) for the complete small
              particle data set

Figure B-lg.  Scatter plot of observed deposition velocity (cm/s) versus model   	B-7
              predicted deposition velocity (cm/s) for the complete small
              particle data set

Figure B-lh.  Scatter plot of observed deposition velocity (cm/s) versus model   	B-8
              predicted deposition velocity (cm/s) for the complete small
              particle data set

Figure B-1L   Scatter plot of observed deposition velocity (cm/s) versus model   	B-9
              predicted deposition velocity (cm/s) for the complete small
              particle data set.

Figure B-lj.   Scatter plot of observed deposition velocity (cm/s) versus model   	B-10
              predicted deposition velocity (cm/s) for the complete .small
              particle data set

Figure B-2a.  Cumulative probability plot of deposition velocity (cm/s) using   	B-ll
              the complete small particle data set.

Figure B-2b.  Cumulative probability plot of deposition velocity (cm/s) using   	B-12
              the complete small particle data set

Figure B-2c,  Cumulative probability plot of deposition velocity (cm/s) using   ...	B-13
              the complete small particle data set

Figure B-2d.  Cumulative probability plot of deposition velocity (cm/s) using   	B-14
              the complete small particle data set.
                                            IX

-------
                                List of Figures - Concluded
Figure D-1.   Predicted deposition velocity for the CARfi-based models for   ........... D-1
              u. » 10 cm/s, 2^, » 10 cm, LAI * 1.0, p • » LO g/cm3, and neutral stability.

Figure D-2,   Predicted deposition velocity for the ADOM-based models for  ........... D-2
              u. = 10 cm/s, za = 10 cm, LAI = LO, p = 1.0 g/cm3, and neutral stability.

Figure D-3.   Predicted deposition velocity for the UAM-based models for  ............ D-3
              u. » 10 cm/s, za = 10 cm, LAI = 1.0, p = 1.0 g/cm3, and neutral stability.

Figure E-l.   Illustration of the depletion factor FQ and the corresponding profile ....... E-2
              correction factor
Figure E-2.    Vertical profile of concentration before and after applying FQ and ........ E-3
                    shown in Figure E-l.

-------
                                     List of Tables
 Table 1-1

 Table 2-1

 Table 2-2

 Table 2-3


 Table 2-4
 Core Particle Deposition Models and Other Algorithms Included in Study  ..  1-3

 Factors Influencing Dry Deposition Rates	2-2

 Typical Surface Roughness Lengths for Various Land Use Types  	  2-17
 Values of Chemical Input Parameters Required by the RADM
 Deposition Model
 Summary of Input Requirements of RADM Deposition Module for
 Gases
    2-29
   2-30
Table 2-5     Leaf Area Index Values as a Function of Land Use Type and Season	  2-32

Table 2-6     Summary of Input Requirements of ADOM-type Deposition Modules ....  2-33
              for Gases
Table 4-1

Table 4-2


Table 7-1

Table 7-2


Table 7-3


Table 7-4-


Table 7-5
Values of Net Radiation Constants  	 4-3

Minimum Values of Monin-Obukhov Length During Stable Conditions for .  . 4-7
Various Land Use Types
A Summary of the Model Designations
A Summary of the Stratifications Made to the Small Particle Data Set  . .
(N = 168)

A Summary of the Fractional and Composite Statistical Measures for  ..
Each of the 10 Models Examined

A Summary of the Fractional and Composite Statistical Measures, for  ..
Each of the Models Examined
... 7-2

. .. 7-3


. . . 7-6


... 7-15
A Summary of the Fractional and Composite Statistical Measures for  ....  7-16
Each of the Models Examined
                                          XI

-------
                                List of Tables - Concluded
Table 7-6      A Summary of the Fractional and Normalized Statistical Measures for
              Each of the Models Examined
.  7-18
Table 7-7      A Summary of the Fractional and Normalized Statistical Measures for
              Each of the Models Examined
.  7-20
Table 7-8      A Summary of the Fractional and Normalized Statistical Measures for
              Each of the Models Examined
.  7-21
Table 7-9     A Summary of the Fractional and Normalized Statistical Measures for  .
             Each of the Models Examined
.  7-23
Table 7-10     Summary of Composite Statistical Measures that Illustrate how the CPM
              Arises for the CARE 1 Model
.  7-26
Table 7-11    Aerosol Mass Fraction as a Function of Size Distribution for Two
             Assumed Aged Sulfate Distributions
  7-30
Table C-1     Observational Particle Deposition Velocity Data Sets  	C-1
                                           xu

-------
1. Introduction
                                 f
       The intermedia transfer Of pollutants from the atmosphere to land, water, and vegetation
is an increasingly important concern in many regulatory environmental impact analyses. The
process of dry deposition is a critical transport route for pollutant movement across the
air/surface interface.  However, current regulatory modeling tools have Limitations in their
ability to evaluate dry deposition.  For example, the Industrial Source Complex (ISC) model
contains an empirical dry deposition algorithm based on reflection coefficients which is
appropriate for large particles dominated by gravitational settling (i.e., particle diameters larger
than approximately 20 jim). However, this algorithm is not designed for small particles or
gaseous pollutants, both of which are of concern for many analyses involving criteria pollutants
and toxic air pollutants.  Recognizing the need for a generalized, scientifically-credible dry
deposition algorithm capable of estimating deposition fluxes for a wide range of pollutants, the
U.S. Environmental Protection Agency (EPA) has sponsored a study to develop such an
algorithm for use in the ISC model The primary objectives of the present study are to  identify
deposition models suitable for regulatory use,  evaluate and intercompare several of the
techniques, and implement the most appropriate approach into the ISC model.

       The dry deposition flux (F) of a pollutant at a reference height above the surface can be
defined as the product of the ambient concentration of the pollutant, (x), and a "deposition
velocity" (vj, Le., F = x vd- Therefore, a model to predict deposition fluxes requires an
estimation of both the ambient concentration and deposition velocity. The EPA has established
procedures for estimating ambient pollutant concentrations, which can be used for deposition
modeling as well. -However, an approved procedure for calculating deposition velocities is not
established. The focus on this study is  the testing and evaluation of algorithms to compute
deposition velocities for particulate matter.

   The following criteria are being applied in the selection of the dry deposition velocity
algorithms for evaluation.

       (1)    The algorithm should parameterize the important physical/chemical processes
              known to determine the rate of  deposition.

       (2)    The scheme should  require only routinely and/or readily available
              meteorological," chemical, and physical input data.

       (3)    The algorithm should have modest computational requirements.
                                                                             m

                                            I-1

-------
       (4)     The formulation should be general enough to accommodate, with the proper
              input parameters, a wide variety of pollutants.

   Although the number of models to be tested must be limited, hybrid approaches, consisting
of the best components of two or more models, have been constructed for some of the particle
deposition velocity models.  Various components of these models have been modified or
enhanced in order to provide increased generality or to eliminate weaknesses in the approach.
Table 1-1 lists the base (core) particle deposition models which have been evaluated in this
study along with the gas deposition, plume depletion, and meteorological processing algorithms
which have been reviewed.

       Because pollutants that are deposited are removed from the plume, there is some
feedback between upwind plume depletion and the ambient concentration at a particular
receptor. Therefore, a second component of deposition which is reviewed in this study is the
method used to track and deplete a pollutant from the plume as it travels downwind and
deposits onto the surface. Among the methods which have been proposed to incorporate plume
depletion into the Gaussian plume framework are source  depletion, surface depletion, modified
source depletion, and K-theory methods. Based on a review of the literature, the advantages
and disadvantages of these techniques are discussed in Section 3.

       Most dry deposition algorithms require as input certain micrometeorological parameters
such as the surface friction velocity and Monin-Obukhov length. In order to provide a complete
system for modeling deposition, a revised ISC modeling system must include a technique for
estimating these parameters. A review of some previous studies which have tested various
methods for the computation of friction velocity and Monin-Obukhov length from routinely
available meteorological observations is discussed in Section 4.

   The model evaluation protocol used in the present study, described in Section 5, is based on
the Cox-Tikvart protocol. It represents EPA's recommended approach to the evaluation of
model performance.  Section 6 contains a summary of the observational particle deposition data
bases used in the model evaluation efforts (described in Section 7.1) as well as some gas
deposition data for possible use in future evaluation work. A summary of the recommendations
and conclusions reached to-date is contained in Section 8.
                                           1-2

-------
                                         Table 1-1
        Core Particle Deposition Models and Other Algorithms Included in the Study
Type
Model
 References
Deposition velocity models
(particles)
 CARB/Sehmel model

 ADOM/CALPUFF/CALGRID
 models

 UAM-V model
 ISC reflection coefficient scheme
 Sehmel and Hodgson (1978)
 Sehmel (1980)
 Pleim et aL (1984)
 Sore et aL (1990)
 Yamartino et aL (1992)
 Gray et aL (1991)
 Dumbauld et aL (1976)
 Overcamp (1976)
 Bowers et aL (1979)
Deposition velocity models
(gases)
RADM model
                               ADOM/CALPUFF/CALGRID
                               models
                               NOAA/ARL models
Wesely (1989)
Walcek et aL (1986)
Shieh et aL  (1986)
Pleim et al. (1984)
Sore et aL (1990)
Yamartino et aL (1992)
Meyers and Baldocchi (1988)
Baldocchi et al. (1987)
Meyers (1987)
Hicks et al.  (1987)
Plume depletion techniques
Source depletion
Surface depletion
K-theory approach
Modified source depletion
Chamberlain (1953)
Horst (1977)
Rao (1981)
Horst (1983)
Meteorological preprocessor
Holtslag and van Ulden
technique
Holtslag and van Ulden (1983)
van Ulden and Holtslag (1985)
Hanna and Chang (1992)  -
                                           1-3

-------
2.     Dry Deposition

       Many complex processes are involved in the transfer and deposition of pollutants at the
surface.  Sehmel (1980) compiled a list of the factors known to influence dry deposition rates
(see Table 2-1).  Although it is not practical to incorporate all of the factors listed in the table
into a routinely-applied regulatory deposition model,  algorithms with only modest data
requirements exist which parameterize many of the processes which typically dominate
deposition rates. Among the most important factors  included in Table 2-2 are pollutant
properties such as the size and density of particles, and the solubility, reactivity, and diffusivity
of gases; surface characteristics such as the surface roughness, vegetation type, amount, and
physiological state; and atmospheric variables including-the stability, friction velocity, and
turbulence intensity.  Hicks (1982) noted the important differences controlling the deposition of
larger particles (e.g., gravitational settling, inertia! impaction) and those controlling gases (e.g.,
turbulence, molecular diffusion).  Deposition of small particles is complicated by the fact that
because of their intermediate size, they may be influenced by the processes affecting gases and
large particles.

       Due to the number and variability of the factors influencing dry deposition rates,
measured values of deposition velocities exhibit considerable variability.  For example, SO2
deposition velocity observations summarized by Sehmel (1980) range over two orders of
magnitude (Figure 2-1). Particle deposition velocities (Slinn et al., 1978) show an even greater
variability. This is illustrated in Figure 2-2, which shows the deposition velocity  as a function of
particle size as measured in two wind tunnel experiments. The deposition velocity ranges from
over 10 cm/s (for > 10 jim diameter particles) to less than 0.01 cm/s (0.25 jim  diameter
particle). Although it is not practical to include in the deposition model ail of the variables listed
in Table 2-1, it is possible to parameterize many of the most important effects known to control
deposition rates in terms of routinely-available variables which describe the state of the
atmosphere, surface conditions, and pollutant properties.

    The treatment of particle resuspension is not being treated in the current study. The
problem of particle resuspension is very complex, involving highly variable and site-specific
factors. Sehmel (1984) notes that experimentally-determined resuspension factors vary over
eight orders of magnitude (i.e., 10'10 to 10"2 m'1).  Among the factors influencing resuspension
rates are soil/surface moisture, mean wind speed, gust intensity, partide/soil/surface properties,
and the nature, frequency, and magnitude of mechanical disturbances of the surface.
Development of a generalized particle resuspension model would involve a substantial effort
beyond the scope of the present study.
                                            2-1

-------
              Table 2-1
Factors Influencing Dry Deposition Rates
          (from Sehmel, 1980)
Micrometeorological
Variables
Aerodynamic roughness
- MaM transfer
(a) Parades
(b) Cases
-Heat
* Momentum
Atmospheric stability
Diffusion, effect of:
-Canopy
- Diurnal variation
•Fetch
Flow separation:
- Above canopy
- Below canopy
Friction velocity
Inversion layer
Pollutant concentration
Relative humidity
Seasonal variation
Solar radiation
Surface heating
Temperature
Terrain
- Uniform
- Nonuniform
Turbulence
Wind velocity
Zero-plane displacements
- Mass transfer
(a) Particles
(b) Gases
-Heat
- Momentum



Depositing Material
Particles

Agglomeration
Diameter
Density
Diffusion
- Brownian
. Eddy equal to
(a)Pardcie
(b) Momentum
(c)Heat
- Effect of canopy on
Diffusiophoresia
p7**f^ffqfln effect s
• Attraction
• Repulsion
Gravitational settling
Hygroscopicity
Impaction
cHOtDenturn
Physical properties
Resuspension
Shape
Size
Solubility
Thermophoresis

Gases

Chemical activity
Diffusion:
• Brownian
-Eddy
Parade pressure in equilibrium with
surface
Solubility
Surface Variable
A-H—flm^*"™-

* IfoflHyM
- Tricfaomes
• Pubescence
• Wax
Biotic surfaces
Canopy growth:
- Dormant
• Expanding
Senescent
Canopy structure:
- Area! density
• Bark
.Bole
- Leaves
-Porosity
• Reproductive structure
.Sou*
-Scon
-Type
BeonMunc nrooerriM
eorosianc properties
Leaf-Vegetation:
. Boundary layer
- Change at high winds
. Flutter
- Stomatal resistance
Non-biotic surfaces
pH effects on:
• Reaction
-Solubility
Pollutant penetration and distribution in
canopy
Prior deposition loading
Water


                 2-2

-------
                M
ST. LOUIS -VIS
ST. LOUIS - 1973
HEDGE
WATER. LAPSE ATM
Fe^ MAX RATE
GRASS, O STABILITY
ALFALFA
GRASS, NEUTRAL ATM.
CEMENT, MAX. RATE
GRASS, LAPSE ATM.
GRASS
WATER, NEUTRAL ATM.
GRASS
CEMENT, MAX RATE
FOREST
GRASS, MEDIUM
STUCCO, MAX RATE
GRASS, D STABILITY
CEMENT, MAX RATE
SNOW, LAPSE ATM.
GRASS
GREAT BRITAIN
SOU, CALCAREOUS
WATER, B STABILITY
SOU, ADOBE CLAY-MAX
STUCCO, MAX RATE
WATER. B STABILITY
STUCCO, MAX RATE
WHEAT
GRASS, D STABILITY
GRASS, B STABILITY
SOU, ADOBE CLAY-MAX
SOU, SANDY LOAM-MAX
SOU, SANDY LOAM-MAX
FOREST, 17 m
WATER, D STABILITY
GRASS, SHORT
SNOW, NEUTRAL ATM.
GRASS, STABLE ATM.
WATER, FRESH
SNOW
ICE
SNOW, LAPSE ATM.
SNOW, STABLE  ATM.
ASPHALT. MAX RATE
                                                                         oa
                                                                          o
                                                                          a
                                                                          &
                                                                          a
                                                                       •*-a
                                                                         x
                                                                      —a
                                                                       aa
   o
X—X
  A
 A
X—X
 A
                                                                  oa
                                                                 a-a
                                                                  A
                                                                  A
                                                               x— x
                                                                 a
                                                                 ?
                                                                 a
                                                                x
        A "MAXIMUM" RATES
        a GRASS
        X WATER
        <7 SNOW
        O OTHER
                                                  Ai , ,,,
                                                    DEPOSITION VELOCITY, cm/sec
Figure 2-1.    Summary of observed SO2 deposition velocities (Sehmel, 1980).

                                                2-3

-------
                                      zo     u (~10cm)
                            (cm s  )  (cm)    (m
                     A --  11    dOQ2      2.2*
                     0 ---  44    d02      7.2*
                              117    dl      13.8*
                     X - -40   ~0.05    ~8**
                      *SEHMEL AND SUTTER (1974)
                     **MOLLER AND SHUMANN (1970)
                                    PARTICl£ DIAMETER, urn
Figure 2-2.    Observed deposition velocities (vj as a function of particle size for 1.5 g/cm3 density
             particles. Measured by Sehmel and Sutler (1974) and Moller and Schumann (1970).
             Figure from Slinn et al. (1978). The gravitational settling velocity (Vj) is also shown.

                                          2-4

-------
2.1     Dry Deposition of Paniculate Matter

        Many models of dry deposition express the deposition velocity as the inverse of a sum of
"resistances" plus, for particles, gravitational settling velocity terms.  The resistances represent the
opposition to transport of the depositing material from a reference height through the turbulent
atmospheric surface layer, and through a quasi-laminar layer just above the surface to the surface
itself.  The major processes are briefly described below.

        Gravitational Settling. The gravitational settling velocity is a function of the particle diameter,
shape, and density.  Figure 2-3 shows the gravitational settling velocity (Vj) as a function of particle
size for several values of the particle density (p)  for solid spherical particles.  In many cases, effective
or aerodynamic particle diameters are reported which express the size and shape of the particle as an
equivalent diameter for a spherical particle of unit density, which simplifies the input to the deposition
model. Note that the  gravitational settling velocity represents a lower limit to the deposition velocity.
It can be seen that for larger particles, in the range of 20-40 jim diameter and higher, the deposition
velocity approaches the gravitational settling velocity, which indicates that the rate  of deposition is
dominated by the gravitational settling mechanism. The gravitational settling velocity decreases with
decreasing particle size. However, for particles smaller than about 20 jim diameter, the deposition
velocity curve shows larger and larger deviations  from the gravitational settling curve as the particle
size decreases. This is due to the effect of other mechanisms, discussed below, in enhancing the
deposition rate of smaller particles above that predicted by gravitational settling alone.

       Atmospheric Diffusion.  The rate of deposition can sometimes be limited by the transfer of
pollutant material to the vicinity of the surface by atmospheric turbulence. For example,  atmospheric
turbulence-limited deposition situations would typically occur during very stable conditions for an
elevated plume of material composed of small-sized particles with small gravitational settling. In the
lowest layer of the atmosphere,  the aerodynamic resistance is used to parameterize the rate of mixing
in terms of the wind speed, atmospheric stability, and surface roughness length.  The aerodynamic
resistance generally decreases (Le., the deposition velocity will increase) with increasing wind speed
and/or surface roughness.

       Quasi-laminar  Layer. Over smooth surfaces, a thin non-turbulent sublayer develops that can
be a significant obstacle to the transfer of the pollutant onto the surface.  For rough, real-world
surfaces, this sublayer  is constantly changing and  is likely to be intermittently turbulent. For this
reason, Hicks (1982) calls this layer the "quasi-laminar" layer.  It is also known as the deposition layer.
Small particles ( < 0.05 pm diameter) are transported through the quasi-laminar layer primarily by
Brownian diffusion. However, Brownian diffusion becomes less efficient as the particle size increases.
Particles in the 1-20 jim diameter range tend to penetrate the quasi-laminar layer by inertial
impaction. Since particles larger than 20 |im diameter are less efficiently captured, the inertial
impaction mechanism  is most effective in the 1-20 um diameter size range. Because particles in the
0.1-1.0 urn diameter size range are not efficiently transported across the quasi-laminar layer by either
Brownian diffusion or  inertial impaction, particles of this size have the lowest deposition velocities.

                                               2-5   -

-------
                i
                *
                §
                I/I
                o
                             STABLE ATMOSPHiRE WITH
                              ROUGHNESS HEIGHT, cm
                                                1               10

                                             PARTICLE DIAMETER. [MB
Figure 2-3.     Predicted deposition velocities for u. =  100 cm/s and particle densities of 1,4, and
               11.5 g/cm3.  Also shown  is the gravitational settikg velocity (v^.  Figure  from
               Sehmel (1980).

                                              2-6

-------
        Based on the discussion above, it is considered essential that the final generalized model
 selected be able to parameterize, at a minimum, the effects of gravitational settling, which is a
 dominate effect for large particles (> ~ 20  pm diameter), inertial impaction (dominates in the
 size range 1.0 to 20 um diameter), and Brownian motion (important for small particles less than
 about 0.1 |im diameter).

 2.1.1   ISC Method

        The basis for the present ISC deposition algorithm is found in Dumbauld et aL (1976)
 and Overcamp (1976). In the ISC approach, the particles are allowed to move toward the
 ground by the combined processes of atmospheric turbulence and gravitational settling.  At the
 surface, a portion of the plume determined by a user-specified reflection coefficient, yn,  is
 assumed to be reflected from the surface and the remainder (1 - YQ) is assumed to be retained
 by the surface. The reflection coefficient scheme is illustrated in Figure 2-4 for YQ = 0.0 (no
 reflection), 0 J (partial reflection), and 1.0 (full reflection).

    The effect of gravitational settling is assumed to result in a tilted plume with an angle,  9, to
 the horizontal given by:

        6 - tan'1 (Vg/u)                                                              (2-1)

 where vg is the gravitational settling velocity and u is the wind speed.

        The concentration in ISC is given by:
              KQD,V
                        exp
(2-2)
where, K is an units scaling factor,
       Q is the pollutant emission rate,
       D, is the decay term,
       V is the vertical term,.
       oy is the standard deviation of the concentration distribution in the
              horizontal,
       oz is the standard deviation of the concentration distribution in the
              vertical,
                                            2-7

-------
                                                                         TOTAL REFLECTION
                                                                                   (r»4.0)
                                                                          50% REFLECTION
                            GROUND
                                                             CONCENTRATION
    (a) SELECTED FOLDED NORMAL DISTRIBUTIONS   (b) RESULTING VERTICAL CONCENTRATION PROFlLf
Figure 2-4.   Illustration of the reflection coefGcient scheme used in ISC for reflection coefEcients
             of 0.0, 0.5, and 1.0.  Figure from Bowers et al., 1979.
                                          2-8

-------
       u  is the wind speed, and
       y  is the crosswind distance from the plume centerline to the receptor.

       The user must subdivide the pollutant size distribution into N particle settling categories.
A maximum of 20 particle settling categories is allowed.  The vertical term for a ground-level
receptor involves a summation over all N particle categories.
v-  E
                                                                                      (2-3)
where,
  Vm - 0.5 A
              E  ft
              i-O
                     +     + £
                             i-l
                                                                            (2-4)
                -  0.5
                                                                                      (2-5)
A2
exp
                 -  0.5
                                                                                      (2-6)
                                                                                      (2-7)
     Y." exp
                - 0.5
                                                                                      (2-8)
                                            2-9

-------
and,   4>B is the mass fraction of the pollutant emitted in the nth particle settling category,
       Ya is the reflection coefficient for particles in the nth particle settling category,
       Hy is the height lost at the downwind distance x due to the gravitational settling
       [Hr - OV x)/u],
       vg is the gravitational settling velocity,
       H,,, is the mixing height, and,
       H, is the effective plume height (stack height + plume rise).

       The ISC manual suggests the use of the following equation (McDonald, 1960) for
calculating the  gravitational settling velocity, vp in cm/a.

                                      P g 4
                                 V a 	r-
                                        1811                                         (2-9)

where, p is the particle density (g/cm3),
       g is the acceleration due to gravity,
       dp is the particle diameter (cm), and
       li is the absolute viscosity of air (ISC manual suggests (i  - 1.83 x 10^* g/cm/s).

The total deposition, D, is summed over all N particle categories, and is given by:
                   ,                                                                 (2-10)
Da  =     "*     tv	^  exp  .    ._  .
            (2*ayoz*)        *  \       (a,;
                                                                                    (2-H)
where the vertical term for deposition, V, is defined as:
                                            2-10

-------
   v1»[ 5 Ht + (i - *)jyv
     I
                         a, - a,}
     i-l
"  b (2tffM - S) -(1 -*)#v
   J2iHm - ff  +
-
-------
                   !	if   I   I   1
                        0.2     .0.4
                            REFLECTION COEFF1CIEHT
1.0
Figure 2-5.    Reflection coefficient as a function  of the gravitational settling velocity (from
             Dumbauld et aL, 1976).

                                       2-12

-------
 55 |im diameter (fuel oil releases).  Less than 1% of the mass of the Duphar was in particles
 with diameters < 30 jim.  For the fuel oil distribution, less than  1% of the mass was in particles
 smaller than 20 um diameter. Particles  of this large size are nearly completely controlled by
 gravitational settling, which, because it is a characteristic of the particle and does not depend on
 factors such as meteorological conditions and surface characteristics,  is compatible with the use
 of spatially and temporally constant reflection coefficients.  However, the deposition of small
 particles is controlled by other physical processes such  as inertia! impaction and Brownian
 motion, which are affected by both meteorological and surface conditions. These effects  are not
 easily parameterized within the framework of a reflection coefficient  model

       For small particles, the ISC reflection coefficient curve (Figure 2-5) was derived based
 on an extrapolation of the gravitational settling-dominated data for larger particles (Dumbauld
 et aL, 1976). The extrapolation led to the assumption that particles with settling velocities less
 than 0.1 cm/s (approximately 5.7 um diameter for unit density particles) are completely
 reflected (Le., experience no deposition). However, this assumption is not consistent with
 measurement studies of small particles, as discussed previously, which show significant
 deposition velocities for particles at and  below 5.7 jim diameter (e.g., see Figures 2-2 and 2-3).
 It was concluded that the ISC scheme could significandy underestimate small particle deposition
 because several important physical processes are not parameterized in its approach.  Therefore,
 die ISC method is not  considered appropriate for small particles. However, it was tested and
 intercompared with the other approaches to serve as a  reference point for the performance of
 die other models.

 2.1.2    CARS Model

       Sehmel and Hodgson (1978) and  Sehmel (1980) proposed a model for predicting
 deposition velocities of particles above smooth surfaces. The basis of the model is a set of wind
 tunnel observations of deposition for monodispersed particles to  surfaces such as gravel,
 artificial grass, brass shim stock, and water.  The model consists of empirical equations for
 transfer resistances derived from a least-squared empirical fit of deposition velocity as a function
 of particle size, density, surface roughness, and  friction  velocity. The  equations were converted
 into a computer code by the California Air Resources Board (CARB) and is widely known as
 the CARB model.  It is also used in the Fugitive Dust Model (FDM)  (Winges, 1990).

       In the CARB model, integrated resistances to mass transfer are computed within two
layers.  The  first layer extends from a reference height of one meter above the surface down to
one centimeter above the surface. In this layer, atmospheric turbulence dominates mass
transfer.  Eddy  diffusivities are used to describe the  transfer rate. The second layer is the

                                      ,    2-13

-------
deposition surface layer within one centimeter of the surface.  The integrated resistance within
the deposition surface layer is derived from a statistical fit of the wind tunnel particle deposition
data.  Sehmel and Hodgson express the deposition velocity as:
                              L - expf-v^ + /3y«.]
                                      1                                              (2-13)
where, vd is the deposition velocity (cm/s),
       vg is the gravitational settling velocity (cm/s),
       Ij2 is the atmospheric diffusional resistance (dimensionless),
       I3 is the surface resistance integral (dimensionless), and,
       uu is the surface friction velocity (cm/s).

       The gravitational settling velocity (cm/s) in the GARB model is calculated as:
                                                                                    (2-14)
where, dp is the particle diameter (um),
       p is the particle density (g/cm3),
       PAJR is the density of air (GARB assumed p^ = 12 x 10"3 g/cm3),
       T is the air temperature (°K),
      • (i is the absolute viscosity of air (GARB used n = 1.78 x 10"4 g/cm/s), and
       the constants q and Cj were assigned values of 9.73 X 10"3 and 1.0 x 10"*,
              respectively.

       The atmospheric diffusion resistance used by Sehmel and Hodgson (1978) is based on
 the flux profile relationships of Businger et aL (1971). For neutral or stable conditions,

       IB-  [Inz^) + 4.7(Zl - z^/LJ/k                                             (2-15)
 where, zl is the upper limit of the atmospheric diffusional resistance integral (i.e.,  100 cm),
        Z2 is the lower limit of integral (i.e., 1 cm),
     '   L is the Monin-Obukhov length (cm), and,
        k is the von Karman constant (Sehmel used a value of 035).
                                            2-14

-------
        For unstable conditions, the atmospheric diffusional resistance integral is:

                                                   - 1.)]}
                                                                                    (2-16)


              (1. - 15VL)0-23                                                        (2-17)

              (1. - tf.Zj/L)'*5                                                        (2-18)
       The surface resistance integral is an empirical relationship based on the wind tunnel
observations.  For particles with a diameter a 0.01
      73 = exp{-378.051 + 16.498 ln(Sc) + to^^-11.818 - 0^863
           0.3226 In(
-------
       The data requirements of the model are relatively simple (particle size, density, surface
roughness, and routine meteorological parameters to compute the friction velocity and
Monin-Obukhov length).

       The plots of deposition" velocity in Sehmel (1980) (e.g., see Figure 2-3) show a
reasonable variation of deposition velocity as a function of model'parameters (i.e., density, size,
surface roughness, friction velocity).  The predicted deposition velocity is dose to the
gravitational settling velocity for large particles (e.g., greater than about 20 \im diameter), and
decreases with decreasing particle size to about 0.1-1.0 pm, where it reaches a minimum.  The
deposition velocity then increases with decreasing particle size for smaller sized particles.  This
behavior is consistent with the importance of Brownian motion in enhancing deposition rates for
very small particles. The GARB scheme produces increased deposition rates for increased
particle density,  surface roughness length, and friction velocity, which is expected based on
physical considerations.

       The main concern about the CARB model deals with the generality of the  highly
empirical relationship for the surface resistance integral (Eqn. 2-19). I3 is based on wind tunnel
data for relatively smooth surfaces conducted under a limited range of conditions.  For example,
in order to avoid extrapolation, the CARB implementation  of the model does not  allow the
surface roughness length used in the algorithm to exceed 10 cm, even though most real world
surfaces have significantly greater roughness lengths (e.g., see Table 2-2). In addition, sensitivity
testing of the model has shown it exhibits some non-physical behavior when the inputs are
varied beyond the range of conditions tested in the wind runnel  For example, the CARB model
showed a very strong sensitivity to temperature which is not exhibited by other deposition
models.  For particles in the 0.1-1.0 jim diameter size, a change on nearly an order of
magnitude in deposition velocity was predicted for a temperature change from 0°F to 100°F (see
Figure 2-6).  This behavior is probably an artifact of the regression equations used to fit the
surface resistance integral to the wind tunnel data.  The original wind tunnel tests  were
performed at a constant temperature, and, as a result, application to a realistic range of
atmospheric conditions involves extrapolation outside the range on which the model was
developed.

       A second problem noted from the sensitivity testing is that the deposition velocities show
a kink in  the curve at about 0.03 um particle diameter. For particles smaller than about 0.1 jim
diameter, the deposition velocity increases with decreasing particle diameter.  However, the
CARB model  shows this trend only to about 0.03 um diameter, beyond which the  deposition
velocity is predicted to decrease or level off with decreasing particle diameter (see Figure 2-7).
The kink in  the  curves appears to be another artifact of the regression equations which occurs
as a result of extrapolation outside the range of conditions on which the model is based.
                                           2-16

-------
                                        Table 2-2
              Typical Surface Roughness Lengths for Various Land Use Types
                                 (From Hjelmfelt, 1982)
                                           Typical
Land Use                            Roughness Length (cm)
Urban - Commercial/Industrial                200
Common residential - single family dwellings     20
Compact residential - multi-family dwelling       50
Metropolitan natural (parks, golf courses)        15
Agricultural - rural                   .         20
Semi-rural                                    20
Undeveloped, wasteland                         5
Forest                                      100
Bottomland agricultural                        15
                                         2-17

-------
                                KUN  tUti
                      PARTICLE DENSITY = 1 g/cm
         10J
§  10
u

s  10
H
N4
2  10
            -2
            -3
          10
            -4
               0.01      0.10      1.00    10.00     100.00

                      PARTICLE  DIAMETER (MICRONS)


                    SENSITIVITY RUN FOR TEMPERATURE

                      PARTICLE  DENSITY = 4 g/cm3
          10
       H    -1
       s10
       §  «'
       H4
       H
       M

       I
       fc  10
            -2
      -3
            -4
          10
               0.01       0.10     1.00    10.00    100.00
                      PARTICLE DIAMETER  (MICRONS)

Figure 2-6.    Deposition velocity  as  a function of particle  diameter  as  predicted  by the
             Sehmel/CARB model for three different values of ambient temperature (0°, 60°,
             100° F).
                                       2-18

-------
             10
10
                                       PARTICLE DIAMETER  (MICRONS)
Figure 2-7.    Deposition  velocity  as  a function  of particle  diameter  as predicted  by the
             Sehmel/CARB model for surface roughness lengths of 0.001, 0.01, 3, and 10 cm.

                                        2-19

-------
       lii addition to evaluating the basic GARB model, several modifications to the model
were made in an attempt to address some of the concerns raised above. In the model
evaluation (Section 7), the basic GARB model described by Eqns. (2-13) to (2-21) is designated
GARB 1. Three variations of the model were  developed and evaluated (GARB 0, GARB 2, and
GARB 3).  The modifications to each version of the model are described below.

GARB 0 -     same as the GARB 1 except the portion of the code which limits the value of z0
              to 10 cm was eliminated. Thus, the GARB 0 revision uses actual surface
              roughness lengths which are allowed to be greater than 10 cm.

GARB 2 -     same as GARB 1, except in Eqn. (2-21), a constant temperature of 75*F is
              always used rather than the actual air temperature.  This change was  intended to
              eliminate the temperature sensitivities of the model noted above.

GARB 3 -     contains changes to eliminate the temperature dependency (as in GARB 2) and
              an adjustment to the surface resistance integral, I3. The value of I3 used in GARB
              3 is that given by Eqn. (2-19) divided by (1.0 + LAI), where LAI is the leaf area
              index. In addition, a fixed reference length of 0.5 cm is used in GARB 3 in the
              calculation of the LAI-adjusted  I3 instead  of z0.

       The Sehmel and Hodgson (1978) empirical relations are based on simple surfaces. For
compound surfaces which are aerodynamically independent, it seems reasonable to expect the
deposition velocity to be increased by the additional surface area. The LAI adjustment in
GARB 3 is made to test this assumption.

       Also, the roughness length in the surface integral resistance, Ij, differs from the
traditional roughness length defined by other researchers such as Plate and Quraishi (1965).
Rather, it is related to the thickness of the laminar sublayer, which is proportional to the ratio
of the molecular diffusivity to the surface friction velocity.  The constant of proportionality for
an established sublayer is approximately 20.  Thus, for a friction velocity of 0.1 m/s and a
diffusivity of 2 x 10"J m2/s, the resulting sublayer depth is .4 cm, which is close to the largest
roughness length used by Sehmel and Hodgson (1978). Thus, for the relatively smooth surfaces
tested by Sehmel, z0 may be a good approximation of the sublayer depth length scales, but for
the LAI-adjusted I3, it may be more appropriate to use a constant reference length..  A constant
value of 0.5 cm is tested in GARB 3.
                                          2-20

-------
2.13  ADOM Model - Particles

       The Acid Deposition and Oxidant Model (ADOM) contains a dry deposition algorithm
(Pleim et aL, 1984) which has served as the basis for the deposition scheme in several other
models, including CALPUFF (Scire et aL, 1990) and CALGRID (Yamartino et aL, 1992).

       The deposition flux, F, is calculated as the product of the concentration, x, at a reference
height and a deposition velocity, vd.

       F = x-vd                                                                   (2-22)

       The general approach used in the resistance methods such as in the ADOM model is to
include explicit parameterizations  of the effects of Brownian motion, inertia! impaction, and
gravitational settling. The deposition velocity is written as the inverse of a sum of resistances to
pollutant transfer through various  layers, plus gravitational settling terms (Slinn and Slinn, 1980;
Pleim et aL, 1984).

                                        	 * vt
                                                                                   (2-23)
where, vd is the deposition velocity (cm/s),
       vg is the gravitational settling velocity (cm/s),
       r, is the aerodynamic resistance (s/cm), and,
       rd is the deposition layer resistance (s/cm).

       Note that for large settling velocities, the deposition velocity approaches the settling
velocity (vd - Vg), whereas, for small settling velocities, vd tends to be dominated by the r, and rd
resistance terms.

       The  lowest few meters of the atmosphere can be divided into two layers: a fully turbulent
region where vertical fluxes are nearly constant, and the thin quasi-laminar sublayer.  The
resistance to transport through the turbulent, constant flux layer is the aerodynamic resistance
It is usually  assumed that the eddy diffusivity for mass transfer within this layer is similar to that
for heat. The atmospheric resistance formulation used in ADOM is based on Wesely and Hicks
(1977).
                                           2-21

-------
                              ^(%/0-tJ
                                                                                   (2-24)
where, t|rH is a stability adjustment factor,
       u. is the surface friction velocity (cm/s),
       k  is the von Karman constant (0.4),
       ZR is reference height (m) (-10 m), and,
       ZQ is the surface roughness length (m).
                      -5 ^l                         0 < zll <  1
                      0
                                                                                   (2-25)
       The approach used by Pleim et a! (1984) to parameterize the deposition kyer resistance
terms is:
                                                                                   (2-26)
where, Sc is the Schmidt number (Sc = u/D) (dimensionless),
       u  is the viscosity of air (- 0.15 cm2/s),
       D is the Brownian diffusivity (cm2/s) of the pollutant in air,
       St is the Stokes number [St = (Vg/g)(u.2 /o)] (dimensionless), and,
       g  is the acceleration due to gravity (981 cm/s2).

       The gravitational settling velocity (cm/s) in the ADOM model is calculated as:

                          „  , (P-P^*^  ,
                                                                                   (2-27)
where, p is the particle density (g/cm3),
            is the air density (- 1J2 x 10"3 g/cm3),
          is the particle diameter (pm),

                                            2-22

-------
        (i is the absolute viscosity of air (- 1.81 x 10"* g/cm/s),
        Cj is air units conversion constant (1 x 10"* cnrYiim2), and
        SQT is the slip correction factor, which is computed as:
                      s
                                         10-* d
                                                                                   (2-28)
 and, Xj, a1} 83, aj are constants with values of 6.5 x 10"*, 1.257, 0.4, and 0.55 x 10"*, respectively.

       The diffusivity of the pollutant (in cm2/s) is computed from the following relationship.

                           D = 8.09 x  10'10 I—-
                                                                                  (2-29)
where T is the air temperature (°K).
       The first term of Eqn. (2-26), involving the Schmidt number, parameterizes the effects of
Brownian motion. This term controls the deposition rate for small particles. The second term,
involving the Stokes number, is a measure of the importance of inertial impaction, which tends
to dominate for intermediate-sized particles in the 2-20 |im diameter size range.

       The data requirements of the ADOM-type deposition model are identical to those of the
GARB model  Particle size, particle density, surface roughness length, friction velocity, and
Monin-Obukhov length are input parameters used by the model

       The ADOM approaches produce patterns of deposition velocities as a function of
particle size, density, etc. similar to those of the GARB model However, the ADOM algorithm
tends to predict somewhat higher deposition  velocities in the 5-15 urn diameter size range than
the GARB model and lower values in the 0.1-5.0 \im diameter range, although the general shape
of the deposition velocity curves are similar.  The minimum deposition velocity tends to occur
closer to 1.0 jim than 0.1 urn diameter.

       The parameterizations of the resistance models for Brownian motion and inertial
impaction effects involve empirical factors derived from field studies and wind tunnel
experiments. The sensitivity analyses with the ADOM model showed no unusual response to

                                           2-23

-------
temperature variations over the range from 0° F to 100°F.  The resistance models ail showed a
steady increase of deposition velocity with decreasing particle diameter in the small particle
range (< 0.1  um diameter).

       The basic (core) ADOM model defined by Eqns. (2-23) to (2-29) is designated as
ADOM 1.  Two sets of changes to the ADOM formulation, described below, were also tested.

       The -3/St exponent in the inertia! term of Eqn. (2-26) may represent too sharp a cutoff
as pointed  out by Slinn (1982).  Slinn suggests that a better function for the inertia! impaction
term, E^ for canopies is:
                1 +
where S^ is a bulk the Stokes parameter defined in Slinn (1982) as S^ = St(jiU/u.A). At the
present time the obstacle scale size A has been set to 1 mm. Note that Slinn (1982) also
suggests an explicit parameterization of the interception mechanism.  However, due to its
reliance on parameters that are not available on a routine basis, it has not been tested in this
version of the model

       The power law dependence on the Schmidt number for the Brownian diffusion term is
likely to be a function of the surface type, with established laminar kyers over a smooth water
surface having an exponent closer to -0J (Slinn and Slinn, 1980), while for more complex
surfaces, the exponent is likely to be nearer -0.7.

       Therefore, in the version designated ADOM 2, a hybrid deposition layer resistance was
computed as:
                              &" *•;—H
                                     i + ag
                                                                                  (2-31)

instead of using Eqn. (2-26), with n »  -0 J for z0 i 10 cm and n = -0.7 for z0 >  10 cm.

       A second change in ADOM 2 is an allowance for a small adjustment to the deposition
rates to account for possible phoretic effects.  Some examples of phoretic effects (Hicks, 1982)
are:
                                           2-24

-------
       THERMOPHORESIS:
       DIFFUSIOPHORESK:
Particles dose to a hot surface experience a force directed
away from the surface because, on the average, the air
molecules impacting on the side of the particle facing the
surface are hotter and more energetic.

Close to an evaporating surface, a particle is more likely to
be impacted by water molecules on the side of the particle
facing the surface.  Since the water molecules  have a lower
molecular weight than the average air molecule,  there is a
net force toward the surface, which results in a small
enhancement of the deposition velocity of the particle.

A second effect is that the impaction of new water vapor
molecules at an evaporating surface  displaces a certain
volume of air. For example, 18 g of water vapor
evaporating from 1 m2 will displace 22.4 liters  of air at
standard temperature and pressure (STP) conditions
(Hicks, 1982). This effect is called Stefan flow.  The Stefan
flow effect tends to reduce deposition fluxes from an
evaporating surface. Conversely, deposition fluxes to a
surface experiencing condensation will be enhanced.

Attractive electrical forces have the potential to  assist the
transport of small particles through the quasi-laminar
deposition layer, and thus could increase the deposition
velocity in situations with high local field strengths.
However, Hicks (1982) suggests this effect is likely to be
small, in most natural circumstances.
       Phoretic and Stefan flow effects are generally small  However, for particles in the range
of 0.1 - 1.0 |im diameter, which have low deposition velocities, these effects may not always be
negligible.  Therefore, the ability to specify a phoretic term to the deposition velocity is added
to ADOM 2 (Le., vd' = vd + v^,^, where vd' is the modified deposition velocity and v^,^ is
the phoretic term).  Although the magnitude and sign of vd(phor) will vary, a small, constant value
of + 0.01 cm/s is used in ADOM 2 for testing purposes.

       Another version of the model, ADOM 3, was tested which included the ADOM 2
changes,  and a leaf area index (LAI) adjustment to rd.  In ADOM 3, rd is that given by Eqn.
(2-31) divided by  (l.Q  +  LAI).
       ELECTROPHORESIS:
                                          2-25

-------
11.4   UAM-V Model

       The deposition formulation variable grid version of the Urban Airshed Model (UAM-V)
(Gray et aL, 1991) is similar to that in ADOM.  The deposition velocity is expressed in terms of
an aerodynamic resistance (rj, deposition layer resistance (rj,  and a gravitational settling
velocity (v^) according to Eqn. (2-23). The formulation for r, in UAM-V is identical to that in
ADOM (Eqn. 2-24). The form of the stability adjustment factor, ty^, is given by Eqn. (2-25),
except  the coefficient for positive L is -4.7 in UAM-V instead of -5."

       The main difference is in the formulation of the inertial impaction term of the
deposition layer resistance.

                      rd * [ 025 (Sc-™  +  Ct A) T1 u.~l
                                                                                 (2-32)
where the parameter Q is computed each hour to force the Sc and St terms  to be equal at a
critical diameter of 03  Jim.

       The form of the gravitational settling velocity equation in UAM-V is  identical to that in
ADOM (Le., Eqns. 2-27 and 2-28). The values of the constants in the vg equations are slightly
different The values of x^ ji, and p^ used in UAM-V are 6.53 x IV6, 1.83 x 10"* g/cm/s, and
1.0 x IQ'3 g/cm3, respectively.

       The data input requirements of the UAM-V dry deposition model (particle size, particle
diameter, roughness length, friction velocity,  and Monin-Obukhov length) are the same as those
required by the CARB and ADOM models.

        The UAM-V dry deposition algorithm described above  is  designated as UAM 1 in the
model  evaluation tests. A second version, called UAM 2, was also evaluated. UAM 2 was
modified to include  a LAI adjustment to the deposition layer resistance (Le., rd in UAM 2 is
that given by Eqn. (2-32) divided by (1.0 + LAI)).

22  Dry Deposition of Gaseous Pollutants

       Many of the available models for predicting dry deposition of gaseous pollutants have a
similar structure.  The deposition velocity is expressed as the inverse of a sum of resistances.

                                                                                 (2-33)
                                          2-26

-------
 where, r, is the aerodynamic resistance (s/m),
        rd is the deposition layer or quasi-laminar layer resistance (s/m), and,
        rc is the bulk canopy resistance (s/m).

        The gravitational settling velocity is not a factor for gases, and therefore does not appear
 in Eqn. (2-33).  The aerodynamic resistance for gases is the same as discussed previously for
 particles, see Eqn. (2-24).  It is determined from meteorological and surface parameters and
 does not depend on the pollutants characteristics.  The deposition layer resistance involves
 transport by molecular diffusion through the  thin quasi-laminar sublayer which intermittently
 exists just above the surface. The deposition layer resistance is a  function of the properties of
 the pollutant (Le., molecular diffusivity in air) and meteorological conditions. The canopy
 resistance involves the physical capture or chemical reaction of the pollutant within the
 vegetative canopy or at the surface.  The canopy resistance, r9 Is often the controlling resistance
 determining deposition flux. Therefore, considerable effort is devoted to estimating re in many
 of the deposition models.

        There are many models based on this simple resistance concept. Three widely-used
 techniques appropriate for deposition of gases have been selected for future further evaluation.
 In some cases, various components of the algorithms are very similar or identical to each other
 (e.g., the formulation of rj. However, other important variations exist,  especially in the
 parameterization of r& which wfll be intercompared, tested and evaluated.  Because the goal of
 this study is to produce an algorithm which can be used on a routine basis for regulatory
 applications, top consideration has been given to simpler parameterizations which capture the
 most significant features of the physical system while requiring modest, readily-available data as
 input.

   The major features of the algorithms selected for the current study  are described below.
 The  original references provide a detailed description of the algorithms.

 2.2.1    RADM Model

       The dry deposition algorithm used in the Regional Acid Deposition Model (RADM)
 meets all of the  model selection criteria listed in Section 1.  It has undergone considerable
 testing and refinement over the past five years. The RADM deposition  scheme is described in a
series of reports and papers, including Sheih et aL, 1986; Walcek et al,  1986; and Weseiy, 1989.
It has also been  incorporated into the latest version of the UAM-V model  (Gray et al.,  1991).

       The early version of the RADM deposition model used fairly standard formulations for
r, and r^ and look-up tables for rc based on land use and season.  However, the difficulty with
the look-up table approach is that detailed rc tables must be developed for  each pollutant of

                                           2-27

-------
interest, which can be difficult for many pollutants. The latest version of the model computes rc
based on the bulk surface resistance along several parallel pathways of mass transfer (Wesely,
1989).  The equation for re is:
       r-  -    7T7-  * T * T-rr * HrH         •                       
              |/*   r»    rta    r
-------
                                         Table 2-3
       Values of Chemical Input Parameters Required by the RADM Deposition Model
                                   (From Wesety, 1989)
Gaseous species
Sulfur dioxide
Ozone
Nitrogen dioxide
Nitric oxide
Nitric acid vapor
Hydrogen peroxide
Acetaldehyde
Formaldehyde
Methyl hydroperoxide
Peroxyacetic acid
Formic acid
Ammonia
Peroxyacetyl nitrate
Nitrous acid
Symbol
SO2
°3
NO2
NO
HN03
HA
ALD
HCHO
OP
PAA
ORA
NHs
PAN
HN02
^.
1.9
1.6
1.6
13
1.9
1.4
1.6
13
1.6
10
1.6
1.0
16
1.6
Solubflity<2)
IxlO5
0.01
0.01
2 x lO'3
IxlO14
IxlO5
15
6xl03
240 .
540
4x10*
2X104
3.6
IxlO5
Reactivity^
0
1
0.1
0
0
1
0
0
0.1
0.1
0
0
0.1
0.1
(1)
(2)
Ratio of molecular diffusivity of water to that of the pollutant

Effective Henry's law coefficient (m/atm)
(3)     Pollutant reactivity parameter as defined by Wesely (1989)
       (0 = nonreactive, 0.1 = slightly reactive, 1 =» highly reactive)
                                           2-29

-------
                                 -Table 2-4
         Summary of Input Requirements of RADM Deposition Module
                                  for Gases
Pollutant Characteristics:
       - solubility (effective Henry's law coefficient)
       • molecular diffusivity
       - reactivity class parameter (nonreactive, slightly reactive, or highly reactive)

Geophysical Data:
       - surface roughness length
       . land use category

Meteorological Variables:
       - friction velocity
       - Monin-Obukhov length
       • air temperature
       - solar radiation
                                     2-30

-------
where dt is an empirical constant (- 2).  The Schmidt number is defined as the ratio of the
viscosity of air (~ 0.15 cm2/s) to the molecular diffusivity of the pollutant in air.

       The canopy resistance is computed by considering three pathways for uptake of the
pollutant within the vegetation or at the surface:

       (1)    transfer through the stomatal pore and dissolution or reaction in the leaf
              interior,
       (2)    reaction with or transfer through the leaf cuticle,  and,
       (3)    transfer into the ground or water surface.

This is expressed as three resistances in parallel

       rc  =  [LAI/rf + LAl/r^ + 1/rJ'1                                            (2-36)

where, LAI is the leaf area index (ratio of leaf surface area divided by ground surface area).
              The LAI can be estimated from land use type and season (e.g.,  see Table 2-5),
       rf is the internal foliage resistance,
       *„£ is  the cuticle resistance, and,
       rg is the ground resistance.

       SO2 is used as a reference species because considerable data are available for it.  The
model inputs for other species are computed by scaling the reference values for SO2 by the
relative adjustment factors for the  other compounds.  For example, the cuticle  resistance for a
pollutant twice as reactive as SO2 is assumed to be one half the reference SO2  cuticle resistance.
Typical input values for many chemical species are described in the ADOM User's Guide.

       The data requirements of the ADOM-type deposition models are very  similar to those of
the advanced RADM deposition model Table 2-6 lists the required input parameters.  All of
the geophysical parameters can be derived from a classification of the land use in the area (e.g.,
urban, agriculture, forest, etc.).  The meteorological variables are either observed routinely (e.g.,
temperature) or can be computed from routinely-available data.  The pollutant characteristics
can generally be obtained from the literature.

2.23   NOAA/ARL Models

       A hierarchy of dry deposition models developed by NOAA/ARL are described in papers
by Meyers and Baldocchi (1988), Baldocchi et aL (1987), Meyers (1987), Hicks et ai. (1987), and

                                           2-31

-------
                                       Table 2-5
             Leaf Area Index Values as a Function of Land Use Type and Season
                                 (From Scire et aL, 1986)
                                     Vegetative Growing Season"
Land Use Type             I            JI            m           IV
Water
Deciduous forest
Coniferous forest
Swamp
Cultivated
Grassland
Urban
Desert shrubland
0.00
6.00
7.00
100
3.00
2.00
0.30
0.10
0.00
6.00
7.00
1.50
LOO
1.50
0.20
0.10
0.00
0.50
7.00
1.00
0.20
1.00
0.05
0.10
0.00
0.40
6.00
0-50
0.01
0.50
0.01
0.05
       * Definitions of season categories:
              I      = peak growing season
              n      = early growing season
              HI     = non-growing season without snow
              IV     = non-growing season with snow
                                          2-32

-------
                                  Table 2-6
      Summary, of Input Requirements of ADOM-type Deposition Modules
                                  for Gases
Pollutant Characteristics:
       - solubility (Henry's law coefficient)
       - molecular diffusivity
       - pollutant reactivity
       -aqueous phase dissociation constant

Geophysical Data:
       - land use category
       - surface roughness length (derived from land category)
       - leaf area index (derived from land use category)

Meteorological Variables:
       - friction velocity
       - Monin-Obukhov length
       - air temperature
       - solar  radiation
                                   2-33

-------
others. There are four basic models of increasing complexity and sophistication (as well as
increasing data requirements and computational cost) referred to as model I through model IV.
The simpler two models (I and n) were reviewed in the current project. Model I is known as the
"big leaf model because it treats the plant canopy as a single surface partitioned into shaded
and sunlit portions to account for differences in stomatal resistances. The aerodynamic
resistance in model I is approximated (Hicks et aL, 1987) by:

                                   neutral, stable conditions
                                  unstable conditions
                                                                                 (2-37)

where, u is the wind speed (m/s),  and,
       oe is the standard deviation of the horizontal wind direction fluctuations (deg.).

       The level II model is similar to model I, except that a somewhat more refined estimate is
made of the aerodynamic resistance and canopy resistance.  The formulation of ra in model n
may be better suited  for regulatory applications because it does not need an estimate of o9,
which is not routinely measured at airport meteorological stations. Model n computes r, based
on the surface friction velocity and Monin-Obukhov length.

The deposition layer  resistance is parameterized in models I and n as:


                           •"•l^lffiT
                                 1    *A  ^               .                       (2-38)

where, Dt is the thermal diffusivity of air (cm2/s), and,
       D, is the molecular diffusivity of the pollutant (caffs).

       The canopy resistance is expressed as a composite resistance composed of several
parallel pathways, including transfer to the leaf stomata into the plant tissue, transfer to the leaf
cuticle, or transfer directly to the ground or water surface.

       The variables used by the model to compute the stomatal resistance include detailed
information on vegetation which would not be available for many routine applications.  The data
and computational requirements of the level HI model (K-theory model) and level IV
(higher-order closure model) are considered too extensive for routine applications. Therefore,
the NOAA/ARL models are not being used as core models in the current study, although the
parameterization of some individual resistances may be included in future hybrid versions of the
core gas  deposition models to be evaluated.

                                          2-34

-------
 3. Plume Depletion Techniques

     Several different approaches to account for the depletion of the pollutant from the plume
 due to dry deposition processes have been reviewed. The schemes examined indude the source
 depletion method (Chamberlain, 1953), the surface depletion method (Horst, 1977), the
 K-theory approach (Rao, 1981), and a modified source depletion technique (Horst, 1983).
 These models determine how the pollutant which is deposited at the surface is  removed from
 the plume.

 3.1 Source Depletion

     The source depletion technique is the simplest method for removing pollutant material from
 a Gaussian plume. At each downwind distance, the source term (emission term) of the
 Gaussian equation is adjusted (decreased) to simulate the effect of increased pollutant removal
 with distance. At any distance, x, the source strength, Q is:

                       Q(x) = 
-------
3.2 Surface Depletion

    The surface depletion technique (Horst, 1977) allows the pollutant to be depleted in the
vicinity of the surface rather than throughout the vertical extent of the plume. Thus, the
resultant plume is allowed to assumed a non-Gaussian shape. This is done by treating the
ground surface as "negative" sources, representing a sink for the pollutant  The surface
depletion equation is:

                    If  7X(M.

                                                                                   (3-2)
where        D is the diffusion function (D
              X is the concentration at spatial coordinates (x,y,z), and
              the other terms have been defined previously.

    The surface depletion method is generally considered to be the most accurate solution of
plume depletion and is frequently used as a reference method for other techniques.  However,
the numerical solution of the complex integral requires between 10 and 100 times more
computer time than the source depletion technique. The large computational cost of the surface
depletion method makes it impractical for routine use. Another drawback of the surface
depletion technique for the current application is that its applicability is restricted to pollutants
with negligible gravitational settling velocities.

33 K-theory Approach

    Rao  (1981) describes a K-theory model for estimating concentrations in the case of
gravitational settling and deposition. The concentration is given by:

                            x(w). «• L *1
                                         u  I  Lt
                                                                                   (3-3)
                                            3-2

-------
where
gt (x,y) is the crosswind diffusion function,
8*2 (*»y) & tne vertical diffusion function modified for deposition effects, and
L^ L, are the length scales of the concentration distribution in the y and z
       directions.
                                                                                   (3-4)


                                                                                   (3-5)
                        g2
                        exp,
                               2o
                          exp
                                             -(z+ff?
                                 2a
                                (vd - v,/2) r
                                       2x
                                                                                   (3-7)
                                                                                   (3-8)
                                                                                   (3-9)
    The K-theory model was included in the evaluation study of Doran and Horst (1985).. They
noted that the K-theory model is valid only when the vertical dispersion coefficient, av varies as
x1^2. For more general forms for oz (e.g., such as those used in regulatory models such as ISC),
the K-theory model does not conserve mass.  Doran and Horst developed a method to  estimate
the missing mass, and then add it back into the plume. Although this correction improved the
                                            3-3

-------
performance of the model, it still did not perform quite as well as the modified source depletion
technique.

    Winges (1990) describes the numerical integration scheme used to compute the mass
conservation correction factors for the K-theory method in the Fugitive Dust Model (FDM).  A
least squares fit to computed values of the correction factors was calculated for the various
combinations of 6 wind speeds, 6 stability classes, 6 particle size classes, and 5 release heights
(Le., 1080 values), and entered into the FDM code in data statements. Winges (1992, personal
communication) has indicated that in  a new release of FDM, an external disk file with a much
larger number of the correction factors is supplied with FDM to cover a wider range of wind
speeds, size classes and release heights than the previous version of the model

    The lack of mass conservation in the unmodified K-theory model or alternatively, the need
to store large numbers of pre-computed mass conservation correction coefficients are viewed  as
the major drawbacks to the K-theory scheme, especially since alternatives without these
liabilities are available.

3.4 Modified Source Depletion

    Horst (1983)  describes modifications to the source depletion technique to account for the
change in the vertical distribution of the pollutant due to deposition at the surface.  In the
modified source depletion method, a profile correction factor, P (x^;) is defined and applied to
the diffusion function to account for plume depletion effects.
                        Ofcz)  -
                           .    • 
-------
                                                         i-i
                                                                                   (3-12)


where the resistance, R, is defined such that
                                            vd
                                                                                   (3-13)
and z,j is the reference height.

       Horst (1983) provides equations for other forms of oz as well, including ax5,
ax(l + bx)"1/7, and ax(l + bx)"1. However, he does not provide solutions for the form
ax(l + bx)1/2, which is needed for McElroy-Pooler coefficients under A and B stability
conditions.

   The modified source  depletion method has been shown to produce results in close
agreement with the exact surface depletion method (Horst, 1984), but only required a small
fraction of the computer  resources. In fact, its computational requirements were found to be
comparable to the source depletion method. In the dual tracer study of Doran and Horst
(1985), the modified source depletion model produced the best agreement with observations of
any of the plume depletion models tested. They used dual tracer field observations of a
depositing aerosol (ZnS)  and a non-depositing gas (SFS) in their evaluation.

   In summary, the modified source depletion method has been shown  to produce accurate
results compared to both the exact surface depletion method and field observations.  In addition,
its formulation is well-suited for use in a Gaussian model, it can treat gravitational settling
effects and plume tilt, it is computationally efficient, and it conserves mass exactly. For these
reasons, the modified source depletion method of Horst (1983) was selected as the approach to
use in the deposition module for assessing plume depletion.

       The  Horst (1983)  method was implemented in ISC2 during a later phase of this study.
Extensions to the method that were required are described in Appendix E.
                                           3-5

-------
4.     Calculation of Meteorological Variables

       Nearly all of the deposition models require estimates of the surface friction velocity (u.),
Monin-Obukhov length (L), and solar radiation.  The methods of Holtslag and van Ulden (1983)
are widely-used to estimate solar radiation and surface sensible heat flux from
routinely-available meteorological and surface (land use) data. The Holtslag-van Ulden scheme
has been implemented into the HPDM model (Hanna and Chang, 1991a), extensively compared
and tested with field data (e.&, Hanna and Chang, 1992), and has been shown to produce
reasonable results.  Therefore, the techniques used in the HPDM meteorological preprocessor
are used to produce the micrometeorology variables required in the deposition calculations.

4.1    Unstable/Neutral Conditions

       The energy balance at the surface can be written  as:


        
-------
             0.38
where T is the measured air temperature (deg. K),
       A is the albedo,
       o  is the Stefan-Boltzmann constant (5.67 x 10"* W/nr'/deg. K4),
       N is the fraction of the sky covered by douds,
       4>  is the solar elevation angle (deg.),
       a  is an empirical surface moisture parameter, and,
       S  is the slope of the saturation enthalpy curve [S=s/YJ, where
              s=a(q,)/3(T) and Y-c/L,
       A.  is the latent heat of water vaporization,
       q, is the saturation specific humidity, and,
       Cp is the specific heat at constant pressure.

       The four terms in the numerator of Eqn. (4-2) account for absorption of short-wave
radiation at the surface, incoming long-wave radiation from gaseous components of the
atmosphere (e.g., water vapor and carbon dioxide), incoming long-wave radiation due to douds,
and outgoing long-wave radiation from the surface, respectively. The factor in the denominator
(1+Cj), results from the use of air temperature rather than the more difficuit-to-determine
surface radiation temperature in the equation. The term in the first set of parentheses in Eqn.
(4-3)  represents short-wave solar radiation in the absence of douds.  The second term
(l+b1Nt)2), accounts for the reduction of incoming solar radiation due to douds (bj is negative).
The values for the empirical constants Cj, Cj,  aa, a,, bt, and t>2 suggested by Holtslag and van
Ulden (1°983) are shown  in Table 4-1.

       The flux of heat into the ground or storage in surface materials, Qp is usually
parameterized during the daytime as a fraction of the net radiation (e.g., DeBruin and Holtslag,
1982; Oke, 1978).

       Q8 =  cgQ.                                                                (4-5)

where cg is an empirical coefficient which depends on the properties of the surface.  Hoitslag
and van Ulden (1983) obtained a value of cg of 0.1 for a grass covered surface in the
Netherlands. Oke (1982) indicates that typical ranges for cg are 0.05 to 0.25 in rural areas, 0.20
to 0.25 in suburban areas, and 0.25 to 030 in urban regions and suggests that typical values of cg
are 0.15, 0.22, and 0.27 for rural, suburban, and urban areas, respectively.
                                            4-2

-------
                         Table 4-1
              Values of Net Radiation Constants
                (Holtslag and van Ulden, 1983)
Constant                                     Value
                                             990 W/m2
                                             -30 W/m2

                                             -0.75
                                             3.4

                                             531xKrBW/mVdeg. K6
                                             60 W/m2
                           4-3

-------
       The anthropogenic heat flux, Qf, can usually be neglected, except in highly urbanized
areas.  Hanna and Chang (1991b) contains a table listing typical values of Qf for various cities.

       The sensible heat flux, Qh and latent heat flux are determined by Holtslag and van Ulden
(1983) as:
                                                                                   (4-7)
where P' is an empirical coefficient (-20 W/m2).

       Typical values of a, based on empirical data of Holtslag and van Ulden and summarized
by Hanna and Chang (1991b) are:

       
-------
               0.128 + 0.0051n(z0/z)      zjz * 0.01                               (4-9)

               , 0.107                     zjz > 0.01                               (4-10)
            1.95 + 3Z6(z0/z)°-'a                      •                              (4-11)
       d,-                                                                       (4-12)
The term djn(l + d2d3) represents the correction due to instability, u., = ku/|Tn(z - d)/z0], k is
the von Karman constant (~ 0.4), and d is the displacement height.

       Hanna and Chang (1990, 1992) tested the analytical formula against values produced by
the iterative solution of u. and L. They found that the Wang and Chen (1980) expression
produced values within 10% of the results determined by the iterative solution for z  = 10 m, d
= 0, za  = 1m, and a large value of Qh (400 W/m2).  Better agreement was found for smaller
roughness elements and smaller sensible heat fluxes.  In addition, the analytical solution was
computationally significantly faster.

       The Monin-Obukhov length can then be computed directly from its definition once u, is
determined from Eqn. (4-8) and Qh from Eqn. (4-6).
42    Stable Conditions

       The Weil and Brower (1983) method for estimating u. is applied in HPDM during stable
conditions. A first estimate of the scaling temperature, 9,, is calculated using Hoitslag and Van
Ulden's (1983) equation:

       9n  =  0.09(1 - 0.5N2)                                                         (4-14)

where N is the total fractional cloud cover and 9. has units of TC  Another estimate of 9. is
made from the profile equation for temperature:
                                           4-5

-------
               18.8 z^

where the neutral drag coefficient C^, is defined as k/ln((z - d)/zj.

       Then, 6. is set equal to the smaller of 9n and 9^.
                                                                                  (4-15)
The sensible heat flux, Qto is defined during stable conditions as:
       Qh = - pCpU.8.,
                                                     (4-16)
For large values of u (or u.), 9n (which depends only on cloud cover) is smaller than 9^ but an
additional check on the product u.8. must be made, since Qh does not keep increasing
indefinitely with higher wind speeds.  In HPDM, the value of 9. is not allowed to exceed
0.05/u., where the numerator has units of "K m/s and denominator has units of m/sec.  This
limit is estimated from observations of heat fluxes during high-wind, stable conditions.

       The friction  velocity, u^ can be calculated from:
       u.
1 -
(4-17)
where u0=   (4.7zg8./T)V2 .
Because 8. is set equal to the smaller of 9n and 9^, the following condition is always met:
                                                                                  (4-18)
       During stable conditions, Hanna and Chang (1992) suggest a lower limit on L in
 recognition of the fact that the atmosphere is less stable over urban areas than over rural
 surfaces.  Their suggested values for use for the various land use categories defined in the Auer
 (1978) scheme are shown in Table 4-2.
                                           4-6

-------
                                       Table 4-2
                        Minimum Values of Monin-Obukhov Length
                                During Stable Conditions
                               for Various Land Use Types
                              (From Hanna and Chang, 1992)
Auer (1978)         Class               Description                 Minimum L
Cl                 Commercial         > 40 story buildings         150 m
                                        10-40 story buildings         100 m
                                        < 10 story buildings          50 m
II, 12               Industrial                                       50 m
R3                 Compact Residential                             50 m
Rl, R2              Residential                                      25 m
A                  Agricultural                                      2 m
                                         4-7

-------
5.      Model Evaluation Protocol

        The major objective of the evaluation exercise is the development and implementation of
an objective model evaluation and scoring methodology that allows good-performing and
poor-performing models to be distinguished, so that an appropriate dry deposition technique
representing the current state-of-the-science may be selected. This protocol outlines an
evaluation approach which allows, to the extent possible, an objective model performance
scoring and selection process which results in the selection of a deposition model for gases and
another for. particles. We have adopted the EPA's model statistical approach to model
evaluation as described by Cox and Tikvart (1990).

        The accuracy of a model performance evaluation is  dependent upon the accuracy and
representativeness of both the observations used for comparison and the model input data used
to produce predictions. Estimates of dry deposition rates presently suffer from numerous
sources of uncertainty (Hosker and Lindberg, 1982) leading to the situation where there is a
wide spread of overlap for sets of deposition velocity observations even under similajr
experimental conditions.  As a consequence, it is expected that, given the small size of the
deposition velocity data sets and uncertainties present in the measurements and model input
data, the resulting uncertainties present in both observations and predictions of deposition
velocity will make it difficult in many cases to distinguish between the statistical performance of
some models. An exact quantitative description of the uncertainties is not possible for many of
the data sets due to limited information on the  reported observed data.

5.1    Evaluation" Approach

       The model evaluation focused on a comparison of observed and predicted deposition
velocities.  Since the ISC model uses reflective coefficients instead of deposition velocities, the
effective deposition velocity for the ISC scheme had to be estimated.  Appendix A contains a
description of the method used to convert the reflective coefficients into effective deposition
velocities for use in the model evaluation effort.

       In the present study we combined all the particle deposition data from the various
observational studies, many of which had less than 10 observations, into one large data set of
168 data points that have  sufficient concurrent data for inputs to the deposition models. Several
subsets  of deposition data were created based on one or more stratification criteria.  On each
subset model performance statistics were calculated. These  statistics were used to estimate a
composite measure of performance where the most accurate model produces the smallest

                                           5-1

-------
composite measure based on the EPA's Cox-Tikvart approach utilizing fractional bias.  In
summary, the model evaluation exercise consisted of the following steps:

       1)     Stratification of the observed deposition velocity data into subsets based on
              several criteria such as physical and chemical properties, surface characteristics,
              meteorological conditions, etc

       2)     Direct statistical and graphical comparison of observations versus predictions of
              deposition velocity for each subset using the EPA's Cox-TOcvart protocol (Cox
              and Tikvart, 1990).

       3)     Ranking of model performance using a composite performance measure (CPM)
              and selection of the best performing model based  on rank and uncertainty in the
              CPM.

       The models which were  evaluated and scored were discussed in Section 2. In addition to
core models,  several hybrid models were created for performance evaluation based on our
review of the physics represented by the models, and on the outcome of the sensitivity analysis.
The technical review provided indications of which components of the various core models are
superior, and which may  have some unrealistic characteristics. We created several hybrid
models from  the core models in order to offer the U.S. EPA the  best performing model (instead
of several that perform about the same and which may have one  or more flaws that can create
problems in the context of the wide ranging applications seen  by  ISC). The hybrid models were
exercised in the same manner as the core models and received the  same statistical treatment
and scoring as the core models.

5.2    Stratification of Deposition Velocity Datasets

       The complete data set of observed deposition velocities includes observations for several
chemical species plus particles made over a variety of surfaces and  under a wide range of
meteorological conditions and sampling times. Model performance was expected to vary
significantly between chemicals  or particle sizes, and for different surfaces and meteorological
conditions. As a result we divided the complete data set into  subsets based on the following
criteria:

       1)     physical and chemical characteristics
       2)     surface conditions
       3)     meteorological conditions
                                           5-2

-------
Figure 5-1 illustrates these subdivisions that were made in the data. A primary division is made
depending whether the substance is a particle or a gas since the physical deposition processes
are different for gases than particles. Further subdivision of the data was made in such a
manner that a minimum number of observations was lost to datasets with sample sizes less than
10. The stratification criteria that appear to provide large enough sample sets and which are
informative and nonredundant are the following:

       > Physical characteristics
              a) Particles
                      1)     2 size ranges (diameters less than 0.1 micron and between 0.1 and
                            20 microns),
              b) Gases
                      1)     Chemical species (SO2 and O3 are only two species with sample
                            sizes in the hundreds)

       > Surface Characteristics
              a) Particles and Gases
                      1)     rough/smooth surface based on surface roughness length
                     2)     dense/sparse vegetation density based on leaf area index

       > Meteorological Conditions
              a) Particles and gases
                      1)     day/night based on time of day
                     2)     large/small atmospheric diffusion based on friction velocity
                     3)     hot/cold based on ambient temperature

       Division criteria were selected in order to (1) provide a logical break between two
conditions (e.g., night = 000-600 1800-2400, day =  600-1800) and (2) provide a breakdown of
the data  set into subsets of approximately the same size.

       The evaluation effort of particle deposition has focused on smaller particles (  < 20 [im
diameter) because this size range is where the most significant differences among the deposition
models occurs. For particles above 20 |im diameter, the deposition velocity predictions for all of
the models tend to quickly approach the gravitational, settling velocity.  Appendix D contains
plots of deposition velocities as a function of particle diameter which illustrates this point.
                                            5-3

-------
           PARTICLES
                                   Deposition velocity  dataset
     GASES
   physical     surface  meteorological
characteristics   type     conditions
surface  meteorolog:
 type      condition
     Figure 5-1.  Schematic illustration of division of deposition velocity data into subsets.

                                            5-4

-------
53    Comparison of Predictions and Observations

       For paired comparisons, the performance measures for the previously described subsets
are based on an analysis of deposition velocity residuals either paired in time, paired in space,
or paired in both space and time. Since deposition velocities are dependent on local surface
conditions, pairings only by time would not be appropriate for the present study.  Likewise, due
to (1) limited simultaneous monitoring and (2) variations in sampling time between data sets,
the use of pairing only by space does not seem to be appropriate either. Consequently the focus
of the present study is on fully paired (space and time) data.

       A number of traditional performance measures such as bias, variance, gross variability,
and average absolute residuals suffer from the lack of ability to intercompare across  the
deposition velocity data subsets.  Furthermore, other statistical measures utilized by the EPA
(e.g., Cox and Tikvart, 1990) possess similar information on bias and precision, and are
normalized in a manner that allows  data set intercomparison and composite scoring of model
performance.  We utilized four types of normalized statistical measures including:

       1)     fractional bias of arithmetic averages (FBA)
       2)     fractional bias of arithmetic standard deviations (FBSD)
       3)     fractional bias of robust extreme statistic for the n smallest values (FBSE)
       4)     fractional bias of robust extreme statistic for the n largest values (FBLE)

       These four measures all possess the same metric (e.g^ have the same range bounding
and limits) and consequently are directly intercomparable. The fractional bias of the average
(FBA) is defined as:
                                      1       /
                                                                                  (5-1)

where the average is given by:
                               O * — ^
                                    N *T
                                                                                  (5-2)
The fractional bias of the standard deviation (FBSD) is defined as:

                             FBSD = 2^°"  " ^
                                           5-5

-------
where the standard deviation is given by:
                                                                                   (5-4)
The FBA is useful to quantify the degree of model over or underprediction.  FBSD is useful to
determine if model predictions have more or less variability than observations.

       Deposition velocity minimums are an  important factor in estimating maximum ambient
concentrations, while maximum deposition velocities are important when estimating maximum
exposure (due to deposited material) estimates.  Thus estimating a robust extreme statistic for
both the n set of largest and smallest deposition velocities is important.

       The  robust extreme statistic (RES) for maximum deposition velocities is expected to
follow an exponential tailed distribution like that of concentration. For such  a distribution the
RES is estimated via the relation:

                 O(RES)  = O(n)  + ( O -  O(n) ) log
                                                   I   'i   I
                                                                                   (5-5)
where the average O is over the n - 1 largest values and O(n) is the nth largest value. There is
also a need  to estimate a similar robust statistic for the smallest deposition velocities.
Development of a similar robust statistic for small deposition velocities is difficult since:

       1)     the functional form of the probability density function (pdf)  is relatively unknown
              for the tail of the distribution representing the smallest  deposition velocities,

       2)     considerable uncertainty is present in observations of small deposition velocities
              due to both  uncertainty in measurement and interference by resuspension
              processes, and

       3)     the presence of a lower bound (zero) insures asymmetry in the distribution of
              extremes, meaning the same measure estimated using Eqn. (5-5) for the
              unbounded maximum deposition velocities cannot be used as a lower bound
              measure.

       Because of the problems cited above we elected to use the average  of the smallest 10%
of the deposition velocities. In an effort to avoid a lack of robustness due to small sample size

                                            5-6

-------
we estimated the extreme statistics for only the overall data set and not for the subsets in the
study.

       The fractional bias of the extreme statistic is given by:

                              SS>
                                      (O(RES) -
                                                                                    (5-6)
The FBLE provides an indication of the degree of over- or underprediction in the largest values
while FBSE does the same for the smallest values.

5.4    Scoring Model Performance Using Composite Measures

       A composite performance measure (CPM) is calculated for each model as a weighted
linear combination of the individual absolute fractional bias components. Each of the four
fractional bias measures previously defined are combined as a weighted linear sum.  The
compensatory effect of sign (over- or underprediction) is removed by taking the absolute value
of each bias measure. The resulting sum which serves as a composite fractional bias (CFB) is
defined for both the overall data set
           CFB. = ^\FBAa\ + w2|FBSDj  * ^\FBSEa\  + *>4\FBLEa\
                                                                                  (5-7)
and for each of the stratification subsets. The CFB for kth subdivision is defined as:

                                                                                 (5-8)
                       CFBt =
All measures are assumed to be equally important and the weights serve merely as an average,
e-g.,

                            i  =  2 =  3 ~  4 - -J
                                                                                 (5-9)


The CFBk is averaged over all subsets (assuming each subset is equally important).
                                       u
                             CFB  - - Y  CFBt
                                                                                 (5-10)
                                           5-7

-------
where in the present case there are 12 subsets (M = 12). The average is then combined with
the CFB0 resulting from the overall data in order to produce the composite measure (CPM).
                         CPM » - CFB + i CFBa
                                 2        2
                                                                                (5-11)
       The CPM thus contains information about both overall model performance as well as
performance under specific sets of conditions. The use of the CPM to select an appropriate
model reduces the possibility that model performance biases under specific sets of conditions
(which may be over represented in the complete data) will unduly influence model selection.  A
perfect model will produce a CPM of zero.  The CPM is independent of whether the model
over- or underpredicts. The model with the smallest CPM is selected as a winner if it  can be
shown that the CPM is lower than other models in a statistically significant sense (e.g., different
at the 95% confidence level).

       In the present study we estimate the 95% confidence interval of the CPM for each
model. The confidence interval is used to establish how robust our selection process may be
based on our single estimate of CPM. A model could have the smallest CPM but if it  is not
very different from that of another model, or if the estimate is relatively uncertain, then the
model should not be selected.
                             •
       The difference in the CPM between all pairs of models is estimated in order to
determine if differences between models are significant    The difference between the
performance of one model and another is the model comparison measure (MCM), defined as;

                      MCM&& = CPM(A) - CPM(B)

where,        CPM(A) = Composite performance measure for Model A, and
              CPM(B) = Composite performance measure for Model B

The MCM is used to judge the statistical significance of the apparent superiority of any one
model over another.  If the MCM is not significant from zero at the 95% confidence level then
the two models cannot be said to perform in a significantly different manner.

       The bootstrap resampling technique is used to estimate confidence  intervals on the
various measures described above.  In applying the bootstrap procedure, observed and predicted
data pairs are resampled.  Sampling is done with replacement, so some data points are
represented more than once.  This process is repeated 1000 times so that sufficient samples are
                                           5-8

-------
available to calculate the standard error of each measure. The resampling recognizes the
stratifications made on each of the model input variables as individual blocks. This assures that
each of the 1000 variants of the original dataset retained the same number of samples from each
stratification.  Had we not blocked the data in this way, one of the 1000 variants might, for
example, only consist of a few samples associated with say the largest friction velocity (repeated
many times). The bootstrap resampling method allows the standard deviation, s^ of any
performance measure to be estimated, from which a confidence interval can be calculated. The
actual CFM or MCM for each model or model pair is assumed to have a 95% chance of lying
within the range given as

                    CPM - cs_ < CPMaamal < CPM  + cs
                                      ~*            *"                          (5-13)
where c is a multiplier for a  specific percentage level of confidence.

       The standard error of the estimate is simply the standard deviation of the measure over
all of the bootstrap-generated outcomes. If the measure involves a single comparison, such as
CPM for a single model, then the value of c can be set equal to the student-t parameter.

       Difference measures  such as the MCM require that simultaneous confidence intervals be
found for each  pair of models in order to ensure an adequate confidence level and to protect
against falsely concluding that two models are different The method of Cleveland and McGill
(1984) is used to calculate c.  In this method, c  is found such that for 95 percent of the 1,000
bootstrap i-tuples,
                                                                                  (5-14)

where, Ay  = model comparison difference measure for model pair i,j,
       AJU = model comparison difference measure for model pair i,j and bootstrap
                     replication k, and
       s,j  = standard deviation of all the Aijk values.

       The 95% confidence intervals for CPM and MCM for each model and pair of models is
presented.  The confidence intervals on the MCM are critical to this study since if the range
that the actual MCM may be found in does not include zero, then the two model's performance
statistics can be said to be statistically different.
                                           5-9

-------
       Model selection was based on model rank, and if the model MCM was significantly non
zero for all other models. Models with overlapping confidence intervals (MCM that are not
significant from zero) were grouped. Models clustered within groups were treated as having the
same model performance.

       Summary of Scoring Scheme

       In summary, the steps taken in providing a scoring of each model analyzed are as
follows:
       (1)    For each model calculate the FB's for the observed and predicted deposition
              velocities paired by time and location over all data and for each stratification
              (block). Calculate the CPM with confidence limits and summarize the model
              performance by category using CFB's. TTie smaller the CFB's and the CPM, the
              better the overall performance of the model

       (2)    Rank CPM from smallest to largest  Calculate confidence intervals for each
              CPM to determine if dearly superior or inferior model performance occurs based
              on ranking and confidence interval.

       (3)    Calculate the MCM with confidence intervals for each possible model pairing.
              Rank MCM's and determine which model pairs have a significantly non zero
              difference.
                                          5-10

-------
6.     Model Evaluation Data Bases

       A list of observational data sets of particle deposition used in the evaluation of the
particle deposition models described in Section 2.1 is presented below. .The observational
deposition velocity data are reproduced in Appendix C.

6.1 Particle Data Sets

       (1)  Doran and Horst (1985).  This paper reports on the results of a field experiment
involving dual tracers. Simultaneous releases were conducted of ZnS, a polydisperse aerosol,
and nondepositing SF6.  Deposition velocities of the particles were computed and are
summarized in the paper. The study used the data to compare four different plume-depletion
models.

       (2)  Lorenz and Murphy (1989).  The aerodynamic profile method was used to determine
deposition rates of 1.0 \im diameter particles to a rough vegetated surface. Figures showing
deposition velocities as a function of meteorological parameters are contained in the paper.
Tables summarizing the mean deposition velocity as a function of particle size contain 61 data
points.

       (3)  Nicholson and Davies (1987).  The profile method was used to derive dry deposition
rates of paniculate sulfate over a rural site in England. Approximately 170 observations were
made over a one year period.  Of these, 78 data points remained after application of a set of
stringent quality criteria. Tabulated values of deposition velocities and meteorological
conditions are presented in the report

       (4)  Sehmel (1980), Sehmel and Hodgson (1978). These studies contain the results of
wind tunnel measurements of particles in the size range from 0.03 jim to  29 \im diameter. Low
roughness surfaces were used in the experiments, ranging from smooth brass to gravel (z0 up to
0.6 cm).  Least squared techniques were used to develop a  set of equations fitting the data
points.  Graphs are provided in the papers showing deposition velocities predicted by the least
squared fit equations (though not  the original data points)  as a function of roughness length,
friction velocity, and particle diameter.  The predictions from the Sehmel and Hodgson
regression equations (as coded in  the CARB model) were plotted and intercompared with the
other deposition models.
                                           6-1

-------
       (5) Hicks et aL (1986).  Eight data points for the dry deposition particuiate sulfur were
obtained from this study which also measured gaseous fluxes.  Measured meteorological
parameters included temperature, wind speed, and friction velocity.

       (6) Garland  (1982). This paper summarizes the dry deposition rates of small particles
to grass in field and wind tunnel experiments. Deposition velocities for lead are presented as a
function of particle diameter. Additional data for an oxide of iron is presented. These data
show similar results from measurements in the field and wind tunnel

       (7) Wesefy et al (1982). Eddy correlation measurements were used to estimate the dry
deposition of particuiate sulfur.  Approximately 19 data points are available including
measurements of sensible heat and friction velocity. The data points were gathered over short
grass during drought conditions in eastern Texas.  Summaries of eddy-correlation experiments
over other surfaces are also presented.

       (8) Wesefy et aL, (1983). Observations of submicron particle deposition velocities were
made in a deciduous forest during winter over the course of a week.  A fine particle sensor was
used to measure particles in the 0.01 to  2 micron size range.  Approximately a dozen eddy
correlation estimates of deposition velocities were reported.  Micrometeorological turbulence
data was collected simultaneously with the deposition data.

62    Gas Data Sets

       The observational data sets of gas deposition which have been collected for use in the
future evaluation of gas deposition models are summarized below.

       (1) Meyers and Baldocchi (1988}. Direct eddy correlation measurements of SO2 and O3
fluxes  are summarized for  both well-watered and water-stressed conditions. Deposition
velocities and meteorological data are presented for approximately 31 periods during two
experiments. The study discusses a comparison of the observations to a hierarchy of deposition
models from simple to highly complex.

        (2) McMiilen et aL (1987).  This report contains measurements of dry deposition of SO2
to a forested site in  Germany.  The measurements were made using the eddy correlation
method.  A tabular listing  of the meteorological and deposition data are presented in an
Appendix of the report  The report discusses limitations to the data and conditions under which
the results are considered  usable.  Basically, good results were obtained when SO2
concentrations exceeded about 1 ppb.

                                            6-2

-------
       (3) Hicks et aL (1989).  Data are summarized from an intensive field study of dry
deposition of SO2 to a variety of crops at a site in Pennsylvania.  Conservative quality-assurance
guidelines were used to screen the raw data base and produce about 22 useful data points. The
report contains tabular listings of the meteorological data and deposition results from the study.

       (4) Harrison et aL (1989).  A gradient technique was used to determine deposition
velocities of HNO3 and HC1 over a variety of vegetative surfaces.  Approximately 34 data points
are provided for the deposition velocity along with selected meteorological parameters, including
the surface friction velocity for each  run.

       (5) Meyers et aL (1989).  The deposition velocity of HNO3 over a fully leafed deciduous
forest was estimated using a gradient technique. The observations were compared to the
predictions of a detailed canopy turbulence model Heat flux and friction velocities are provided
for each experiment. Approximately  10 data points are listed in tables in the paper.

       (6) Godowitch (1990). Vertical ozone fluxes were measured from aircraft over several
different land use types, including agricultural crops and forested areas. An analysis was
performed to derive_ ozone deposition velocities for each experiment.  Measurements of selected
meteorological parameters are also provided.

       (7) Wesefy et aL, (1983).  Eddy correlation measurements of ozone fluxes were
performed above a leafless deciduous forest Measurements of fine particles with a diameter of
approximately 0.1 \im were also  performed. Ranges of deposition velocities and meteorological
parameters are tabulated  in the paper.  The results of 19 half-hour averaged measurements are
also provided for both ozone and particuiate sulfur deposition, although the particle data contain
a low signai-to-noise ratio. However, at least the ozone measurements (10 points) appear to be
usable.

       (8) Wesefy et aL, (1978).  An eddy correlation technique was used to determine vertical
fluxes of ozone above both mature and senescent maize canopies.  The results of approximately
26 .runs are tabulated. Both deposition velocities and meteorological measurements (e.g.,
friction velocity, heat flux) are provided.

       (9) Wesefy et aL, (1981).  Measurements of ozone deposition velocity over snow, wet
bare soil, and water were made using the eddy correlation technique.  Over 60 data points are
listed.  Meteorological data during each experiment is provided.  An interesting component of
this study is the importance of evaporation (Stefan flow) in influencing the deposition rate.
                                           6-3

-------
       (10) Hicks et aL, (1986). Eddy correlation measurements of vertical fluxes of gaseous
and paniculate sulfur compounds were conducted over two different surface types. Tables of
meteorological conditions and deposition velocities are provided for 26 different time periods.
Data is also presented for NO, fluxes, but it is highly variable due to a low signal-to-noise ratio
in the NOX measurement system.  It appears that 8 points of SO2 deposition are usable.

       (11) Fowler and Cape (1982).  The eddy correlation method was used to calculate SO2
fluxes over a Scots pine forest during daytime hours. Tables of flux, deposition velocity,  and
sensible heat are presented for 20 points. Dry deposition rates ranged from 0-5 to 10 mm/sec.

       (12) Davies and Mitchell (1982). This paper presents dry deposition rates for SO2 based
on gradient method measurements over grass in rural eastern England during a period of
relatively constant atmospheric conditions.  Twenty-five data points are available including wind
speed, friction velocity,  sensible heat flux, and Monin-Obukhov length.

       (13) Huebert (1982).  A modified Bowen Ratio method was used to measure nitric acid
fluxes in this study. Measurements were made over a pasture near Champaign, Illinois.
Meteorological parameters measured include wind speed and friction velocity.  The results of
die study showed that dry deposition is capable of depositing HNO3 at a rate comparable to that
of wet deposition measured in previous experiments.

       (14) Padro et aL (1991). In this paper, the ADOM dry deposition model was compared
to observed ozone deposition velocities taken over a fully leafed mixed deciduous  forest  during
July and August of 1988.  Meteorological data collected during the study include wind  speed,
wind direction, temperature,  dew point, solar radiation, net radiation, friction velocity, sensible
heat flux, latent heat flux, and stability. Dr. Padro has kindly supplied us with  the data from this
experiment on a floppy diskette.
                                            6-4

-------
7.     Results of Model Evaluation

7.1    Particle Deposition Models

       For identification purposes the model designations listed in Table 7-1 identifies each of
the ten models and describes the differences between the models. The modifications adopted
for the hybrid models are summarized in the table.

       The particle data set consists of 168 cases in total where non-zero observations of
deposition velocity were made and for which sufficient concurrent meteorological data exists to
exercise each deposition model Assumptions needed to be made regarding the size
distributions of the particles that were measured during some experiments where only an
average size or a range of particle sizes were reported.  For some particles that are formed in
situ such as sulfate particles there may be a reason for selecting an a priori distribution.  For
primary particle sources there may not be a  reason for selecting any particular distribution.  In
our initial analyses we have assumed a uniform particle size distribution.  Later we introduce a
particle size distribution that is peaked in the 'accumulation' size range centered around 0.2
microns in order to see if model performance characteristics would change for 'aged' particle
distributions. A total of twelve data subsets  were developed in order to learn  if model
performance biases and imprecision appeared systematically in the data.  Table 7-2 summarizes
the stratifications made, along with the stratification criteria.  For each of the  stratifications
there are greater than 10 cases,  sufficient for minimal performance statistics.  In each of the
following sections we discuss the results for each stratification.

       The resulting composite  performance measures depend on several factors including:

       •      particle size distribution assumed
       •      random numbers used  and number of replications
       •      number of sample points in extremes
       •      treatment of zero and negative observed deposition velocity

In the present study we present  model performance and selection results for both a uniformly
distributed particle distribution and a  more realistic peaked particle distribution for sulfate
particle observations. In the present study we use 1000 replications for bootstrap estimation of
confidence intervals. Tests show little sensitivity to sources of random numbers. The extremes
are represented by a sample of the 11 largest velocities  and 17 smallest velocities (lowest 10
percentile). Due to issues of robustness related to sufficient sample size the RES was estimated
for only the overall data set (168 cases). Initially all zero  or negative observations were thrown
                                            7-1

-------
                                    Table 7-1
                        A Summary of the Model Designations
Model Name
Description of Model
CARB1*
CARBO
CARB2
CARB3
ADOM 1*
ADOM2
ADOM 3

UAM1*
UAM2

ISC*
Unmodified GARB model
Removal of CARS roughness length restriction of z0 < 10 cm
Same as GARB 1 except uses fixed temperature (T = 75 degrees F)
Uses fixed temperature (as in GARB 2), a LAI adjustment to I3, and
      a constant reference length of OJ cm in I3

ADOM dry deposition algorithm
Modified Stokes/Schmidt relations and phoretic effects term
Changes made to ADOM 2 + LAI adjustment to rd

Unmodified UAM-V model
Contains LAI adjustment to rd

ISC model with boundary layer estimator for H,./x
* Core (unmodified) models
                                       7-2

-------
                                 •Table 7-2
A Summary of the Stratifications Made To the Small Particle Data Set (N = 168)
Stratifying Variable
Particle Diameter (microns)
Roughness Length (m)
Leaf Area Index
Day /Night Insolation
Friction Velocity (m/s)
Temperature (deg K)
Definition
Threshold(s)
0.1 < d < 20.0
sO.25
> 0.25
> 3.0
night
day
> 0.25
* 290.0
> 290.0
Threshold
Selection
non inertia!
inertia!
simple canopy
complex canopy
nonforest
forest
-
sample size
sample size
sample size
sample size
Sample Size
13
155
97
71
105
63
42
51
58
110
62
106
                                   7-3

-------
 out since relative measurement uncertainty is large for small deposition velocities, and factors
 like resuspension could become important We examined the potential bias caused by dropping
 such  data by assigning a minimum velocity of 0.005 cm/s.

-7.1.1    Full Data Sets

        One of the first steps in the model performance evaluation exercise was to examine the
 scatter plots of the observed versus the predicted deposition velocities. A complete set of
 scatter plots for each of the ten models is presented in Appendix B as Figures B-la through
 B-lj.  From these scatter plots we can note that the observed deposition  velocity ranges through
 nearly three orders of magnitude. Most of the models with the exception of ISC also reproduce
 the wide range of variation. There appears to be a tendency for the original model algorithms
 (GARB 1, ADOM 1, UAM 1, and ISC) to underpredict the deposition velocity. The scatter
 plots also indicate that there is a very wide range of scatter through most ranges of observed
 deposition velocity. The major exception that can be noted is a tendency for the scatter to
 decrease with increasing deposition velocity, suggesting that the fractional (percentage) error
 decreases with increasing deposition velocity.

        The scatter plots indicate that there appears to be a significant decrease in average bias
 in the hybrid model results compared with the original model formulations.  For example,
 comparing B-la with B-ld, or B-le with B-lg, or B-lh with B-li we find  that several groups of
 observation which were originally underpredicted in a significant manner, are predicted
 significantly better by the hybrid model formulation.

        Several of the models such as ADOM 1 (Figure B-le) exhibit a tendency to underpredict
 in the mid-range of observed deposition velocity.  The  underpredictions are sometimes by as
 much as an order of magnitude.  The range of observed deposition velocities extend over a
 factor of several hundred and are plotted  on a logarithmic scale which visually understates the
 contribution of the largest deposition velocities.  Caution should be exercised in interpreting the
 scatter plots since some models do rather well at either extreme (or both) and  the various
 performance measures which.are designed to describe  a specific characteristic of performance
 can produce a conflicting picture of performance which can reduce the discriminating power of
 the composite measure.

        A purely qualitative survey of the  scatter plots in Figures B-la through  B-lj suggests that
 models such UAM 2 (Figure B-li) and CARB 3 (Figure B-ld) appear to have  similar scatter
 over all ranges of observed deposition velocity and appear to do the best job of predicting. Of
 these two models the UAM 2 appears to  be better at predicting the very smallest observed
 deposition velocities. The ADOM 1 model appears to perform well at the extremes, but
 considerably less well in the mid-range. The hybrid model ADOM 2 does not significantly

                                            7-4

-------
improve the mid-range undeiprediction bias and in fact causes the predicted range of deposition
velocities to shrink by causing small deposition velocities to be overpredicted and large
deposition velocities to be underpredicted.

       The ISC model exhibits the most unique scatter plot signature in Figure B-lj.  Due to
the fact that small particles with a settling velocity of less than 0.1 cm/s are assumed to have a
reflection coefficient equal to one, a majority of particles have essentially a zero deposition
velocity (small values in Figure B-lj are lower bounds that are set so that the deposition velocity
does not identically equal zero).  The non-zero values of the predicted deposition velocity in
Figure B-lj are strongly underpredicted due to the fact that the deposition velocity components
are multiplied by  one minus the reflection factor,  ct, which is generally smaller than 0.1.

       All of the  fractional and several composite statistical performance measures are
summarized in Table 7-3.  The UAM 2 model (with GARB 3  a distant second) appears to
predict the average deposition velocity the best The bias in the standard deviation is best
predicted by the ADOM 1 model, with the UAM 2 model a close second. The GARB 3 model's
overpredictions of the largest observed deposition velocities offsets its good performance in
other areas. These overpredictions made by CARB 3 appear to be due to the LAI modification
of the deposition  velocities as evidence by an improvement in performance when LAI is not
considered as in the case of CARB 2 in Table 7-3. The ADOM 2 model appears to most
accurately predict the smallest deposition velocities while the ADOM 1 model shows a tendency
to overpredict them. The LAI modification of the deposition velocities produces overpredictions
of the smallest deposition velocities for both the CARB 3 and ADOM 3 models. The largest
deposition velocities appear to be most accurately predicted by the ADOM  1 model with the
UAM 2 model coming in second.

       The composite fractional bias for the 168 cases is shown in Table 7-3. The best
performing models in order of rank are UAM 2, CARB 3, and ADOM 1. The UAM 2 model
appears to have the most well rounded performance since the fractional bias was at or nearly
the smallest bias for all of the four bias measures. The ADOM 1  model on the other hand was
helped significantly by the fact that what it lost in performance by overpredicting the smallest
deposition velocities, it gained back by producing  an appropriate estimate of the variation in
deposition velocity.  In all cases the ISC model performed significantly poorer than all other
models and so was ranked last in terms of performance.

       The composite measure of fractional bias for each of the subsets was computed using
only the fractional bias of the average and the standard deviation. The average of the
composite factional bias was averaged over all twelve subsets and the complete data set and is

                                           7-5

-------
                                 Table 7-3

A summary of the fractional and composite statistical measures for each of the 10 models
examined.  The   is the composite  fractional  bias measure estimated  as  the
fractional bias of the average and standard deviation averaged over all 12 subsets.  The
overall sample size is 168.  The largest extreme sample  size is 11; the smallest extreme
sample size is 17. A uniform particle size
Model Mane FBA FBSD FBSE
CARS 0
CARB 1
CARS 2
CARB 3
ADOM 1
ADON 2
ADOM 3
UAH 1
UAH 2
ISC
0.991
1.070
0.976
0.800
1.005
1.146
0.895
0.992
0.609
1.854
0.852
0.846
0.764
0.815
0.448
0.944
0.886
0.604
0.539
1.719
0.815
0.815
0.366
-0.276
0.999
0.071
-0.892
0.784
0.383
1.976
distribution is assumed.
FBLE  CFB.
1.090
1.090
1.013
0.907
0.494
1.094
0.987
0.858
0.686
1.773
1.137
1.031
0.991
0.986
1.023
1.160
0.937
1.121
0.829
1.854
0.937
0.955
0.780
0.700
0.736
0.814
0.915
0.810
0.554
1.831
                                        7-6

-------
 presented in Table 7-3 as weJL This composite measure indicates that UAM 2, GARB 3, and
 ADOM 3 have the best average performance as measured by the smallness of the average
 composite.

        Figure 7-1 shows the fractional biases present in the data taken as a single data set.  The
 results indicate that UAM 2 has the smallest bias in the mean and standard deviation.
 According to FBSE, the smallest deposition velocities appear to be best predicted by the ADOM
 2 as noted earlier in Table 7-3. According to the FBLE the largest deposition velocities are
 predicted best by the ADOM  1 model. The confidence intervals on  these measures for the
 entire data set are rather large, sometimes being as large as 50% of the measure.  In order to
 reduce  this wide error bar an average was taken over all data stratifications, reducing the
 confidence interval by over a factor of 3. The resulting averaged composite measures are
 graphically summarized in Figure 7-2.  This figure indicates UAM 2  is the best performer for
 average bias and GARB 0 and ADOM 3 are nearly tied runners up.  The ISC model is clearly
 the poorest  performing model out of the 10 models.

       The  scatter plots of the fractional measures of model performance for the overall data
 set are presented in Figures 7-3a through 7-3d. A co-plot of the fractional bias in the average
 and standard deviations is shown in Figure 7-3a. This plot indicates  that model performance for
 all of the models except UAM 2 and ISC are clumped outside the factor of two box in the upper
 right hand quadrant where underpredictions of both the average and standard deviation occur.
 The UAM2 model also underpredicts both measures and lies just inside the factor of two box.
 The ISC model is an outlier and underpredicts both measures in an extreme sense.  The co-plot
 of the fractional bias in the average of the entire data set versus that of the 11 largest deposition
 velocities  in Figure 7-3b shows a similar pattern as that of Figure 7-3a. When the average bias
 of the RSE of the 17 smallest deposition velocities is co-plotted with the fractional bias of the
 average for the entire data set (Figure 7-3c) the UAM 2 and CARB  2 and 3 models appear to
 best predict small deposition velocities. The ADOM 3 model significantly overpredicts while
 ISC significantly underpredicts. The co-plot of the fractional bias in  the 17 smallest and 11
 largest deposition velocities (Figure 7-3d), indicates that only UAM 2 prediction of deposition
 velocities fall within a factor of 2 for both extremes.

 7.1.2   Stratification by Particle Diameter

       The stratification by particle diameter is important in order to see which models perform
 best for intermediate sized particles that behave neither as rapidly sedimenting particles or as a
gas.  Thus we will be evaluating model performance for very small particles ( < 0.1 microns)
versus intermediate size particles of size 0.1 through 20  microns.

                                           7-7

-------
00
           Absolute Fractional Bias
        2.5
        1.5  -
        0,5
         0
            CARB 0 CARB  1  CARB 2  CARB 3  ADOM 1 ADOM 2 ADOM 3  UAM  1  UAM 2   ISC
                                    Deposition Model
                           Fba
Fbsd
Fbse
Fble
    n = 160
              Figure 7-1. A summary of the absolute value of the fractional bias using all data.

-------
VO
              Absolute  Fractional Bias
          1.5
          0.5
            0
                       I
                              t
I
a.

                             p
                             I
               UAM 2  CARB 3 ADOM 1  CARB 2  ADOM 3 CARB 0  UAM  1  ADOM 2 CARB 1    ISC
                                       Deposition  Model
                                           Fba
                          Fbsd
                 Figure 7-2. A summary of the absolute value of the fractional bias averaged over all 12 subsets.

-------
z.u
C 1.5
0
tandard Deviat
o »-*
en b
CO
5 o.o
4-)
'o
CO
cd -0.5
ffl
*•»*
od
.0 -1.0
o
ca
^-1.5
-2.Q
i i i
••
•i*

M
H»
M*

••»
—

-
-
^
1 l_ 1
i i i










i i i
i i


















1 1 !
1 1 1












1 1 1
1 1 1












1 1 1
1 1 1



ffl?






A^

-------
7)
0)
£
 0)
 5iO
 J-,
 CO
«*-!
 O

 M
 cfl
 CO
 C
 o
2.U
1C
.5
1f\
.0
OS
.0
Of\
.0
Oc
.D

1 K.
— i ..o
2r\
_
.Ml



—
—
—
— —
-
^
1 1 1
2.0 -:
111 1 i i









i i i
1.5 -
I I









i i
1.0
1









I
— (
1 1 1









1 1 !
).5 0
I 1 1









i i i
.0 0
"T









1
.5
TT~


A
^
\





1 I
1


3J>
7
f






.0 1
1 1 1
OQ-
-

-

-




I i
.5 2











.G
Key
QCARR 1
O CARB 2
A CARB 3
-f- CARB 0
ADOM 1
O ADOM 2
>K ADOM 3
V UAM 1
© UAM 2
Q ISC






                 Fractional  Bias of  the  Average
      Figure 7-3b.   Co-plot of fractional bias of the 11 largest deposition velocities for each of the

                    deposition models.  The sample set includes all data points.  The box indicates

                    the region within which the predictions are within a factor of 2 of observations.
                                                7-11

-------
CO

1
0)
-t-J
X
CO

'd
s
 CO
 cd
•*•*
ffl
r«-i
 ca
 fi
 o
 o
 cd
     2.0
     1.5
     1.0
     0.5
«M   0.0
   -0.5
   -1.0
      1.5
i 1 1
•urn
-
••B
^_
.•»
^^
1 1 1
i i i






i i i
i i i












! 1 1
1 1 1








1 1 1
1 1 1








! 1 !
1 I I



.



$

c
A

*
1 1 1
1 1 1

<
-'
1


1 1 1
I 1 £2)
^™
—
—
_
^~

1 1 1
2.0 -1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 2
.C
Key
CD CARB 1
CD CARB 2
A CARB 3
+ CARB 0
X ADOM 1
O ADOM 2
* ADOM 3
V UAM 1
© UAM 2
Q ISC
)
                 Fractional  Bias  of the  Average
       Figure 7-3c.   Co-plot of fractional bias of the 17 smallest deposition velocities for each of the

                    deposition models.  The sample set includes all data points.  The box indicates

                    the region within which predictions are within a factor of 2 of observations.
                                               7-12

-------
w
X
W

CO
^
13
S
CO

"o
CO

-------
       The results of the performance evaluation are quantitatively summarized by" Table 7-4.
The most notable points to be drawn from this table are the following:

       1)     The CFB suggest that the ADOM 2, ADOM 3 and GARB 0 models are the most
              accurate models for small particles however the GARB model variants do
              perform poorer for intermediate sized particles due to significant
              underpredictions.

       2)     The overall most accurate model with the smallest CFB for larger particles seems
              to be the UAM 2 and ADOM 1 models which seem to treat larger particles
              considerably better than small ones.

       3)     All measures indicate that the ISC model consistently performs the worst of all of
              the models with significant underpredictions of the deposition velocity.

Table 7-4 indicates that no single model does well for both small and larger particles. The
ADOM 3 model appears  to do best for small particle diameters, while UAM 2 does the best for
larger particles.  Both models provide significantly small fractional biases for both the average
and standard deviation. All models underpredict all measures of bias. The ISC model cannot
predict deposition for particles smaller than  0.1 microns since the deposition velocity is basically
set equal to zero by setting the reflection coefficient equal to 1.  This effect results in the
extremely large CFB in Table 7-4 for the ISC model.

7.13   Stratification by Roughness Length

       The roughness length figures prominently in most of the deposition models undergoing
evaluation. For small roughness lengths where a complex canopy is not present the models
would be expected to perform at their best  The fractional and composite  performance statistics
which are presented in Table  7-5 indicate the following points:

       1)     According to the CFB, the UAM 2 model performs in a superior manner
              regardless  of the underlying surface.  The model produces underpredictions on
              the average.

       2)     The statistical measures indicate that all models perform poorer over rough
              surfaces than smooth ones as might be expected.
                                          7-14

-------
Table 7-4
The first row is for particle diameters less than 0.1 microns, the second
0.1 to 20 microns. The rank is by CFB and the particle distribution is
Model Name Sanroles PBA FBSD
GARB 0

GARB 1

GARB 2

GARB 3

ADOM 1

ADOM 2

ADOM 3

UAM 1

UAM 2

ISC

13
155
13
155
13
155
13
155
13
155
13
155
13
155
13
155
13
155
13
155
0.100
1.042
0.899
1.077
0.641
0.991
1.186
0.787
0.971
1.006
0.925
1.156
0.443
0.916
1.673
0.970
1.500
0.583
1.998
1.849
1.435
0.846
1.827
0.839
1.669
0.756
1.591
0.815
1.622
0.439
0.856
0.942
0.232
0.895
1.895
0.601
1.761
0.542
2.000
1.717
is for particles in the range
for a uniform distribution.
CFB. Rank
0.768
0.944
1.363
0.958
1.155
0.874
1.388
0.801
1.297
0.723
0.891
1.049
0.337
0.905
1.784
0.786
1.630
0.562
1.999
1.783
2
7
6
8
4
5
7
4
5
2
3
9
1
6
9
3
8
1
10
10
  7-15

-------
                                       Table 7-5

A summary of the fractional and composite statistical measures for each of the models examined.
The first row is for roughness lengths less than 0.25 m, the second is for roughness lengths greater
than 0.25 m. The rank is by CFB and the particle size distribution is for a uniform distribution.


Model Name        Samples	FBA	FBSD	CFB..	Rank
GARB 0
GARB 1
GARB 2
GARB 3
ADOM 1
ADOM 2
ADOM 3
UAM 1
UAM 2
ISC
97
71
97
71
97
71
97
71
97
71
97
71
97
71
97
71
97
71
97
71
0.902
1.196
0.902
1.490
0.813
1.385
0.800
0.800
0.732
1.795
0.916
1.766
• 0.687
1.453
0.725
1.761
0.571,
0.691
1.789
1.999
0.846
1.263
0.846
1.765
0.763
1.631
0.793
1.356
0.446
1.614
0.970
1.747
0.925
1.700
0.616
1.920
0.521
1.068
1.720
2.000
0.874
1.229
0.874
1.628
0.788
1.508
0.796
1.078
0.589
1.705
0.943
1.756
0.806
1.577
0.671
1.840
0.546
0.879
1.754
1.999
3
3
7
6
4
4
5
2
2
7
9
8
6
5
3
9
1
1
10
10
                                          7-16

-------
       3)     All measures indicate that the ISC model consistently performs poorest of all
              models, regardless of surface, and always produces significant underpredictions as
              noted from the large positive FB's.

The average deposition velocity for both rough and smooth surfaces is predicted best by UAM
2. The UAM 2 also predicts the variance of the deposition velocities under rough surfaces the
best, while ADOM 1 does this best for smooth surfaces. All models underpredict all measures.

7.1.4  Stratification by Leaf Area Index

       Several hybrid models possess an explicit dependence of deposition velocity on Leaf
Area Index (LAI). The LAI for each observation was separated into complex surfaces  (e.g.,
forests) where the LAI is 3 or greater and simple surface ceases where the LAI is less than 3
(e.g., grass). If LAI represents an appropriate increase in collection area, then adding a
adjustment for LAI should, in principle, improve model predictions. The normalized
performance measures summarized in Table 7-6.

From Table 7-6 we can note the following:

       1)     The UAM 2 model produces superior composite performance according to the
              CFB regardless of LAI.
                   •
       2)     The CFB indicates that all models tend to perform poorer under
              large LAI situations with a consequent increase in the average
              magnitude of model residuals (e.g., FBA). -

       3)     The ISC model consistently performs the worst of all of the models regardless of
              LAI

Both the average and standard deviations of the deposition velocity distribution are
underpredicted by  all models. The UAM and ADOM family of models exhibit an improvement
for both small  and  large LAI samples. The results for the CARB family is mixed, with there
being no  improvement for small LAI samples, but for large LAI samples the improvement is
dramatic.
                                          7-17

-------
Table 7-6
The first row is for leaf area index (LAI) less than 3.0, the second is for LAI greater than 3.0. The
rank is by CFB and the particle size distribution is for a uniform distribution-
Model Name Samel es FBA FBSD CFB,. Rank
GARB 0
GARB 1
GARB 2
GARB 3
ADOM 1
ADOM 2
ADOM 3 '
UAM 1
UAM 2
ISC
105
63
105
63
105
63
105
63
105
63
105
63
105
63
105
63
105
63
105
63
0.844
1.422
0.902
1.572
0.304
1.499
0.326
0.738
0.743
1.915
0.935
1.817
0.711
1.469
0.760
1.765
0.609
0.610
1.798
1.999
0.856
1.800
0.350
1.867
0.769
1.847
0.790
1.552
0.449
1.979
0.963
1.861
0.918
1.616
0.612
1.912
0.515
1.237
1.719
2.000
0.850
1.611
0.876
1.720
0.787
1.673
0.308
1.145
0.596
1.947
0.951
1.839
0.814
1.543
0.686
1.838
0.562
0.923
1.759
1.999
7
4
3
6
4
5
5
2
2
9
9
8
6
3
3
7
1
1
10
10
   7-18

-------
7.1.5   Stratification by Day vs Night

       The particle deposition velocity is dependent on the degree of atmospheric turbulence
which in turn is dependent on the atmospheric stability. Atmospheric stability generally
undergoes a significant diurnal variation. At night turbulent transport is generally conducted
under neutral or stable conditions. Any day-night difference in performance is likely to be
directly connected with the aerodynamic resistance formulation utilized. From Table 7-7 which
summarizes the normalized and composite performance statistics, the following points can be
noted:

       1)     The CFB indicates that during the night the ADOM 1 model is the best
              performing model while during the day the ADOM 3  model is best.

       2)     Models tend to perform better during the night than during the day.  During the
              day even typically good performing models such as UAM 2 perform markedly
              poorer.

       The fractional bias measures compiled in Table 7-7 suggests that ADOM 1 does quite
well for night samples for both the average and the standard deviation. During the day the
GARB 0 model has the  smallest bias in the standard deviation, while the ADOM 3 model has
the smallest bias in the average.  The ISC model performed the worst, and during the day
essentially showed no predictive skill

7.1.6   Stratification by Friction Velocity

       Friction velocity is related directly to the vertical turbulent transport of momentum.  In
addition, the friction velocity plays a role in determining the laminar boundary layer  near the
surface.  Consequently the friction velocity is a relatively important determinant of deposition
velocity.  We have stratified the small particle cases into high  and low friction velocity sets with
a threshold set  to divide the sample into halves. The model performance statistics are presented
in Table 7-8.  From this table we can make the following observations;

       1)     Based on CFB the  UAM 2 model is the best performing model
              regardless of friction velocity while ISC is the worst

       2)     CARB 3 is the next best performing model with its best
              performance occurring under low friction velocity conditions.
                                           7-19

-------
                                     Table 7-7

A summary of the fractional and composite statistical measures for each of the models examined.
The first row is for night, the second is for day.  The rank is by CFB and the particle size
distribution is for a uniform distribution.
Model  Name	Samples	FBA	FBSD	  CFB,         Rank
GARB 0
GARB 1
GARB 2
GARB 3
ADOM 1
ADOM 2
ADOM 3
UAM 1
UAM 2
ISC
42
51
42
51
42
51
42
51
42
51
42
51
42
51
42
51
42
51
42
51
0.987
1.001
0.987
1.380
0.893
1.243
1.014
1.272
0.665
1.558
0.946
1.069
0.853
0.481
0.787
1.739
0.741
1.493
1.766
1.998
0.994
0.703
0.994
1.410
0.904
1.207
1.087
1.717
0.493
1.396
1.021
1.648
1.041
1.067
0.779
1.876
0.759
1.729
1.737
2.000
	 =*^c 	
0.991
0.852
0.991
1.395
0.899
1.225
1.051
1.495
0.579
1.477
0.984
1.358
0.947
0.774
0.783
1.808
0.750
1.611
1.752
1.999
8
2
7
5
4
3
9
7
1
6
6
4
5
1
3
9
2
8
10
10
                                        7-20

-------
Table 7-8
The first row is for friction velocity less than 0.25 m/s, the second is
than OJ25 m/s. The rank is by CFB and the particle size distribution is
Model Name Samoles FBA FBSD
GARB 0

GARB 1

CARS 2

GARB 3

ADOM 1

ADOM 2

ADOM 3

UAM 1

UAM 2

ISC

58
110
58
110
58
110
58
110
58
110
58
110
58
110
58
110
58
110
58
110
1.116
0.966
1.135
1.056
1.036
0.964
0.806
0.799
1.522
0.914
1.103
1.155
0.770
0.923
1.151
0.960
0.801
0.572
1.933
1.838
0.773
0.859
0.772
0.843
0.670
0.764
0,769
0.821
1.047
0.406
1.001
0.925
1.011
0.867
0.604
0.590
0.511
0.550
1.798
1.709
for friction velocity greatei
for a uniform distribution.
CFB. Rank
0.944
0.913
0.954
0.949
0.853
0.864
0.787
0.810
1.285
0.660
1.052
1.040
0.890
0.895
0.878
0.775
0.656
0.561
1.865
1.774
6
7
7
8
3
5
2
4
9
2
3
9
5
6
4
3
1
1
10
10
  7-21

-------
Of the core models ADOM 1 performs rather well under large friction velocity conditions, but is
the next to worst performer under small friction velocity conditions. The UAM 2 fractional bias
for both the average and standard deviation was the smallest of all models evaluated.

7.1.7   Stratification by Temperature-

       The dependence of deposition velocity model performance on temperature was
examined. The observed data was broken up into 'hot' and 'cold' subsets based on a 17° C
threshold which was applied to split the overall data set up into two large subsets.  While most
particle deposition algorithm do not have an explicit temperature dependence, the original
GARB formulation (GARB 1) does.  The resulting model normalized performance statistics and
performance scores are summarized in Table 7-9.  The resulting fractional bias and composite
performance measures indicate that:

       1)      The CFB indicates that the UAM 2 model is the best performer under warm
              temperatures and is the second best performer under cool temperatures. The
              model always underpredicted  the observed deposition velocities.

       2)      The CFB indicates that the ADOM 1 model is the best performer under cool
              temperatures, while the GARB 3 model is the second best performer under warm
              temperatures.

       3)      All measures consistently show the  ISC model as the worst performing model
              regardless of temperature.

Under warm temperatures the UAM 2 shows significantly smaller fractional bias measures for
both the average and the standard deviation. Under cool temperatures the ADOM 1  model
produces the smallest fractional biases for both the average and the standard deviation.

       The results of the findings or each model and  for each subset is summarized by the
model specific bar chart of CFB in Figure 7-4.  This figure shows the composite fractional bias
averaged over the high  low categories. This  figure indicates that the UAM 2 is the best
performing model (smallest CFB) over.many of the stratifications while the ISC model is the
worst performed over all stratifications.  The runners  up for best performance are the ADOM 3
and the CARB 3 models.  The performance  of these two models alternate in ranking from
subset to subset. For example from Figure 7-4 for the two stratifications that UAM 2 does
poorly on, namely the particle size and day/night stratifications, the best performing model was
ADOM 3 in both cases.
                                          7-22

-------
                                      Table 7-9

A summary of the fractional and composite statistical measures for each of the models examined
The first row is for temperatures less than 290.0 deg K, the second is for temperatures greater than
290.0 deg K. The rank is by CFB and the particle size distribution is for a uniform distribution.


Model  Name	Samples	FBA	FBSD	CFB..	Rank
GARB 0
CARB 1
GARB 2
CARB 3
ADOM 1
ADOM 2
ADOM 3
UAM 1
UAM 2
ISC
62
106
62
106
62
106
62
106
. 62
106
62
106
62
106
62
106
62
106
62
106
0.987
0.994
1.115
1.045
0.972
0.979
1.091
0.659
0.811
1.127
0.972
1.253
0.796
0.953
0.999
0.988
0.932
0.457
1.820
1.874
1.064
0.661
1.067
0.648
0.952
0.587
1.116
0.595
0.481
0.411
1.014
0.870
1.012
0.757
0.817
0.410
0.793
0.340
1.737
1.697
	 	 K 	
1.025
0.827
1.091
0.847
0.962
0.783
1.104
0.627
0.646
0.769
0.993
1.062
0.904
0.855
0.908
0.699
0.863
0.398
1.779
1.786
7
6
8
7
5
5
9
2
1
4
6
9
3
3
4
3
2
1
10
10
                                        7-23

-------
  Composite Fractional Bias
3 -
0
   2g =
   1 =
1  li
•*•  •a =
   CARB  1  CARB 2 CARB 3 CARB 0 ADOM 1  ADOM 2  ADOM 3   UAM 1   UAM 2    ISC


                              Deposition Model
       Particle Size


       Day/Night
Roughness


Friction Velocity
      Leaf Area Index

,riTY,...,,,;..

P:: •'' 1  Temperature
   Figure 7-4. A summary of the total CFB for the six types of data subsets. The total represents a

   sum over both high and low categories for each subdividing variable.

-------
7.1.8   Estimation of CPM from Tables

       The estimate of a CPM can be conducted from the statistics presented in the tables.
The CFB's for CARB 1 will serve as an illustrative example. Table 7-10 illustrates how one can
extract fractional bias information from the tables to estimate the CPM.

72    Discussion of Model Performance

       The composite performance measure (CPM) defined by Equation (5-11) was used to
rank the models.  The model with the smallest CPM is ranked highest  The 95 percent
confidence interval indicates how much the estimated CPM might vary if measurements and
predictions of deposition velocity were repeated under identical meteorological and site
conditions. Model performance was determined using the reported size distribution except for
the experiments involving sulfates where size distribution were not reported.  In those cases, two
different particle distributions assumed for those data sets involving sulfate: a predicted
distribution with most of the mass between 0.16 to 0.29 \im diameter (Richards et aL, 1989) and
a second distribution with the sulfate uniformly distributed between 0.1  -1.0 um diameter. In
the following sections the results of the statistical analysis is summarized.

7.2.1   Uniform Size Distribution

       Figure 7-5 illustrates the top to bottom ranking for the 10 models assuming a uniform
size distribution.  With the exception of ISC all models fall within a narrow range of CPM.  The
three top ranked models are UAM 2, CARB 3, and ADOM 1. The first two models represent a
hybrid variant of core models'with an LAI adjustment The only core model in the top three
ranked models is the ADOM 1 model The confidence intervals suggest that none of the three
models has any obvious performance advantages.  The other core models themselves appear to
have essentially the same composite performance.  The only exception is ISC which appears as
an outlier  with a confidence interval that is narrowed by the many zero  predictions it produces.

       The overlap of the confidence intervals on  the CPM in Figure 7-5 suggests that most
models have performance indistinguishable from their ranked neighbors. The model comparison
measure (MCM) defined by Equation 5-12  is an appropriate measure to compare one model
versus another.  If the difference is not significant  from zero at the 95% confidence level the
two models can be said to be statistically identical Figure 7-6 ranks the MCM for all unique
model pairings.  The MCM's reveal that over 50% of the MCM confidence intervals cross zero
thereby indicating a lack of significant difference in performance. There is no significant
difference among  UAM 2, CARB 3, and ADOM 1 regardless of pairing. There are significant

                                          7-25

-------
                            Table 7-10
Summary of Composite Statistical Measures that Illustrate how the CFM
                   Arises for the CARB 1 Model
   Stratification
 Source
        CFB
    Particle Size
       small
       large
 Roughness  Length
       < 25
       > 25
  Leaf Area Index
        < 3
        > 3
      Sunlight
       night
        day
  Friction Velocity
     < 25cm/s
  •   > 25'cm/s
    Temperature
      < 290°K
      > 290°K
      Average
       CPM0
       CPM
Table
Table 7-5
Table 7-6
Table 7-7
Table 7-8
Table 7-9
Table 7-3
        1.01

        1.01
        1.53

        0.99
        1.56

        125
        1.41

        0.99
        1.00

        1.06
        0.93
        1.17
        0.99
1.08 (Table 7-3
                                   1.08)
                                7-26

-------
    Deposition  Model
UAM 2
CARB 3
ADOM 1
CARB 2
ADOM 3
CARB 0
UAM 1
ADOM 2
CARB 1
ISC
-
-
-
-
-
-
-
-
-
-
* +- *
* -f *
\|^ _|_ S|X
xpt ™|"T /y^
¥ + ^
* +
\j/ |
yyt ~|
Nj/ 1
/T\ ~|
^x 1
*




X
^
K
*
*
*
4- *



•






*
                               0           0.5            1           1.5
                                     Composite Performance Measure  (CPM)
95% low
+   Mean
                                                                       95% high
Figure 7-5.  A ranking of the models by CPM estimated from Equation 5-11. The smallest CPM represents the best performance.
          The CPM is for a uniform particle size distribution.

-------
    Model Names
£
oo
VAU I-CAU a
AMU a-CAU a
UAtI I-ADOU 1
A>au a-CAU a
AMU I-CAU a
AMU I-CAU a
AMU a-CAU a
VAU i-AMy a
Aaoy a-AMy i
CAU a-CAU a
AMU a-CAU 1
VAU I-CAU a
CAU a-CAU a
VAU I-CAU 1
AMU a-CAU a
AMU a-CAU a
AMU I-CAU a
CAU a-CAU i
VAU I-ADOU |
AOOU a-ABou i
ABOU a-CAU 1
UAU I-CAU a
CAAA a— CAU i
AMU a-CAu a
VAU a-CAU a
ABOU I-CAU I
UAU a-AMy i
CAU a-CAU i
VAU a-CAu a
VAU a-AMy a
VAU a-CAU a
UAU a-UAU i
UAU a-AMy a
UAU a-CAU i
ISC-CAU 1
ISC-UAU 1
ISC-CAU a
isc-ABay a
ISC-ABOy 1
UC-CAU a
uc- UAU a
• -
— •
— •
— •
— •
— •
— •
— •
— •
— •
— •
— •
•
r
- •
- •
— •
- •
- •
-
~ *,
- •
-
—
— i
—
~
—
—
—
—
-
—
—
—
_
—
—
_ .
•
~ g
- •
•
- •
- •
- •
— m
•
- •
- •
•
j_ " „
— ' •
— •
— •
— •
~ 7 *
T * ,
it B"
IT *
i-J— •
— J-- •
IT "
• -U— •
• i|j- •
• -U— •
• -Ur- •
• 11— •
• —I— •
••$+•'.





































• + •
• 4-*"
• ~l *
• -jy
",TL
• L




































".
— •
' + '



*


































                         -0.5
0          0.5           1           1.5

Model Comparison  Measure (MCM)
                 Figure 7-6. A summary of the MCM for each unique model pair. The MCM was computed using
                 predictions made assuming a uniform particle size distribution.

-------
 differences among ISC and the other models and certain model pairs, usually involving one of
 the core (unmodified) models (e.g., GARB 1, ADOM 1, UAM 1).

 122.  Sulfate Particle Distribution

       The performance evaluation exercise was repeated using a more realistic distribution for
 the experiment data sets where sulfate particle deposition was observed Aerosol observations
 such as those reported by Hidy (1984) and Richards et aL (1989)  generally exhibit a multi-
 peaked mass distribution as a function of mean particle diameter. Peaks have been observed in
 the mass fraction at 0.2, 1-2, and 6-10 microns. Each of these peaks are associated with a
 particular pathway of particle emission and/or formation. Many of the particle  experiments
 observed aged sulfate aerosol with a mass fraction peak at 0.2 microns.  The model performance
 evaluation statistics may be sensitive to changing assumptions of the size distribution of the
 particle mass fraction. To address this concern the model performance evaluation exercise was
 repeated with a size distribution taken from Richards et al. (1989) and which is  summarized in
 Table 7-11. The mass fraction peak occurs at 0.2 microns and is nearly twice the value of the
 uniform distribution for a 0.1 micron size range.

       The CPM for the sulfate particle distribution case is illustrated in Figure 7-7. The model
 results are presented in top to bottom ranking by CPM with the highest ranked  (most favored)
 model having the smallest CPM. The ranking indicates that the top three models, CARS 3,
 UAM 2, and ADOM 1 identified previously remain as the top ranked models. Figure  7-8 shows
 that the  MCM and its confidence interval indicates that there is in fact no statistically significant
 difference between any combination of the top three models. The results of the performance
 evaluation and selection exercises are relatively unaffected by the  changes tested in particle
 distribution.

 123  Model Performance When Zeroes are Included

       A test that was performed was to examine the model performance statistics if negative or
 zero observed deposition velocities were retained in the data  base as small positive values.  Five
 such cases  out of 173 observations with valid meteorological data were noted. The effects of
 these five cases were examined by setting the deposition velocity equal to a minimum of 0.005
 cm/s.  Figure 7-9 illustrates the results of ranking the models top-to-bottom by CPM. The
 sulfate particle distribution was used for this exercise due to its greater realism.  Two of the
•same models CARB 3 and UAM 2 remain as the top ranked models and are separated by
 relatively small differences in CPM. The only notable change is that the hybrid ADOM 3 model
 succeeds the core model ADOM 1 as the third best performer.
                                           7-29

-------
                         Table 7-11
Aerosol Mass Fraction as a Function of Size Distribution for Two
             Assumed Aged Sulfate Distributions
Diameter
(Microns)
0.10
0.13
0.16
0.19
0.23
029
036
0.44
0.54
0.66
0.81
1.00
Mass Fraction
(Uniform)
0.083
0.083
0.083
0.083
0.083
0.083
0.083
0.083
0.083
0.083
0.083
0.083
Mass Fraction
(Richards et ai 1989)
0.053
0.095
0.144
0.189
0.161
0.111
0.053
0.046
0.053
0.062
0.035
0.000












                            7-30

-------
Deposition  Model
UAM 2
CARB 3
ADOM 1
ADOM 3
CARB 2
ADOM 2
CARB 0
UAM 1
CARB 1
ISC





-
-
-
-
r
O/ 1 yi/
i*l\ ~~| <^\
* + *
sJ/ ) ^ >J/
^x "~f~ ^r\
* + ^
\L^ 1
^T^ i
^Jx I
^^ 1
"/J^
\^x
^|Ni






*
*
-*
+ \|,/
/|N
/ J__ NJX
C" | ^N










NXW/
TTllv
                           0           0.5            1            1.5
                                Composite Performance  Measure (CPM)
95% low
+   Mean
                                                                  95%  high
           Figure 7-7. A ranking of the models by CPM. The smallest CPM represents the best performance.
           The CPM is for a peaked particle size distribution.

-------
          Model   Names
                             ADOU a-ABOU I
                             ABOH a-CAU a
                              uui I-CAII a
                             ADOtl I —CiH 1
                              luti a-ciu a
                              UAH 1-CiBB I
                             AAOIt 1—C
                              UUI 1-1
                             ABOU a-c
                             ABOU J-ABOH a
                             ABoy I-CABB a
                             UUI a-ABOU I
                             ABOH a-ABOH I
                             ABOU a-CAaa a
                             ABOU a-CABB I
                             uuia-A
                             CABB a-CAU I
                             ABOU I-CAU a
•|4

Ni
                             ABOII a-CAU a
                             UAH I-ABOM I
                             UAH a-CAU a
                             CAU a-CAU a
uui a-ABou a
uui I-CAU a
ABdU I-CAU I
UAH a-CAU a
 uui a-UAU
CAU a-CAU
  UC-CAU
  ISC-UAH
  ISC-CAU t
  ISC-ABOH a
  ISC-CAU a
  ISC-ABOH a
  ISC-ABOH I
  ISC-CAU a
  ISC-UAH a
                                                                                +-
                                                                                  -
                                                                                   ~L
                                     -0.5
                           0              0.5              1               1.5

                           Model Comparison Measure  (MCM)
                           Figure 7-8. A summary of the MCM for each unique model pair.  The MCM was computed using
                           predictions made assuming a peaked particle size distribution.

-------
      Deposition  Model
£
U)
UAM 2
ADOM 1
CARB 3
CARB 2
APOM 3
CARB 0
UAM 1
ADOM 2
CARB 1
ISC






-
-
-
-
* 4- *
* + *
* +
* + *
yi/ _ 1 _
\^/ 1
^^ l~
* 4
¥




^

5K
*
*
1- *
c 4- *










m
                                 0            0.5            1            1.5
                                       Composite  Performance Measure  (CPM) .
95% low
                                                       +  Mean     *  95% higji
                    Figure 7-9. A ranking of the models by CPM. The smallest CPM represents the best performance.
                    The CPM is for a uniform particle size distribution with the observed zeros set to 0.005 cm/s.

-------
       This result was not unexpected since ADOM 1 previously performed the best for the
smallest deposition velocities, and the larger range of observations due to the small deposition
cutoff enables ADOM 3 to predict the standard deviation better. The similarity of the ranking
of models and the magnitude of the CPM for each model suggests that the presence of the five
cases of minimal deposition velocity does not  result in a major change in the results presented in
previous sections.

7.2.4   Cumulative Distribution Results

       The Cumulative Distribution Function (CDF) plots presented in Appendix B illustrate
how the models tend to perform over the range of deposition velocities.  Figure B-2a in the
appendix shows the performance of the core models.  The curves indicate that all models
underpredict  deposition velocities less than 0.2 cm/s, but did much better for larger deposition
velocities. The ISC model is the notable exception. The CDF of the CARS series of models
are illustrated in Figure B-2b.  For small deposition velocities the CARB 3  hybrid model
appears to track the observed CDF most closely. However for deposition velocities greater than
0.5 cm/s there seems to be no appreciable difference between the model variants.  Figure B-2c
shows the closely matching CDPs of the observations and UAM 2 and the  poor match obtained
with ISC.  The improvement in UAM due to the addition of the LAI influence is  best illustrated
by Figure B-2c.  The overprediction tendency of the UAM 1 model is significantly lessened in
the UAM 2 model, although both models perform similarly for the largest deposition velocities.

12J5   Selection of Best Performing Deposition Models

       The results indicate that two hybrid models with an  LAI adjustment perform best.  Three
models, CARB 3, UAM 2, and ADOM 1 have approximately the same overall composite
performance. Although any of the three best performing models is a substantial improvement in
the predictive ability over the current scheme in ISC, each of the new methods has drawbacks
and limitations.  For example, the CARB formulation is empirical which raises questions on  its
generality outside the limits of particle size and surface roughness on which it is based. The
UAM 2 formulation is based on the assumed  equivalence of the Schmidt number term and the
Stokes number term  in the resistance equation for particles of 03 urn diameter, which is an
assumption not fully supported or documented with data. Finally, ADOM  1 shows a distinct
trend for underpredicting deposition velocities for particle size ranges which may  be important
for many combustion sources.  However, although additional improvements in modeling
deposition will undoubtedly be made in the future, a significant benefit can be realized by
replacing the scheme in ISC now with one of the best performing schemes  (CARB 3, UAM 2,
or ADOM 1).
                                          7-34

-------
8.     Summary and Conclusions

       Hie purpose of this study was to review, refine, and test dry deposition techniques that
are suitable for use in regulatory models such as the Industrial Source Complex (ISC) model
Dry deposition is the process by which particulate matter and gaseous pollutants are transferred
from the air to land, water, and vegetative surfaces through "dry" (Le^ non-precipitation)
mechanisms.  Because indirect risk assessment pathways such as fish, food chain, and water
ingestion, commonly dominate total intake and exposure to many pollutants, an accurate
estimate of dry deposition is an important element of many regulatory analyses.  Dry deposition
may also be important for a refined estimate of air quality concentrations for sources subject to
significant plume depletion.

       The dry deposition flux can be written as F = x vd> where F is the flux (g/m2/s), x is the
ambient pollutant concentration (g/m3), and vd is the deposition velocity (m/s), all defined at a
reference height. Standard procedures  can be used for estimating the concentration term of the
flux equation, with  appropriate modifications made to account for plume depletion effects. The
main focus of this study was the testing and evaluation of various methods for computing the
deposition velocity.

       A review of the technical literature identified several models that are suitable for
predicting the dry deposition velocity within the framework of a regulatory model  These
models are listed in Section 1 and described in more detail in Section 2.  Three resistance-based
particle deposition  models were identified which fit the required criteria of this study (i.e.,
methods of sound technical basis that are suitable  for regulatory use for both large and small
particles).  The technical literature suggested that certain parameterizations in these models
could be improved. Therefore, several modifications and enhancements to the core models were
developed and tested in this study. As a result,  a total of ten deposition velocity models were
evaluated (see Section 7).

       A second literature review identified observational data sets which could be used to test
deposition velocity  algorithms.  Based on this review, eight datasets for particulate matter and
fourteen datasets for gases were assembled in a database. Although the ultimate goal is to
evaluate dry deposition for both particulate matter and gases, only particulate  matter deposition
was evaluated in this study. Appendix C lists the observational particle deposition velocity
datasets.  One of the recommendations  of the study is that additional evaluation efforts be
conducted to test the dry deposition models for gases (see Section 2.2) with the datasets listed in
Section 7.
                                            8-1

-------
       As explained in Section 2.1, large particles (Le., above - 20 |im diameter for unit density
particles) tend to be dominated by gravitational settling effects. The concept of gravitational
settling is incorporated into the deposition velocity relationship described in Section 2.13 as well
as the reflection coefficient scheme used in the current ISC model  Particles in the size range
from 1.0 to 20.0 nm diameter are significantly influenced by inertial effects, which enhance the
rate of deposition over that obtained by considering gravitational settling alone. The deposition
of very small particles ( <  ~ 0.1 jim diameter) are dominated by Brownian diffusion.  This
process increases in importance as the size of the particles decreases.  Particles in the size range
from 0.1 to 1.0 urn  diameter show a minimum in the deposition velocity because they are not
efficiently deposited by any of the processes described above.  Although the deposition velocity
database consists of particles in the range from 0.1 to 20  urn diameter sizes, the resistance-based
modeling techniques tested in this  study apply to larger particle as well  For all of the models,
the deposition velocity approaches the same gravitational settling velocity as the size of the
particle becomes large. Therefore, the recommended deposition model is considered to be
applicable to the full range of particle sizes of interest as might be encountered in typical
regulatory studies.

       Two related components necessary for a complete deposition model are (1) a method for
tracking mass conservation and plume depletion, and (2)  a meteorological module for estimating
the micrometeorologicai parameters required by the deposition model In Section 3, four
algorithms for computing plume depletion (source depletion, surface depletion, K-theory
method, and modified source depletion) were reviewed. As discussed  in Section 3, the modified
source depletion model of Horst (1983) is recommended as the overall best approach for use in
a regulatory model  Although evaluations of plume depletion algorithms in the literature against
field data are very limited,  one such study (the dual tracer study of Doran and Horst, 1985) and
intercomparisons of the various techniques with the reference surface depletion method support
the use of the modified source depletion technique. This algorithm is  computationally efficient,
conserves mass, and can account for gravitational settling effects. In Appendix E,
implementation issues associated with the use of the  modified source depletion method are
discussed.

       Methods suitable for estimating the necessary micrometeorologicai parameters for the
dry deposition model are outlined  in Section 4. As required for regulatory applications, these
data must be obtained from routinely available observations. In particular, the dry deposition
models require an estimate of the  surface friction velocity (u.) and the Monin-Obukhov length
(L). The meteorological literature contains  several techniques for estimating these input
parameters.  The techniques selected  here have been shown to produce reasonable results.
Although other mathematical relationships may eventually be used when the deposition

                                            8-2

-------
algorithm is incorporated in the ISC model, the. effects of the change are likely to be minimal
since experience shows that differences among the most commonly-used techniques are small for
most conditions.

       An objective model evaluation methodology Was used to distinguish between the
performance of the various models for predicting particle deposition velocities. Only those
models deemed from the scientific review to parameterize the major known processes affecting
deposition of small and large particles, as discussed in Section 2, were considered for
recommendation as the preferred model.  The model evaluation approach, discussed  in Section
5, is based on the EPA's statistical model evaluation protocol. This approach was used because
it has been successfully demonstrated for many other regulatory model evaluation studies.

       The results of the model evaluation exercise described in section 7 was inconclusive in
picking a single model with statistically significant (e.g. at the 95% confidence level) superior
overall performance. Three models (UAM 2, ADOM 1, and GARB 3 described in section 2)
appeared  to have one or more performance characteristics that were superior to the rest of the
models. The addition of a factor to account for increased deposition area due to leaf area index
(LAI) appears to consistently improve core model performance.

       The recommended procedures for computing the deposition velocity, plume depletion,
and meteorological variables have been implemented in  a revised version of the ISC2 dispersion
model and a companion meteorological processor.  Modified versions of both the short term
(ISC2ST)  and long term (ISG2LT) models and the meteorological processor will be made
available through the EPA's SCRAM bulletin board system.  Draft revisions to the user's guide
and model formulation documents will also be made available for the purposes of public review
and comment

       In future work, it is recommended that an analysis be made to compare the revised
version of the ISC model to the previous version of the model to determine likely changes in
modeled design concentrations. It is also recommended that some additional analysis be
conducted to examine the combined sensitivity of the recommended deposition velocity model
and the modified source depletion model to various input variables.  This can be done within
the new ISC model, since both models have  been included in the revised code. For example,
the relative sensitivity of the deposition fluxes to the particle size distribution, particle density,
surface characteristics (e.g., surface roughness) and meteorological conditions should be
assessed.
                                           8-3

-------
9.     References

Auer, AJH. Jr., 1978:  Correlation of land use and cover with meteorological anomalies. /. AppL
       Meteor^ 17, 636-643.

Baldocchi, D.D., B.B. Hicks and P. Camara, 1987: A canopy stomatal resistance model for
       gaseous deposition to vegetated surfaces. Atmos. Environ., 21, 91-101.

Bowers, JJF., J.R. Bjorklund and CS. Cheney, 1979:  Industrial Source Complex (ISC)
       Dispersion Model User's Guide.  Volume L  EPA-450/4-79-030, U.S. Environmental
       Protection Agency, Research Triangle Park, NC.

Businger, J_A., J.C. Wyngaard, Y. Izumi and E.F. Bradley, 1971:  Flux-profile relationships in the
       atmospheric surface layer. /. Atmos. Sd, 28, 181-189.

Chamberlain, A.C.,  1953: Aspects of travel and deposition of aerosol and vapor clouds. Atomic
       Energy Research Establishment, HP/R 1261.

Cleveland, W.S. and R. McGill, 1984:  Graphical Perception:  Theory, Experimentation, and
       Application to the Development of Graphical Methods. /. Am. Stat. Assoc., 79, 531-444.

Cox, W.M. and J.A. Tikvart, 1990: A Statistical Procedure for Determining the Best Performing
       Air Quality Simulation Model. Atmos. Environ., 24, 2387-2395.

Davies, T.D. and J.R. Mitchell, 1982:  Dry deposition of sulfur dioxide onto grass in rural
       eastern England (with some comparisons with other forms of sulfur deposition).
       Proceedings of the Fourth International Conference on Precipitation Scavenging, Dry
       Deposition, and Resuspension, Volume 2, Santa Monica, CA, 29 November-3 December.

DeBruin, ELA.R. and A-A.M, Holtslag, 1982: A simple parameterization of the surface fluxes of
       sensible and latent heat during daytime compared with the Penman-Monteith concept. /.
       Clim. AppL Meteor., 21, 1610-1621.

Doran J.C. and T.W. Horst, 1985:  An evaluation of Gaussian plume-depletion models with
       dual-tracer field measurements. Atmos. Environ., 19, 939-951.
                                           9-1

-------
Dumbauld, ILK., J.E. Rafferty and H.E. Cramer, 1976:  Dispersion-deposition from aerial spray
       releases. Proceedings of the Third Symposium on Atmospheric Turbulence, Diffusion, and
       Air Quality, October 19-22, Raleigh, NC

Fowler, D. and J.N. Cape, 1982: Dry deposition of SO2 onto a Scots pine forest.  Proceedings of
       the Fourth International Conference on Precipitation Scavenging, Dry Deposition, and
       Resuspension, Volume 2, Santa Monica, CA, 29 November-3 December.

Garland, J.A^ 1982: Dry deposition  of small particles to grass in field conditions.  Proceedings of
       the Fourth International Conference on Precipitation Scavenging, Dry Deposition, and
       Resuspension, Volume 2, Santa Monica, CA, 29 November-3 December.

Godowitcfa, J.M., 1990: Vertical ozone fluxes and related deposition parameters over
       agricultural and forested landscapes. Boundary-Layer MeteoroL, 50, 375-404.

Gray, HLA., M.P. Ligocki, G.E. Moore, CA. Emery, R.C. Kessler, J.P. Cohen, C.C. Chang, S.L
       Balestrini, S.G. Douglas, R.R. Schulhof, J.P. Killus, OS. Burton, 1991:  Deterministic
       Modeling in the Navajo Generating Station Visibility Study. Volume EL  Appendix E
       (Description of deposition algorithms). Systems Applications, International, San Rafael,
       CA.

Hanna, S.R. and J.C. Chang, 1990: Modification of the Hybrid Plume Dispersion  Model
       (HPDM) for urban conditions and its evaluation using the Indianapolis data set. Vol.
       in. Analysis of urban boundary layer data. Sigma Research Corp., Concord,  MA.

Hanna, S.R and J.C. Chang, 199 la:  Modification of the Hybrid Plume Dispersion Model
       (HPDM) for urban conditions and its evaluation using the Indianapolis data set. Vol. I.
       User's guide for HPDM-Urban.  Sigma Research Corp., Concord, MA.

Hanna, S.R. and J:C. Chang, 1991b:  SIGPRO - A meteorological preprocessor for dispersion
       model applications to stack plumes in urban areas.  Seventh Joint AMS-AWMA Conf.
       on AppL of Air Poll Meteor., New Orleans, LA, Jan., 1991.

Hanna, S.R. and J.C. Chang, 1992: Boundary-layer parameterizations for applied  dispersion
       modeling over urban areas. Boundary-Layer Meteorology, 58, 229-259.
                                           9-2

-------
Harrison, R.M, S. Rapsomanikis and A. Turnbull, 1989:  Land-surface exchange in a
       chemically-reactive system:  Surface fluxes of HNO^ HO and NH^ Atmos. Environ., 23,
       1795-1800.

Hicks, B.B., 1982: Critical assessment document on acid deposition.  ATDL Contrib. File No.
       81/24, Atmos. Turb. and Diff. Laboratory, Oak Ridge, TN.

Hicks, B.B., D.D. Baldocchi, T.P. Meyers, R.P. Hosker, Jr. and D.R. Matt, 1987: A preliminary
       multiple resistance routine for deriving dry deposition velocities from measured
       quantities. Water, Air, and Soil Poll, 36, 311-330.

Hicks, B.B., D.R. Matt and R.T. McMillen, 1989:  A micrometeorological investigation of
       surface exchange of trace gases: A case study.  NOAA Tech. Memo.  ERL ARL-172, Air
       Resources Laboratory, Silver Spring, MD.

Hicks, B.B., M.L. Wesely, R.L. Coulter, R.L. Hart, J.L. Durham, R. Speer and D.H. Stedman,
       1986: An experimental study of sulfur and NOX fluxes over grassland. Boundary-Layer
       MeteoroL, 34, 103-121.

Hidy, G.M., 1984: Aerosols: An Industrial and Environmental Science.  Academic Press, Inc.,
       New York, NY.

Hjelmfelt, M.R., 1982:  Numerical simulation of the effects of St. Louis on mesoscale
       boundary-layer airflow and vertical air motion:  Simulations of urban vs. non-urban
       effects.  /. AppL Meteor., 21, 1239-1257.

Holtslag A-A.M. and A.P. van Ulden,  1983: A simple scheme for daytime estimates of the
       surface fluxes from routine weather data. /. dim.  and AppL Meteor.,  22, 517-529.

Horst, T.W., 1977: A surface depletion model for deposition from a Gaussian plume.  Atmos.
       Environ., 11, 41-46.

Horst, T.W., 1983: A correction to the Gaussian source-depletion model.  In Precipitation
       Scavenging, Dry Deposition and Resuspension, H.R.  Pruppacher, R.G. Semonin, W.G.N.
       Slinn, eds., Elsevier, NY.

Horst, T.W., 1984: The modification of plume models to account for dry deposition. Boundary-
       Layer Meteor., 30, 413-430.
                                          9-3

-------
Hosker, RJ*. and S.E. Undberg, 1982:  Review: Atmospheric Deposition and Plant Assimilation
       of Gases and Particles. Atmos.  Environ., 16, 889-910.

Huebert, BJ., 1982; Measurements of the dry-deposition flux of nitric acid vapor to grasslands
       and forest.  Proceedings of the Fourth International Conference on Precipitation Scavenging,
       Dry Deposition, and Resuspension, Volume 2, Santa Monica, CA, 29 Nov.-3 Dec.

Lorenz, R, and CE. Murphy, Jr., 1989:  Dry deposition of particles to a pine plantation.
       Boundary-Layer MeteoroL, 46, 355-366.

McDonald, J.E., 1960: An aid to computation of terminal fall velocities of spheres. /.  Met, 17,
       463.

McMillen R.T., D.R. Matt, BJB. Hicks and J.D. Womack, 1987: Dry deposition measurements
       of sulfur dioxide to a spruce-fir forest in the Black Forest: A data report. NOAA Tech.
       Memo. ERL ARL-152, Air  Resources Laboratory, Silver Spring, MD.

Meyers, T.P., 1987: The sensitivity of modeled SO2 fluxes and profiles to stomatal and boundary
       layer resistances.  Water, Air, and Soil PolL, 35, 261-278.

Meyers, T.P. and D.D. Baldocchi, 1988: A comparison of models for deriving dry deposition
       fluxes of O3 and SO2 to a forest canopy.  Tellus, 40B, 270-284.

Meyers, T.P., BJ. Huebert, and B.B. Hicks, 1989: HNO3 deposition  to a deciduous forest.
       Boundary-Layer MeteoroL, 49, 395-410.

MoJler U. and G. Schumann, 1970: Mechanisms of transport from the atmosphere to  the
       earth's surface. /. Geophy. Res., 75, 3013-3019.

Nicholson, K.W. and TD. Davies, 1987:  Field Measurements of die dry deposition of
       particulate sulphate.  Atmos. Environ., 21, 1561-1571.

Oke, T.R., 1978: Boundary Layer Climates.  John Wiley & Sons, New York, NY.

Oke, T.R., 1982: The energetic basis of the urban heat island.  Quart. J.R. Meteor. Soc., 108,
       1-24.
                                           9-4

-------
Overcamp, TJ., 1976:  A general Gaussian diffusion-deposition model for elevated point
       sources. J.AppL Meteor., 15, 1167-117L

Padro, J., G. den Hartog, and H.H. Neumann,  1991:  An investigation of the ADOM dry
       deposition module using summertime O3 measurements above a deciduous forest.
       Atmos. Environ., 25A, 1689-1704.

Plate, E. and AA. Quraishi, 1965:  Modeling of velocity distribution inside and above tall crops.
       /. AppL MeteoroL, 4, 400-408.

Pleim, J., A. Venkatram and R. Yamartino, 1984: ADOM/TADAP model development
       program. Volume 4. The dry deposition module. Ontario Ministry of the Environment,
       Rexdale, Ontario.

Rao, K. S., 1981: Analytical solutions of a gradient-transfer model for plume deposition and
       sedimentation.  NOAA Tech. Memo. ERL ARL-109, Air Resources Laboratory, Silver
       Spring, MD.

Richards, L.W., JA. Anderson, D.L. Blumenthal, JA. McDonald, P.S. Bhardwaja, R.B.
       Candelaria and D.W. Moon, 1989:  Nitrogen and sulfur chemistry and aerosol formation
       in a western coal-fired power plant plume. AWMA/EPA International Specialty
       Conference. Ester Park, CO, October,  1989.  *

Scire J.S., F.W. Lurmann, P. Karamchandani, A. Venkatram, R. Yamartino, J. Young and J.
       Pleim, 1986: ADOM/TADAP model development program. Volume 9. ADOM/TADAP
       User's Guide.  Ontario Ministry of the Environment, Rexdale, Ontario, Canada.

Scire, J.S., D.G. Strimaitis and RJ. Yamartino, 1990: Model formulation and user's guide for
       the CALPUFF dispersion model Sigma Research Corp.,  Concord, MA.

Scire, J.S. and D.L. Wojichowski, 1987: Modeling deposition and dispersion in an urban
       environment MA.S.S.-APCA 33rd Anniv. Technical Conference, November 3-6,
       Atlantic City, NJ.

Sehmel, GA., 1980: Particle and gas dry deposition: A review. Atmos. Environ., 14, 983-1011.

Sehmel, G A., 1984: Deposition and Resuspension.  Chapter 12 in Atmospheric Science  and
       Power Production, DOE/TIC-27601. U.S. Department of Energy, D. Randerson, Ed.

                                         9-5

-------
Sehmei, G-A. and W.H. Hodgson, 1978:  A model for predicting dry deposition of particles and
       gases to environmental surfaces.  PNL-SA-6721, Battelle Pacific Northwest Laboratory.

Sehmei, G.A. and S.L. Sutter, 1974: Particle deposition rates on a water surface as a function of
       particle diameter and air velocity. /. Rechs. Atmos., in, 911-918.

Shieh, CM., MJL Weseiy and CJ. Walcek, 1986:  A dry deposition module for regional acid
       deposition. EPA/600/3-86/037, U.S. Environmental Protection Agency, Research
       Triangle Park, NC.

Slinn, W.G.N., L. Hasse, B.B. Hicks, A.W. Hogan, D. Lai, P.S. Uss, K.O. Munnich, G.A. Sehmei
       and O. Vittori, 1978: Some aspects of the transfer of atmospheric trace constituents past
       the air-sea interface. Atmos. Environ., 12, 2055-2087.

Slinn, S.A. and W.G.N. Slinn, 1980: Predictions for particle deposition on natural waters.
       Atmos. Environ., 14, 1013-1016.

Slinn, W.GJN., 1982:  Predictions for particle deposition to vegetative canopies. Atmos. Environ.,
       16, 1785-1794.

van Ulden, A.P. and A.A.M. Holtslag,  1985:  Estimation of atmospheric boundary layer
       parameters for diffusion applicati6ns. /. AppL  Meteor., 24, 1196-1207.

Walcek, CJ., RA. Brost, J.S. Chang and ML. Weseiy, 1986: SO^ sulfate, and HNO3 deposition
       velocities computed using regional land use and meteorological data. Atmos. Environ.,
       20, 949-964.

Wang, LT. and P.C. Chen, 1980: Estimations of .heat  and momentum fluxes near the ground.
       Proc. 2nd Joint Conf. on Applications of Air Pott. Meteor., American Meteorological
       Society, Boston,  MA, 764-769.

Weil, J.C. and R.P. Brower,  1983:  Estimating Convective Boundary Layer Parameters for Diffusion
       Application.  Draft Report prepared by Environmental Center, Martin Marietta Corp.,
       for Maryland Dept of Natural Resources.

Weseiy, ML., 1989:  Parameterization of surface resistances to gaseous dry deposition in
       regional-scale numerical models.  Atmos. Environ., 23, 1293-1304.
                                           9-6

-------
Wesely, MJL, D.R. Cook, and R.M. Williams, 1981: Field measurement of small ozone fluxes to
       snow, wet bare soil, and lake water. Boundary-Layer MeteoroL, 20, 459-471.

Wesely, MJL, D.R. Cook and RJL Hart, 1983:  Fluxes of gases and particles above a deciduous
       forest in wintertime.  Boundary-Layer MeieoroL, 27, 237-255.

Wesley, MJL, D.R. Cook, RJL Hart, B.B. Hicks, J.L. Durham, R.E. Speer, D.H. Stedman and
       RJ. Tropp, 1982:  Eddy-correlation measurements of the dry deposition of particulate
       sulfur and submicron particles.  Proceedings of the Fourth International Conference on
       Precipitation Scavenging, Dry Deposition, and Resuspension, Volume 2, Santa Monica, CA,
       29 November-3 December.

Wesely, MX., JA. Eastman,  D.R. Cook and B.B. Hicks, 1978:  Daytime variations of ozone
       eddy fluxes to maize.  Boundary-Layer MeteoroL, 15, 361-373.

Wesely, M.L. and B.B. Hicks, 1977: Some factors that affect the deposition rates of sulfur
       dioxide and similar gases on vegetation. /. Air PolL Control Assoc., 27, 1110-1116.

Winges, KD., 1990: User's guide for the Fugitive Dust Model (FDM) (revised).  Volume 1:
       User's Instructions. EPA-910/9-88-202R.  U.S. EPA, Region 10, Seattle,  WA.

Winges, K.D., 1992: Personal communication to J. Scire.

Yamartino, RJ., J.S. Scire, S.R. Hanna, G.R. Carmichael and Y.S. Chang, 1992:  The
       CALGRID mesoscale photochemical grid model. Volume I. Model formulation.
       Atmos. Environ., 26A,  1493-1512.
                                          9-7

-------
           Appendix A





Estimation of ISC Deposition Velocity

-------
                            Estimation of ISC Deposition Velocity
       The basis for the present ISC deposition algorithm is found in Dumbauld et aL (1976)
and in Overeamp (1976). In this approach the particles are assumed to move towards the
ground with a total velocity equal to the sum of the gravitational settling velocity and an average
turbulent velocity which determines the rate of plume spreading. This turbulent velocity is given
by
                            v, = (uHt - v^c)
                                             bM>
                                                                                   (A-l)

where u is the stack height wind speed, H,. is the effective plume height, vg is the gravitational
settling velocity, x is the downwind distance, and the vertical dispersion coefficient, av is given
by the relation:
                                         B

                                                                                   (A-2)
The coefficients A and B are stability dependent, and are treated as average values in the ISC
deposition model By differentiating Equation A-2 and substituting into Equation A-l we have:
                                                                                   (A-3)'

       The turbulent velocity is thus a function of the ratio of the plume centeriine height to the
downwind distance.  For small particles, the uH,./x term is much larger than the settling velocity
which can be ignored.

       In nearly all of the small particle experiments there is no specific plume-receptor
information in order to directly estimate an He/x.  Furthermore, in some experiments it is
possible that several sources may be contributing to the deposition fluxes.  How then  does one
estimate an He/x that will be appropriate and consistent with the information provided  to the
other deposition velocity models?

       The other deposition models estimate a turbulent velocity near the surface as  being
equal to the inverse of the aerodynamic resistance, rr  Thus the He/x term can be estimated
from the relation:
                                           A-l

-------
                                                                                   (A-4)
where the aerodynamic resistance is given by:
                                                                                   (A-5)

which is a formulation common to the ADOM and UAM models. Using the information on
Pasquill Gifford Turner stability class and the friction velocity and Monin-Obukhov length, and
assuming a reference height of 10 m, a displacement plane height of zero and using the given
surface roughness length and wind speed we estimated the equivalent H,./x needed by the ISC
model

       The actual deposition velocity used in the ISC model is, following Overcamp (1976),
equal to:
                                                                                   (A-6)

The settling velocity used in ISC, vp is given by the Stokes relation:
                                 v<
                                                                                   (A-7)
where p is the particle density, g is the acceleration due to gravity, d is the panicle diameter,
and u is the absolute viscosity of air (p.  • 1.83 x 10"* g cm"1 s"1).

       The reflection coefficient, «, in Equation A-6 is the fraction of the image plume source
remaining.  In the limit of a fully reflecting plume the image plume experiences no depletion
and cc approaches-one.  The Dumbauld  et aL (1976) paper indicates that when the settling
velocity drops below 0.1 cm/s the reflection coefficient is set equal to 1.  As a result, ISC will
predict a zero deposition velocity for many of the cases in the small particle dataset. This limit
on deposition velocity especially affects  sulfate paniculate matter since the size range for such
paniculate matter peaks in the submicron diameter range.
                                            A-2

-------
     Appendix B




Supplemental Graphics

-------
Series 1: Scatter plots of Observed Versus Model
         Predicted Deposition Velocities

-------
                                                                       Key
                                                                      3   GARB 0
1E-4
 1E-3        1E-2
Observations
1E-1
Figure B-la.   Scatter plot  of  observed deposition velocity (cm/s) versus  model predicted
             deposition velocity (cm/s) for the complete small particle data set

                                        B-l

-------
1E-4
                                                                      Key
                                                                      CD   GARB 1
 1E-3        1E-2
Observations
1E-1
1EO
1E1
Figure B-lb.  Scatter plot  of observed  deposition velocity (cm/s) versus  model predicted
             deposition velocity (cm/s) for the complete small particle data set.
                                        B-2

-------
                                                                      Key
                                                                      Z]   CARS  2
1E-4
 1E-3        IE
Observations
1E-1
1EO
1E1
Figure B-lc.   Scatter  plot of  observed deposition velocity (cm/s) versus  model predicted
             deposition velocity (cm/s) for the complete small particle data set
                                        B-3

-------
                                                                      Key
                                                                          CARS  3
IE-4
 1E-3        1E-2
Observations
1E-1
1EO
Figure B-ld.  Scatter plot  of  observed deposition velocity (cm/s) versus  model predicted
             deposition velocity (cm/s) for the complete small particle data set.

                                        B-4

-------
                                                                      Key
                                                                      Zl   ADOM 1
1E-4
 1E-3        1E-2
Observations
1E-1
1EO
1E1
Figure B-le.   Scatter plot  of  observed deposition velocity (cm/s) versus  model predicted
             deposition velocity (cm/s) for the complete small particle data set.
                                        B-5

-------
                                                                       Key
                                                                           ADOM 2
1E-4
 1E-3       1E-2
Observations
1EO
1E1
Figure B-lf.   Scatter plot  of observed  deposition  velocity  (cm/s)  versus model predicted
             deposition velocity (cm/s) for the complete small particle data set

                                        B-6

-------
                                                                      Key
                                                                     H   ADOM 3
1E-4
 1E-3        1E-2
Observations
1E-1
1EO
1E1
Figure B-lg.   Scatter plot  of  observed deposition velocity  (cm/s)  versus model  predicted
             deposition velocity (cm/s) for the complete small particle data set.

                                       B-7

-------
                                                                      Key
                                                                      H   UAM 1
1E-4
 1E-3        1E-2
Observations
1E-1
1EO
1E1
Figure B-lh.   Scatter plot  of  observed  deposition  velocity (cm/s) versus  model predicted
             deposition velocity (cm/s) for the complete small particle data set.
                                        B-8

-------
                                                                     Key
                                                                        UAM 2
1E-4
 1E-3        1E-2
Observations
1E-1
1EO
1E1
Figure B-1L  Scatter plot  of observed deposition velocity  (cm/s)  versus model predicted
            deposition velocity, (cm/s) for the complete small particle data set

                                       B-9

-------
Cd
     1
tl
o C*.
a n
g:-1
§u
* s
 ^« w
•S-8"

^•3
 09 0)
>-X O-
 5J>
 ° 8-
 S-'S
 &8

 §i:
1°
 Sii

 i!
S.J
S" s?
p p
8

r|

  t

 "S
  (D

  I
  a
        Predictions

      1E-3       1E-2
                                                    1E-1
                                                                1EO
                   o ^

                   u W
                   n> I"
                   3 co
                   o
                   e
                     W -
                 M -
                 I '
                     W-
                     o
                     W-
iiti mi
                          -e-
           t  i  i i mi
                                              i _ i  t i mi
                                                                               1E1
                                                          i _ i i i mi
                                                                        i _ i  i HIM
                                                                           (D
                                                                         W
                                                                         o
                                                                         s

-------
Series 2:  Cumulative Distribution Function (CDF) Plots for Observations
             and Model Predictions of Deposition Velocity

-------
                                                                           Key
                                                                          D   OBSERVED
                                                                          O   CARB  1
                                                                          A   ADOM  1
                                                                          +   UAM 1
                                                                          X   ISC
I   I I  I I  11
        1E-2
             Deposition  Velocity
        Figure B-2a. Cumulative probability plot of deposition velocity (cm/s) using
                        the complete small particle data set.

-------
KJ
X
                                                                                            Key
                                                                                                OBSERVED
                                                                                                GARB 1
                                                                                                CARB 2
                                                                                                CARB 3
                                                                                                CARB 0
                                             TTTT
                            1E-2             1E-1
                                 Deposition Velocity
                            Figure B-2b. Cumulative probability plot of deposition velocity (cm/s) using

-------
                                                                            Key
                                                                            (3   OBSERVED
                                                                            (D   ADOM 1
                                                                            A   ADOM 2
                                                                            +   ADOM 3
                                                                            X   ISC
Figure B-2c. Cumulative probability plot of deposition velocity (cm/s) using
                 the complete small particle data set.

-------

                                                                                         Key
                                                                                        E)  OBSERVED
                                                                                        (D  UAM 1
                                                                                        A  UAM 2
                                                                                        *  ISC
?E-3
TIT	1	1   I  I I  I II
IE1-2              1E-1
     Deposition Velocity
                   Figure B-2d. Cumulative probability plot of deposition velocity (cm/s) using
                                   the complete small particle data set.

-------
                  Appendix C




Observational Particle Deposition Velocity Data Sets

-------
                                     Technical Note
               Description  of Particle Deposition  Velocity Data Sets
        There are 24 original data sets altogether which are used in the data analysis.  These data sets
are read into the fortran program PARTVD and are analyzed  and processed  in the manner shown
schematically in Figure C-l.  The PARTVD software reads in each observed deposition velocity case
in one of two formats.  If the global roughness length zO is present it is used for the entire data set and
data on Richardson number (Ri) and nondimensional fluxes of heat and momentum are assumed to NOT
be present.  If the global roughness length is set to -999.7  then this data is assumed to be present  for
each observed deposition velocity present in the sample. One can note this difference when comparing
sample sets numbers 1 and 2. If the reference height for wind speed is missing a default value of 10
m is used which is typical of the assumptions made in applications of the deposition models. If the leaf
area index  is missing a value is assigned based on the the land use type (vegetation state), or whether
is is a special wind tunnel study.  The effects of nonuniform particle size distributions is  input as a
fractional mass weight for each size range. Presently this is done for only the sulfate particle samples.
In cases where case specific Ri, phim, and phih are assigned -999.9 there is NOT sufficient information
to generate deposition velocity estimates for all models and these cases are dropped. For example  the
fifth case in sample set number 3 would be dropped. Only samples with positive deposition velocities
are kept.

        The PARTVD  program produces  a set of 173 predictions for 9 models. In addition to the 9
model  predictions, a number  of additional meteorological variables are also  output so  as to provide a
means  of stratifying the deposition velocity data.  This data set is read in and used to produce estimates
of deposition velocity for ISC which becomes the 1 Oth model  prediction set and is added to the input
data set and  written out.   These data sets are displayed  in Table C-l  for  the uniform particle size
distribution and Table C-2 for the sulfate peaked particle distribution.  Footnotes provide definitions  for
each column  in the tables.

-------
           PARTVD.FOR
         model deposition
            velocity
           ISCDEP.FOR
           Estimate ISC
           deposition
>PERFSTAT.FOR
   Does CPM
   estimation
     printed
     output.
                                 EVALSTAT.INP
EVALSTAT.FOR/
Plots various ^
  statistics
         Figure C-1.

   Schematic diagram of
       data processing

-------
Table C-1. The data making up the overall dataset for a uniform particle distribution. The data set contains 173 data points including observed
zero deposition velocities which have been set to a lower limit of 0.005 cm/s.  The footnotes define the variable A thru X.
Deposit ion Modol
OBS CAM 1 CAM 2 CAM 3 CAM 0 ADOM 1 ADOM 2 ADOM 3 UAH 1 UAM 2 ISC ABCD E F G H
!:11H S:iiiS Hill l:S!!i 111! l:i!Sl lilii !:H?1 H1!1 1'«" °:"" l a « a » s » » z.» «:s
j:Sjj ifi jijjjj j:5ij iisB liiiS s.sjl sIH, lillli iHi! sisll i 11 II i I3 » H 1! Hi |i:i
1./4O0 O.BQ16 0.1)5 96 0.7043 0.6016 0.9154 0.6852 0.7050 0 $724 0057 0 1169 5 26 fl5 Tn*; 11 5J ot «i i 01 ia CA
3.1400 0.7904 0.8774 0.7165 0.7904 1.0436 0.7185 0.7409 1.0131 !o496 t.ltM 6 5 83 SI 22 10 22 40 111 ll il
3.0100 0.7904 0.8771 0.7165 0.7904 1.0436 0.7185 0.7409 1.0131 .0496 0.1490 6 5 83 Si 22 10 22 40 540 ij'lo
2.8400 0.7904 0.8774 0.7165 0.7904 1.0436 0.7185 0.7409 1.0131 .0196 0.1490 6 5 83 Ss 22 10 22 40 632 1750
1.7500 0.6808 0.7706 0.6434 0.6808 0.4911 0.5564 0.5663 0.6207 .8119 0.0822 6 12 83 SI 22 43 23 13 300 it'll
1.6200 0.6808 0.7706 0.6131 0.6808 0.4911 0.5564 0.5663 0 8207 0 8449 0 0822 6 12 83 ZnS 22 13 23 13 1 14 11 M
1.3100 0.6808 0.7706 0.6434 O.(808 0.1911 0.5561 0.5663 0.8207 0 8419 I'.llll ( 11 11 Si 22 43 23 13 I'll It'll
1.5(00 0.7357 0.8606 0.7018 0.7357 0.9218 0.6887 0.7089 0.9787 1.0129 0.1314 6 21 83 SI 23 6 23 28 307 14 10
1.4700 0.7357 0.8606 0.7048 0.7357 0.9248 0.6887 0.7089 0 9787 1 0129 0 1314 6 24 83 Ss 21 f. 31 >l 1 °I i! i«
1.1100 0.7357 0.8606 0.7048 0.7357 0.9248 0.6887 0.7089 0 9787 I'.llll i.lllt 1 " " Ss 23 6 23 28 3 t( It 11
1.1700 0.8752 0.9275 0.7524 0.8752 1.3838 0.8085 0.8371 1.0860 1 1266 01601 6 27 83 SI 21 31 22 1 3 17 3050
1.1500 0.8752 0.9275 0.7524 0.8752 1.3838 0 8085 0.8371 1.0860 1 1266 0 1601 6 27 83 SI 21 31 22 1 '380 3050
1.1000 0.8752 0.9275 0.7524 0.8752 1.3838 0.8085 0.8371 1 0860 1 1266 01601 6 27 83 Ss 21 31 33 i I'll ll'll
0.0400 0.0190 0.0243 0.0732 0.0190 0.0054 0.0370 0.0948 oioOTO I'.llll I'.llll 6 " 79 £5 " 0 0 0 ill li"
I'llll S'S!« J J199 °-°691 °-°136 '•°OS1 °-°353 °-0858 °-00" 0.0183 0.0001 6 21 79 S04 2 36 0 0 I tl ill
S'S SJ J-°197 0'0213 °-0550 O-0197 0-OU2 O-0752 O-2120 0.0151 0.0481 0.0001 6 21 79 SO4 0040 253 21 20
S^SS S'SHl S S2" 0'0943 0-°m 0-0°" ° 02« 0.0173 0.0014 0.0109 0.0001 6 21 79 S04 3 55 0 0 1(4 830
0.3300 0.0220 0.0287 0.0563 0.0220 0.0121 0.0815 0.2306 0.0164 0 0526 0 0001 7 18 79 SGI 003 13 269 1410
0.0200 0.0125 0.0210 0.0327 0.0125 0.0075 0.0512 0.0865 0.0100 0.0183 0 0001 2 19 80 £1 1? 23 0 0 2 'if 1'"
0.2000 0.0086 0.0140 0.0326 0.0086 0.0115 0.0773 0 135J 0.0155 0.0291 0.0001 2 20 80 S04 0 0 4 40 373 5 ' 90
0-0900 0.0083 0.0137 J-0319 0.0083 0.0086 0.0578 0.0980 0.0114 0.0210 0.0001 2 21 80 S04 9 55 0 0 3^13 i.ll
J'SJSJ i!'?.1!5 i!'0261 °-0592 o!oi65 o!oi37 0!0923 0^625 0^0187 o!o603 oioOOl 10 17 79 SO4 6 54 0 " I'll I'll
l'l,ll J'JJ!4 S-0.31?. °-°585 °-0194 °-0064 O-0441 0-UO2 0.0085 0.0252 0.0001 10 18 79 SO4 7000 153 5 10
S'JJSS S'Si 1 0.0237 0.0600 0.0174 0 0143 0.0954 0.2722 0.0195 0.0628 0.0001 10 19 79 SO4 004 38 381 1300
0.0800 0.0216 0.0370 0.0550 0.0246 0.0080 0.0545 0.1467 0.0106 0 0328 0 0001 10 19 79 SC4 6 58 0 0 1 50 S BO
0.1000 0.0173 0.0303 0.0264 0.0173 0.0064 0.0444 0.0596 0.0085 0.0119 0 0001 11 21 79 »4 10 50 0 0 I'll I'll
0.0700 0.0105 0.0177 0.0244 0.0105 0.0082 0.0562 0.0768 0.0110 0.0157 0 0001 li 22 79 M4 1 1 1 si 1 30 ill
S'lISS 2-°jS29 S'01!1 °-0249 0-0108 O'OO" O'0509 0-°«84 0.0100 0.0141 0.0001 11 27 79 S04 10 53 0 0 252 ill
0.0500 0.0091 0.0152 0.0211 0.0091 0.0097 0.0651 0.0894 0.0129 0.0185 0.0001 11 29 79 SO4 00 4 7 2 31 520
0.0200 0 0108 0.0188 0.0258 0.0108 0.0068 0.0458 0.0611 0.0089 0 0126 0.0001 11 29 79 iol 11 1 0 0 2 41 3 10
0.3200 0.0086 0.0152 0.0241 0.0086 0.0096 0.0647 0.0888 0.0129 0 0184 0.0001 12 3 79 iol 11 9 0 0 345 290
0.0700 0.0163 0.0250 0.0273 0.0163 0.0061 0.0418 0.0555 0.0080 0 0112 0 0001 12 6 79 SOI 11 12 0 0 1 ?« 7 so
0.2300 0.0099 0.0152 0.0242 0.0099 0.0087 0.0584 0.0798 O.Olis I.lllt I'.llll 11 7 79 ISJ J " 2 35 I'll , 00
1 ,,ll I'ltl', l'llll S-0254 °-0078 °-01" °-0837 o-1164 o-0170 o-0246 o-oooi " « 79 =°4 11 20 o o 4 72 in
!'J1SS S'*1M 0-°l" °-0296 O-O"2 0.0162 0.1077 0 1514 0.0221 0.0322 0.0001 1 22 80 SO4 0 0 3 18 6 03 770
2'24°.S S 222! 2-21'8 "-02" °-009S °-0053 O-0364 0 0473 0.0068 0.009J 0.0001 1 23 80 SOI 10 59 0 4 207 1 70
J'SJSJ 2-2°!' "-0119 "•"" O'0068 O'0050 O'0340 O-0438 0-0063 0 0087 0.0001 1 24 80 Iol ll 3 0 0 2 19 1 40
0 2100 0 0054 0.0096 0.0253 0.0054 0.0071 0.0481 0.0646 0.0093 0.0132 0.0001 1 28 80 SOI 0054 275 1 20
0.2200 0.0076 0.0118 0.0241 0.0076 0.0094 0.0629 0 0861 0.0125 0.0179 0.0001- 1 29 80 Sot 10 42 0 0 3 17 7 30
2-?.222 2-2°" °-0099 "'O"4 0-0065 0.0126 0.0838 0.1163 0.0171 0.0217 0.0001 1 30 80 S04 0047 502 820
S'2J?J 2 2J?1 °-°159 0-02" ff-0101 O-0072 O-0488 0-06« O-0094 0 0133 0 0001 1 30 So SOI 10 49 0 0 2 40 670
0.0050 0.0073 0.0118 0.0241 0.0073 0.0093 0.0627 0.0859 0.0121 0 0178 0.0001 2 14 80 SOI 11 22 0 0 3 46 590
0.2800 0.0069 0.0119 0.0319 0.0069 0.0086 0.0578 0 0986 0 0114 0 0209 0 0001 3 i 80 »4 9 59 0 0 3 12 I'M
S'5!SS S-S522 °-0118 °-0318 O-0072 0-0»90 ° °«01 0.1036 0.0119 0 0220 0.0001 3 6 80 SO4 00 4 36 323 5 30
0.0400 0.0074 0.0121 0.0327 0.0074 0.0075 0.0509 0.0853 0.0099 0.0180 0.0001 3 6 80 S04 9 56 00 270 1 70
0.0200 0.0113 0.0188 0.0321 0.0113 0.0083 0.0560 0.0956 0.0110 0.0203 0 0001 3 11 80 S04 10 3 0 0 2 52 ill
0.0800 0.0069 0.0120 0.0183 0.0069 0.0079 0.0535 0.0574 0.0101 0.0113 0.0001 1 25 80 SO4 004 56 253 I'll
S'^SS S'SS'l °-0093 °-°l82 0-°°" O-0083 0-0555 00596 0.0110 00119 0.0001 2 15 80 SO4 0 0 S 39 310 710
0.0100 0.0087 0.0143 0.0188 0.0087 0.0072 0.0191 0 0527 0.0094 0 0102 0 0001 2 27 80 SOI 004 52 206 S 10
0 3800 0.0076 0.0121 0.0186 0.0076 0.0075 0.0515 0.0552 0.0100 0.0108 0.0001 2 29 80 SO 4 004 4 230 I'M
0 2500 0.0094 0.0165 0.0180 0.0094 0.0100 0 0676 0.0728 0 0135 0 0117 0.0001 3 lo 80 io! 9 59 0 0 3 01 360
I'llll S'SJS? J-S?1? S'Sff2 "'"'""' 0-°°50 0.0311 0 0768 0.0064 0.0181 0.0001 10 21 79 So* 6 52 0 0 l'?7 ill
S'JISJ !'°101 °-<)1*1 O-0630 0.0101 0.0076 0.0515 0.1466 0.0100 0.0338 0 0001 6 9 80 S04 5 11 0 0 2 15 11 20
I SJSJ S J ' 0.0139 0.0629 0.0109 0.0113 0.0759 0.2313 0.0153 0 0539 0.0001 6 10 80 SOI 0055 3 58 14 70
! IJ2J J-"08 "-01" °-°"0 0-0108 0-0076 0 0516 0.1183 0 0100 0 0318 0.0001 6 12 80 S04 5 12 0 0 2 28 1000
0.6100 0.0617 0 0622 0.0571 0.0617 0.0056 0 0393 0.0923 0 0073 0 0189 0.0001 9 17 79 S !5 5 15 30 1 83 23 60
!'?2SS S °495 °-°S27 0-0499 0.0195 0 0071 0.0193 0.1209 0.0091 0 02S1 0.0001 9 25 79 I 11 3! 11 I ill ll'll
0.4100 0 0501 0.0526 0.0499 0.0501 0 0071 0 0493 0 1207 0 0091 0.0251 0 0001 9 25 79 S ll I 12 30 1 44 2200
S ,1™ S'"!08 0-°S26 °-0499 O'0508 0.0071 0 0191 0 1208 0 0091 0 0251 0.0001 9 25 79 S 12 35 13 0 1 34 2250
J'J^J ° °500 ° OSn O-0489 0 0500 0.0075 0 0517 0 1268 0 0099 0 0266 0 0001 9 25 79 S 13 5 11 30 1 29 22 90
S'SSIS °-OS13 °-"26 °-0498 O-0513 0-00n O-0493 0.1204 0.0094 0.0251 O.OOOi 92579! It 35 ll 1 ill llll
?'2^S °-°512 0-°5" °-0198 0 0512 0.0071 0 0191 0 1188 0 0091 0.0250 0 0001 9 25 79 S IS 35 16 0 1 S3 22 90
?'?SSS * '289 l'7923 2'1277 !-6989 2'1279 i-1775 1 6611 2 1015 2 1131 0 2S20 0 0 0 Pb 00 0 0-999'90 20 00
J'iSSS J'J04J J'J9" 1'7508 1 3043 1 3767 O-6268 ° 9671 l-«'24 2 0935 0.1158 0 0 0 Pb 000 0-99990 2000
l'°J^ S l,,l i-°i5° l'"S1 °'93" °-4l28 0 3031 0 1725 1 2507 1 7111 0 0809 0 0 0 Pb 000 0-999 90 2000
1 ,111 S «5f J 5"2 °-8"° °-5418 °-0800 0 122° 0 U51 0 7195 1 2381 0 0001 0 0 0 Pb 000 0-999 90 2000
J'J5JS S 1 0.1563 0.2726 0 0316 0 0615 0 1037 0 3724 0 72S7 0 0001 0 0 0 Pb 000 0-999 90 2000
0.0100 0 0896 0.0991 0 1517 0 0886 0 0122 0 0116 0 0967 0 1060 0 2376 0 0001 0 0 0 Pb 000 0-999 9° 20 00

I
-999.90
-999.90
-999.90
-999.90
-999.90
-999.90
-999.90
-999.90
-999.90
-999.90
-999.90
-999.90
-999.90
-999.90
-999.90
-999.90
-999.90
-999.90
-999.90
-999.90
-999.90
-999.90
-999.90
-999.90
-999.90
-999.90
-999.90
-999 90
-999.90
-999.90
-999.90
-999.90
-999.90
-999.90
-999.90
-999.90
-999.90
-999.90
-999.90
-999.90
-999.90
-999 90
-999.90
-999.90
-999.90
-999.90
-999.90
-999.90
-999.90
-999.90
-999.90
-999.90
-999.90
-999.90
-999.90
-999.90
-999.90
-999 90
-999.90
-999.90
-999.90
-999.90
-999.90
-999.90
-999 90
-999.90
-999.90
-999.90
-999.90
-999.90
-999.90
-999 90
-999.90
-999 90

J K
0.40 0.166E>03
0.40 0.1((E*03
0.40 0.166E»03
0.26 0.440E«02
0.26 0.440E*02
0.26 0.440E<02
0.27 0.770£»02
0.27 0.770E*02
0.27 0.770E+02
0.20 0.340E*02
0.20 0.340E.02
0.20 0.340E>02
0.26 0.590E.02
0.26 0.590E«02
0.26 0.590E<02
0.30 0.710E.02
0.30 0.710E»02
0.30 0.710E*02
0.11 0.141E<02
0.10-0.183E»02
0.26-0.237E>02
0.06 O.S57E.01
0.29 0.900E.10
0.17 0.573E+02
0.28-0.101E+03
0.20 0.38SE«02
0.25 0.900E.10
0.34 0.245E«03
0.14 0.204E»02
0.35-0.333E*03
0.18 0.913E.02
0.14 0.858E.02
0.19-0.167E«03
0.17 0.243E«02
0.23 0.661E«02
0.15 0.17SE«02
0.23 0.116E»02
0.13 0.211E.02
0.20 0.120E+03
0.31 0.120E«03
0.41 0.215E*03
0 11 0.152E.02
0.10 0.176E.02
0.16 0.8S8£<02
0.22 0.215E«03
0.31 0.900E*10
0.16 0.505Et02
0.22 0.162E<03
0.20 0.191E403
0.21 0.900E+10
0.17 0.779E»02
0.19 0.952E.02
0.18-0.(30E*02
0.19 0.120E.03
0.16-0.118E«02
0.17-0.289E>02
0.21 0.245E.03
0.10 0.886E»01
0.17 0.162E403
0.27-0 988E.02
0.17 0 900E*10
0 11-0.256E<01
0 1S-0.373E»01
0.1S-0.470E+01
0 15-0.471E+01
0.16-0 977E+01
0.15-0 622E<01
0.15-0.150E+02
0.35 0.900E«10
0.35 0.900E.10
0.35 0.900E+10
0 35 0.900E+10
0 35 0 900E»10
0 35 0 900E.10

L
-999.900
-999.900
-999.900
-999.900
-999.900
-999.900
-999.900
-999.900
-999.900
-999.900
-999.900
-999.900
-999.900
-999.900
-999 900
-999 900
-999.900
-999.900
-999.900
-999.900
-999.900
-999.900
-999.900
-999.900
-999.900
-999.900
-999.900
-999.900
-999.900
-999.900
-999.900
-999.900
-999.900
-999.900
-999.900
-999.900
-999.900
-999.900
-999.900
-999.900
-999.900
-999.900
-999.900
-999.900
-999.900
-999.900
-999.900
-999.900
-999.900
-999 900
-999.900
-999.900
-999.900
-999.900
-999.900
-999.900
-999.900
-999.900
-999.900
-999.900
-999.900
51.000
88 000
70.000
70.000
11 000
53 000
22.000
-999 900
-999 900
-999 900
-999 900
-999 900
-999 900

H N
3.00 1
3.00 1
3.00 1
3.00 1
3.00 1
3.00 1
3.00 1
3.00 1
3.00 1
3.00 1
3.00 1
3.00 1
3.00 1
3.00 1
3.00 1
3.00 1
3.00 1
3.00 1
0.80 10
0.40 10
1.00 -10
0.30 10
2.20 10
0.90 10
0.30 10
0.30 10
0.20 10
1.50 10
2.30 10
1.10 10
4.30 10
2.00 10
0.60 10
0.60 10
0.40 10
0.60 10
0.40 10
1.10 10
0.40 10
0.20 10
0.30 10
0.30 10
0.10 10
0.10 10
0.20 10
0.10 10
0.40 10
0.20 10
0.20 10
0.20 10
0.20 10
0.70 10
0 20 10
0.10 10
0 . 30 10
0.20 10
0.50 10
1.30 10
0.30 10
0.30 10
0.40 10
9.40 10
9.40 10
9.10 10
9.10 10
9.10 10
9.40 10
9.40 10
2 00 1
2.00 1
2.00 1
2 00 1
2.00 1
2.00 1

O
(.00
(.00
(.00
6.00
(.00
6.00
(.00
6.00
(.00
(.00
(.00
(.00
6.00
(.00
(.00
(.00
6.00
6.00
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
13.00
10.00
7.50
5 00
3.20
1 60

F
6.00
(.00
(.00
6.00
(.00
(.00
(.00
(.00
(.00
(.00
(.00
(.00
(.00
(.00
6.00
(.00
(.00
(.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
13.00
10.00
7.50
5.00
3 20
1.60


at era

V
4. 00-999.90 0.1 1.0 11
4.00-999.90 .1 l.a 11
4.00-99
4.00-19
4.04-99
4.00-99
4.00-99
4.00-99
4.00-99
1.00-91
4.00-99
4.00-99
4.00-99
4.00-99
4.00-99
4.00-99
4.00-99
4.00-99
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
i.oo
1.00
1.00
1.00
1.00
1 00
1.00
1.00
1.00
1.00
1.00
1. 00-99!
1.00-99!
1.00-99!
1.00-99!
1 00-99!
1.00-99!
1.10
1.10
1.10
1.10
9.90-
9.94
1.90
1.10
1.14
9.90
9.94
1.94
9.14
9.94
9.14
1.14
1.40
.00
.44
.44
.44
.04
.44
.04
.1
.1
.1
.1
.1
i!
9.
9.
9.
9.
9.
9.
9.
9.1
.1
.4
.5
.5
.5
.S
.4
.4
.0
.44 .4
.44 .5
.04 .5
.04 .S
.04 .%
.44 .S
.44 .5
.44 .S
.44 .5
.44 .S
.44 .5
.44
.44 4.
.44 4.
.44 4.
.44 4.
.40 4.
.40 0.
.40 4.
.40 .5
.40 .S
.00 .5
.00 .4
.04 .4
.44 .4
.44 1.0
.04 4.1
.44 4.1
.40 4.1
.40 4.1
.44 4.1
.44 2.5
.00 3.0
.00 3.0
.00 3.0
.00 2.0
.04 2.4
.44 2.4
.40 2.0
.40 2.4
.40 2.0
.40 2.0
.90 1.5
.90 15
.90 1.5
.90 1.5
.90 1.5
.90 1.5 1
L.4 11
1.4 11
1.4 11
L.4 11
1.4 11
1.4 11
1.4 11
1.4 11
L.4 11
.4 11
.4 11
.4 11
.4 11
.4 11
.4 11
.4 11
.4 2
.4 14
.4 14
.4 14
.0 14
.4 3
.4 3
.0 ?
.4
.0 1
.0 1
.0 1
.4 1
.4 1
.0 1
.4 1
.0 1
.0 1
.4 1
.0 1
.0 1
.4 1
.0 1
.0 1
.0 1
.0 1
.0 1
.0-1
.4 1
.4 1
.4 1
.4 1
.4 1
.0 1
.0
.0
.0
.0
.0
.0 1
.0
.0
.0
.0
!o
.0
.0
.0
.0
.0
.0
.0
.0
1
)
)
)
)
)
>
)
>
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
10
11
12
13
14
15

-------
0.0200 0.0297 0.0333
0.0200 0.0165 0.0184
0.0900 0.1068 0.1167
0.0500 0.1687 0.1862
0.1200 0.1725 0.1911
0.1800 0.1790 0.1988
0.1200 0.1790 0.1988
0.4200 0.2539 0.2779
0.0100 0 1147 0.1676
0.0200 0.0614 0.0991
0.1200 0.0814 0.1007
0.1100 0.0873 0.1133
0.0700 0.0913 0.1133
0.1400 0.0807 0.0986
O.C800 0.0973 0.1481
0.1200 0.1012 0.1504
0.8000 0.0889 0.1310
O.SSOO 0.1010 0.1481
0.2000 0.1047 0.1S05
0.0900 0.0771 0.1089
0.2200 0.0249 0.0211
0.1400 0.02SO 0.0212
0.1100 0.02S1 0.0213
0.1500 0.0263 0.0222
0.2300 0.0250 0.0213
0.2400
0.1SOO
0.1500
0.1900
0. 1500
0.1600
0 . 29*00
0.2000
0.2000
0,2500
0 . 1SOO
0.0700
0.1400
0.1500
0.0200
0.0400
0.1300
0.1300
0.1100
0.1700
0.2200
0.1900
0.1700
0.1400
0.1SOO
0.2400
0.2300
0.3000
0.3400
0.2800
0 . 3200
0.2500
0. 1400
0.1400
.0249 0.0212
.0248 0.0211
.0266 0.0224
.0249 0.0211
.'0252 0.0214
.0266 0.0224
.0268 0.0227
.0263 0.0222
.0250 0.0212
.0252 0.0214
.02S4 0.0215
.0249 0.0212
.0253 0.0216
.0468 0.0554
.0464 0.0551
.0498 0.0596
.0464 0.0550
.0463 0.0550
.0464 0.0551
.0463 0.0550
.0463 0.0550
.0463 0.0550
.0464 0.0552
.0466 0.0554
.0466 0.0554
.0464 0.0552
.0464 0.0552
.0465 0.0553
.0467 0.0555
.0468 0.0557
.0469 0.0559
.0469 0.0559
.0472 0.0563
.01A1 O.0474
0.3&00 0.0471 0.0561
0.3100 0.0472 0.0563
0.2600 0.0473 O.OS64
0.2300 0.0478 0.0571
0.2500 0.0489 0.0584
0.2000 0.0518 0.0620
0.4500 0.0465 O.OS53
0.4400 0.0465 0.0553
0.5200 0 0466 O.OS55
0.4900 0.0467 0.0556
0.5200 0 0472 0.0562
0.4100 0.0472 0.0562
0.3900 0-0474 0.0565
0.3900 0.0475 0.0566
0.4300 0.0478 0.0570
1.1700 0.0464 0.0552
0.4900 0 0484 O.OS78
0.5600 0 0493 0.0589
0.6100 0 0493 0.0590
0.6300 0.0497 0.0595
0.7500 0 0502 0.0600
0.9000 0.0199 0.0597
0.9700 0.0499 0.0596
0.3400 0.0492 0.0588
0.3900 0 OSOO 0.0598
0.4800 0 0505 0.0605
0.3200 0.0515 0.0617
0.3800 0.0524 0.0628
0.4400 0 0510 0.0636
0.4200 0 0538 0.0646
0.3600 0 0541 0.0649
0.4600 0.0552 0 0662
.0501 0.0297 0.0086 0.0667 0.1440 0.0257 0 0603 0.0001 0 0 0 Pb 00 0-999.90 20.00 -999.90 0 35 0.900E*10 -999.900 2.00
.0272 0.0165 0.0116 0.0904 0.2000 0.0097 0.0232 0.0001 0 0 0 Pb 00 0-999.90 20.00 -999.90 0.35 0.900E+10 -999.900
.1546 0.1068 0.1267 0.4353 0.8437 0.0312 0.0812 0 0001
.2642 0.1687 0.0261 0.0172 0.0551 0.1479 0.2861 0.0001
.1175 0.1725 0.0263 0.0399 0.0686 0.1731 0 1841 0.0001
.2907 0.1790 0.0265 A. 0426 0.0682 0.1982 0.3925 0.0001
.2907 0.1790 0.0265 0.0426 0.0682 0.1982 0.3925 0.0001
.4282 0.2539 0.0461 0.0557 0.0758 0.2395 0 4645 0.0001
.0750 0.2835 0.1183 0.0886 0.1398 0.0303 0.0454 0 0001
0412 0.2073 0.0487 0.0375 0.0634 0 0125 0.0187 0.0001
.0421 0.2073 0.0521 0.0402 0.0674 0 0135 0.0201 0.0001
.0484 0.2213 0.0712 0.0539 0.0881 0.0182 0.0273 0.0001
.0484 0.2308 0.0714 0 0541 0.0883 0.0183 0.0274 0.0001
.0409 0.2058 0.0471 0.0362 0.0617 0.0121 0.0181 0.0001
.0655 0.2431 0.1032 0.0774 0.1211 0.0265 0.0396 0.0001
.0666 0.2523 0.1054 0.0790 0.1256 0.0270 0.0404 0.0001
.0571 0.2239 0.0888 0.0668 0 1074 0.0227 0.0340 0.0001
.0655 0.2540 0.1038 0.0779 0 1239 0.0266 0.0398 0.0001
.0666 0.2608 0.1058 0.0793 0 1261 0.0271 0.0405 0.0001
1.0463 0.1969 0.0650 0.0494 0.0811 0.0167 0.0249 0.0001
.0471 0.0249 0.0106 0.0714 0.. 1777 0.0143 0.0394 0.0001
.0473 0.0250 0.0110 0.0738 0 1841 0.0148 0.0409 0.0001
J.0475 0.0251 0.0114 0.0763 0.1911 0.0154 0.0125 0.0001
).0500 0.0263 0.0116 0.0911 0.2307 0.0186 0.0519 0.0001
.0471 0.0250 0.0091 0 0616 0 1514 0.0122 0.0331 0.0001
).0169 0.0249 0.0094 0 0639 0.1572 0.0127 0 0346 0.0001
).0469 0.0248 0.0098 0.0661 0.1617 0.0132 0.0361 0.0001
1.0505 0.0266 0.0140 0.0931 0.2358 0.0191 0.0534 0.0001
.0471 0.0249 0.0106 0.0713 0.1768 0.0143 0.0393 0.0001
1.0478 0.0252 0.0117 0 0785 0.1953 0.0159 0.0439 0.0001
.0505 0.0266 0.0140 0.0931 0.2340 0.0191 0.0533 0.0001
1.0511 0.0268 0.0144 0.0952 0.2388 0.0196 0.0547 0.0001
3.0500 0.0263 0.0136 0.0901 0.2236 0.0185 0.0515 0.0001
1.0472 0.0250 0.0109 0.0725 0.1747 0.0148 0.0404 0.0001
1.0477 0.0252 0.0117 0.0773 0.1874 0.0158 0.0435 0.0001
1.0481 0.02S4 0.0121 0.0799 0.1947 0.0164 0.0451 0.0001
>.0469 0.0249 0.0090 0.0598 0.1388 0.0121 0.0324 0.0001
).0474 0.0253 0.0082 0.0544 0.1230 0.0110 0.0291 0.0001
.00
0 0 Pb 00 0-999.90 20.00 -999.90 0 35 0.900E«10 -999 900 2.00
21 81 FECOH 000 2.50 20.00 -999.90 0.16 0.900CUO 0.000 2.00
23 81 FECOH 000 1.00 20.00 -999.90 0.19 0.900E*10 0.000
1 81 FEOOH 000 3.50 20.00 -999.90 0.23 0.900E*10 0.000
17 91 FECCH 0000 3.50 20.00 -999.90 0.23 0.900EUO 0.000
30 82 FECOH 0000 2.40 20.00 -999.90 0.15 0.900C«10 0.000
26 81 Fin* Prt 10 0 10 30 4.00 9.20 -999.90 0.64-0.499E*03 49.000 10<
26 81 Fin* Prt 17 30 18 0 2.10 15.40 -999.90 0.26 0.105E*03 -16 000 10
26 61 Fin* Prt 18 0 18 30 2.50 14.60 -999.90 0.28 0.104E<03 -20.000 10(
27 81 Fin* Prt 10 30 11 0 2.50 13.20 -999.90 0.38-0.610E«02 85.000 10<
27 81 Fin* Fit 11 30 12 0 2.60 15.00 -999.90 0.38-0 .912E»02 56.000 101
27 81 Fin* Prt 12 30 13 0 1.80 15.20 -999.90 0.25-0 .S72E*02 26.000 101
26 81 Fin* Prt 10 30 11 0 3.60 7.50 -999.90 0.56-0.378E«03 43.000 10
28 81 Flo* Prt 11 30 12 0 2 70 .40 -999.90 0 S7-0.824E»02 209.000 10<
28 81 Fin* Prt 12 0 12 30 2.90 .50 -999.90 0.46-0 .116E*03 69.000 10
28 81 Fin* Prt 12 30 13 0 3.30 .20 -999.90 0.56-0 .709E*02 231.000 10
28 81 fin* Prt 13 0 13 30 2.80 .70 -999.90 O.S7-0.730E>02 237.000 10(
28 61 Fin* Prt 15 0 IS 30 1.50 .70 -999.90 0. 35-0.728E*02 55.000 10
25 78 S 11 18 11 46-999.90 30.00 -999.90 0.24-0 .961E«01 144.000
25 78 S 11 48 12 18-999.90 30.00 -999.90 0.2S-0.116E*02 135.000
25 79 S 12 18 12 48-999.90 30.00 -999.90 0.26-0 .113E*02 1S6.000
25 78 S 12 48 13 18-999.90 30.00 -999.90 0.32-0 .193E*02 170.000
25 78 S 13 16 13 48-999.90 30.00 -999.90 0. 20-0 .463E<01 171.000
25 78 S 13 48 14 16-9"99.90 30.00 -999.90 0.21-0.828E>01 112.000
25 76 S 14 18 14 46-999.90 30.00 -999.90 0.22-0.237E*02 45.000
25 78 S 14 48 IS 18-999.90 10.00 -999.90 0.33-0.3431*02 105.000
25 76 15 18 IS 48-999.90 30.00 -999.90 0. 24-0.136E<02 102.000
25 78 15 48 16 16-999.90 30.00 -999.90 0. 27-0.299E*02 66.000
25 78 16 18 16 48-999.90 30.00 -999.90 0.33-0.692E*02 52.000
25 78 16 46 17 18-999.90 10.00 -999.90 0.34-0.262E*03 15.000
25 78 17 18 17 48-999.90 30.00 -999.90 0.32 0.234E»03 -14.000
25 78 17 48 18 18-999.90 30.00 -999.90 0.25 0.539E*02 -29.000
25 78 16 16 11 48-999.90 30.00 -999.90 0.27 0.481E*02 -41.000
2S 78 18 46 19 18-999.90 30.00 -999.90 0.28 0.610E»02 -36.000
25 76 19 18 19 48-999.90 30.00 -999.90 0.20 0.229E*02 -35.000
25 78 19 48 20 16-999.90 30.00 -999.90 0.18 0.139E+02 -42.000
.1666 0.0670 0.0062 0.0148 0.0382 0.0168 0.1400 0.0001 000 ART 0000 1.84 17.80 -999.90 0.21 0 900E*10 0.000 2
i 1706 0 0662 0 0070 0.0153 0.0436 0.0199 0 1686 0.0001 000 ART 0000 2.22 17.80 -999.90 0.25 0.900E*10 0.000 2
1.1942 0 0705 0.0092 0.0168 0.0579 0.0279 0.2436 0.0001 000 PART 0000 3.22 17.80 -999.90 0.36 O.SOOEHO 0.000 2
1.1686 0.0663 0.0067 0.0151 0.0414 0.0187 0.1573 0.0001 000 PART 0000 2.07 17.80 -999.90 0.21 0.900E«10 0.000 2
1.1698 0.0662 0.0069 0.0152 0.0427 0.0191 0.1641 0.0001 000 PART 0000 2.16 37.80 -999.90 0.24 0.900E*10 0.000 2
1.1713 0.0662 0.0071 0.0154 0.0441 0.0202 0.1716 0.0001 000 PART 0000 2.26 17.80 -999.90 0.25 0.900E*10 0.000 2
1.1690 0.0662 0.0067 0.0152 0.0419 0.0189 0.1595 0.0001 000 PART 0000 2.10 17.80 -999.90 0.24 0.900EtlO 0.000 2
).169« 0.0662 0.0069 0.0152 0.0427 0.0194 0.1641 0 0001 000 PART 0000 2.16 17.80 -999 90 0.24 0.900E«10 0.000 2
1.1699 0.0662 0.0069 0.0153 0.0129 0.0195 0.1648 0.0001 000 PART 0000 2.17 17.80 -999.90 0.24 0.900E»10 .000 2
1.1716 0.0662 0.0071 0.0154 0.0444 0.0203 0.1731 0.0001 000 PART 0000 2.28 17.80 -999 90 0.26 0.900E«10 .000 2
1.1735 0.0664 0.0074 0.0156 0.0460 0.0212 0.1811 0.0001 000 PART 0000 2.39 17.80 -999.90 0.27 O.SOOE.10 .000 2
1.1731 0.0663 0.0073 0.0156 0.0459 0.0211 0.1306 0.0001 000 PART 0000 2.38 17.80 -999.90 0.27 0.900E*10 .000 2
) 1714 0.0662 0 0071 0.0154 0.0443 0.0203 0.1723 0.0001 000 PART 0000 2 27 17.80 -999.90 0.26 0.900EtlO .000 2
1.1721 0.0662 0.0072 0.0155 0.0449 0.0206 0 1753 0 0001 000 PART 0000 2.31 17.80 -999.90 0.26 0.900E*10 000 2
3.1724 0.0663 0.0072 0.0155 0.0451 0.0207 0.1768 0.0001 000 PART 0 0 11 0 2.33 17.80 -999.90 0.26 0.500E.10 .00» 2
3.1743 0.0665 0.0075 0.0156 0.0466 0.0215 0.1843 0.0001 000 PART 0000 2.43 17.80 -999.90 0.27 0.900E«10 .000 2
1.1753 0.0666 0.0076 0.0157 0.0473 0.0219 0.1881 0.0001 000 PART 0000 2.48 17.80 -999.90 0.28 0.900E*10 .000 2
>.1763 0.0667 0.0077 0.0158 0.0480 0.0223 0.1918 0 0001 0 0* 0 PART 0000 2.53 17.80 -999.90 0.28 0.900E*10 .000 2
3.1765 0.0668 0.0077 0.0158 0.0481 0.0224 0.1926 0 0001 000 PART 0000 2.54 17.80 -999.90 0.29 0.900E*10 .000 2
3 1787 0.0671 0.0079 0.0159 0.0496 0.0232 0 2001 0.0001 000 PART 0000 2.64 17.80 -999.90 0.30 0.900E.10 .000 2
1.1843 0.0682 0.0084 0 0163 0.0529 0.0251 0.2173 0.0001 000 PART 0000 2.67 17.80 -999 90 0.32 0.900E*10 0.000 2
1 1776 0.0669 0.0078 0.0159 0.0469 0.0228 0.1963 0.0001 000 PART 0000 2.59 17.80 -999.90 0.29 0.900E»10 0.000 2
1.1787 0.0671 0.0079 0 0159 0.0496 0.0232 0.2001 0.0001 000 PART 0000 2.64 17.80 -999.90 0.30 0.900E*10 0.000 2
>.1791 0.0(72 0.0080 0.0160 0.0499 0 0234 0.2016 0.0001 000 PART 0000 2.66 17.80 -999.90 0.30 0.900E«10 0.000 2
> 1628 0.0679 0.0083 0.0162 0.0520 0 0246 0.2128 0 0001 000 PART 0000 2.91 17.80 -999.90 0.32 0.900EtlO 0.000 2
3.1889 0.0693 0.0088 0.0165 0 0553 0.0264 0 2301 0.0001 0 0 PART 0000 3.04 17.80 -999.90 0.34 0.900E.10 0.000 2
1.2044 0.0731 0.0099 0 0172 0.0625 0.0304 0 2676 0.0001 0 0 PART 0000 3.54 17.80 -999.90 0.40 0.900E+10 0 000 2
».1721 0.0662 0.0072 0.0155 0.0450 0.0207 0 1761 0.0001 0 0 PART '0000 2.32 17.80 -999.90 0.26 0.900E«10 0.000 21
3 1730 0.0661 0.0073 0.0155 0.0456 0 0210 0.1791 0 0001 0 0 PART 0000 2.36 17.80 -999.90 0.27 0.900E*10 0.000 2
3 1739 0.0664 0.0074 0.0156 0.0463 0.0214 0.1828 0.0001 0 0 PART 0000 2.41 17.80 -999.90 0.27 0.900E«10 0.000 2
3 1749 0.0665 0.0075 0.0157 0.0470 0.0218 0.1866 0.0001 0 0 PART 0000 2.46 17.80 -999 90 0.26 0.900E.10 0.000 2
> 1782 0.0671 0.0079 0.0159 0.0493 0.0231 0 1986 0 0001 0 0 PART 0000 2.62 17.80 -999.90 0.29 0.900E«10 0.000 2
1.1765 0.0671 0.0079 0.0159 0 0494 0.0231 0 1993 0.0001 000 PART 0000 2.63 17.60 -999.90 0.30 0.900E+10 0.000 2
1.1801 0 0674 0.0080 0.0160 0.0504 0.0237 0.2046 0.0001 000 PART 0000 2.70 17.80 -999.90 0.30 0.900E+10 0.000 2
1.1806 0.0675 0 OOB1 0 0160 0.0507 0.0239 0.2061 0.0001 000 PART 0000 2 72 17.80 -999.90 0.31 0.900E*10 0.000 2
> 1625 0.0679 0.0083 0.0162 0 0519 0 0245 0 2121 0 0001 000 PART 0000 2.60 17.80 -999.90 0.32 0.900E+10 0.000 2
1 1719 0.0662 0.0072 0 0154 0 0447 0 0205 0.1746 0 0001 000 PART 0000 2.30 17.60 -999 90 0 26 0.900E*10 0.000 2
J.1862 0.0687 0.0086 0.0161 0.0539 0.0256 0.2226 0 0001 000 PART 0000 2.94 17 80 -999.90 0.33 0 900E*10 0.000 2
9 1912 0.0698 0.0090 0 0166 0.0564 0.0271 0 2361 0.0001 000 PART 0000 3.12 17 80 -999.90 0.35 0 900E+10 0.000 2
9.1915 0.0699 0.0090 0.0166 0.0566 0 0271 0.2368 0.0001 000 PART 0000 3.13 17 80 -999.90 0.35 0.900E*10 0.000 2
1 1936 0.0704 0.0091 0.0167 0.0576 0.0277 0.2421 0.0001 000 PART 0000 3.20 17 80 -999 90 0.36 0 900EtlO 0 000 2
1.1960 0.0710 0 0093 0 0169 0 0587 0.0283 0.2481 0 0001 000 PART 0000 .28 17.80 -999 90 0.37 0.900E+10 0.000 2
M94S 0.0706 0.0092 0 0168 0.0580 0.0279 0.2443 0.0001 000 PART 0000 .23 17.60 -999.90 0.36 0.900E+10 0.000 2
1 1942 0 0705 0 0092 0.0168 0.0579 0.0279 0.2436 0 0001 000 PART 0000 .22 17.80 -999 90 0.36 0.900E*10 0 000 2
9.1906 0.0697 0.0089 0.0166 0 0562 0.0269 0.2346 0 0001 000 PART 0000 .10 17.80 -999 90 0 35 0 900E«10 0.000 2
9.1951 0.0707 0.0092 0 0168 0.0583 0.0281 0.2456 0 0001 000 PART 0000 25 17.80 -999.90 0.37 0.900E.10 0.000 2
9.1979 0.0714 0 0094 0 0169 0 0596 0.0288 0 2526 0 0001 000 PART 0000 34 17.80 -999.90 0 38 0.900E.10 0 000 2
9.2031 0.0727 0.0098 0 0172 0 0619 0.0301 0 2616 0 0001 000 PART 0000 50 17.80 -999 90 0 39 0 900E+10 0 000 2
9.2078 0.0740 0.0101 0.0174 0 0639 0.0312 0 2751 0 0001 000 PART 0000 64 17.80 -999 90 0 41 0 900E*10 0,000 2
9.2110 0.0746 0.0103 0 0175 0.0652 0 0319 0.2816 0 0001 000 PART 0000 .73 17 80 -999 90 0 42 0.900E+10 0 000 2
1 2149 0.0758 0.0105 0 0177 0 0666 0 0326 0 2901 0 0001 000 PART 0000 84 17.60 -999.90 0.43 0.900E+10 0 000 2
0 2164 0.0762 0 0106 0 0177 0 0674 0 0331 0 2931 0.0001 000 PART 0000 .86 17.90 -999 90 0 44 0 900E+10 0 000 2
9 2216 0.0776 0.0109 0 0179 0 0694 0.0342 0.3036 0 0001 000 PART 0000 02 17 80 -999.90 0 45 0 900E+10 0 000 2
.00
.00
.00
.00
.00
> 00
).00
.00
>.oo
> 00
1.00
.00
1.00
1.00
>.oo
1.00
.75
.40
.04
.60
.80
.80
.60
.80
.05
.05
.05
.05
.05
.05
.05
.05
.05
.05
.os
.os
.00 10 .10
.00 10 .10
.00 10 .10
.00 10 .10
.00 10 .10
.00 10 .10
.00 10 0.10
.00 10 0.10
.00 10 0.10
.00 10 0.10
.00 10 0.10
.00 10 0.10
.00 10 0.10
.00 10 0.10
.00 10 0.10
.00 10 0.10
.00 10 0.10
.00 10 0.10
.00
1.00
1.00
.00
1.00
1.00
.00
.00
1.00
.00
.00
.00
.00
.00
1.00
.00
.00
.00
.00
.00
.00
1.00
.00
.00
1.00
.00
.00
.00
.00
.00
.00
i.OO
.00
.00
1.00
.00
.00
.00
I 00
00
.00
00
.00
00
.00
.00
.00
) 00
00
00
.00
00
00
0.50
0.50
0.50
o.so
o.so
0.50
o.so ,
0.50
0.50
o.so
0.50
0.50
0.50
o.so
o.so
0.50
o.so
0.50
o.so
o.so
o.so
0.50
o.so
o.so
o.so
o.so
o.so
o.so
o.so
0.50
0.50
o.so
0.50
o.so
o.so
0.50
0.50
0.50
O.JO
o.so
0.50
o.so
0.50
0.50
0.50
0.50
0.50
0.50
0.50
0.50
0.50
0 SO
0.50
.75 1.00-999.90
.40 1.00-999.90
.04 1.00-9.9}. 90
.80 1.00 10.00
.80 1.00 10.00
.80 1.00 10.00
.60 1.00 10.00
.80 1.00 10.00
.10 1.00 39.00
.10 1.00 39.00
.10 1.00 39.00
.10 1.00 39.00
.10 1.00 31.00
.10 1.00 39.00
.10 1.00 39.00
.10 1.00 39.00
.10 1.00 19.00
.10 1.00 39.00
.10 1.00 39.00
.10 1.00 39.00
.00 1.00 39.00
.00 1.00 39.00
.00 1.00 39.00
.00 1.00 39.00
.00 1.00 39.00
.00 1.00 39.00
.00 1.00 39.00
.00 1.00 39.00
.00 1.00 39.00
.00 1.00 39.00
.00 1.00 39.00
.00 1.00 39.00
.00 1.00 3J.OO
.00 1.00 39.00
.00 1.00 39.00
.00 1.00 39.00
.00 1.00 39.00
.00 1.00 39.00
.00 1.00
.00 1.00
.00 1.00
.00 1.00
.00 1.00
.00 1.00
.00 1.00
.00 1.00
.00 1.00
.00 1.00
.00 1.00
.00 1.00
.00 1.00
.00 1.00
.00 1.00
.00 1.00
.00 1.00
.00 1.00
.00 1.00
.00 1.00
.00 1.00
.00 1.00
.00 1.00
.00 1.00
.00 1.00
.00 1.00
.00 1.00
.00 1.00
.00 1.00
.00 1.00
.00 1.00
.00 1.00
.00 1.00
.00 1.00
.00 1.00
.00 1.00
.00 1.00
.00 1.00
.00 1.00
.00 1.00
.00 1.00
.00 1.00
.00 1.00
.00 1.00
.00 1.00
.00 1.00
.00 1.00
.00 1.00
.00 1.00
.00 1.00
.00 1.00
.00 1.00
.00 1.00
0 5900 0.0543 0.0651 0.2171 0.0764 0.0107 0 0178 0 0677 0.0333 0.2946 0 0001 000 PAPT 0000 3.90 17.80 -999 90 0 44 0 900E*10 0 000 28.00 6 0.50 00 1.00
.80
.80
.10
.60
.(0
.10
.60
.60
.10
.10
.10
.60
.60
.60
.to
.60
.60
.10
.60
.80
.10
.60
.80
.80
.80
.60
.60
.80
.60
.60
.80
.60
.60
.80
.10
.60
.80
.60
.60
.80
.80
.80
.80
.80
.80
.60
.60
.80
.80
.80
.8<»
.80
.60
.5
.S
.S
.S
.0
.S
.S
.0
.5
.S
.5
.S
.S
.S
.5
.S
.S
.S
.S
.5
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.«
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
0
.0
.80 .0
.80 9 0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.«
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
_o
.0
.0
,0
.0
.0
.0
.0
.0
.0
.0
.0
.0
o
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
, Q
.0
.0
.»
.0
.0
.0
.0
.0
.0
.0
.a
.<
.0
.0
.0
.0
.0
_o
to
_o
_o
o
'o
.0
1C
17
It
19
19
20
20
21
22
22
22
2]
22
22
22
22
2]
22
22
22
23
21
21
23
23
23
23
13
23
23
23
23
23
23
23
23
23
23
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
14
24
24
14
21
24
24
24
24
24
34
24
24
24
It
24
21
24
.0 S 24
.0 5 24
.0 t 24
.0 5 24
.0 S 24
.0 S 24
.0 S 21
.0 S 24
0 S 24
.0 S 24
.0 S 24
.0 S 24
.0 5 24
0 S 24
.0 5 24
.0 S 24
.0 5 24

-------
p
u
V
11™ S'JHS J'SfS! °-"" 0.0783 0.0111 0.0180 0 0703 0 0347 0.3061 0 0001 000 PART 0000 1.06 17.80 -999.90 0.46 0 900E.10
£M Hill 1 SH? J'«S S-^« S'S'J' °-°185 °-°7" ° "" °-"14 ° °001 ° ° ° PAI!T 0000 4.39 17.80 -999.90 0.49 0.900E.10
i«JJ J Jll? S J™ 0.2153 0.0759 0.0106 0.0177 0 0670 0 0329 0 2908 0.0001 000 PART 0000 3.85 17.80 -999 90 0 43 0 900E»10
JS2J J'S!« J'S'S1 ° "" °-°8" °-0118 °-01" °-0751 °-°"« »•»" 0-«001 0 0 OPART 0000 4.41 17.80 -999?0 0500900E.10
SSIS !'Jf»J !'J!Ji J-1"! "-0"9 °-00" °-0162 °-0520 °-02« °-2128 « 0001 0 0 0 PART 0000 2.81 17 80 -999.90 0.32 0.900E.10
0050 0.0508 0.0608 0.1995 0.0718 0.0096 0.0170 0.0603 0 0292 0.2563 0.0001 000 PART 0000 3.39 17.80 -999.90 oile o!900E*10
•ontb
day
yatr
pollutant
beginning hour
•nding hour
wind ap««4
t**p*ratur*
•oUr radiation
friction valoclty
Honia-Obutthov length
•Ixtng height
roughness length
number of diameter •
•malleat diameter
largest dlaaetar
den*lty
reference height
leaf area index .
LAI correction exponent
land use type
sample aet number
0.000 20.00 € O.SO .00 .00 .80 .0 .0 24
0.000 28.00 « O.SO .00 .00 .80 .0 .0 31
0.000 28 00 < O.SO .00 .00 .80 .0 .0 24
0.000 28.00 € 0.50 .00 .00 .80 .0 .0 24
0.000 28.00 6 O.SO .00 .00 .80 .0 .0 24
0.000 28.00 6 O.SO .00 .00 .80 .0 .0 24

-------
 Table C-2.  The data making up the overall dataset.for a sulfate peaked particle distribution.  The data set contains 173 data points including
 observed zero deposition velocities which have been set to a lower limit of 0.005 cm/s. The footnotes define the variable A thru X.
        Dtpoiltlon Modtl

CARS 1 C«B 2- CARS J CARS 0  ADCH 1 ADOM 2 ACCM 3 U»M 1
.2100 0.9370 1.1215 0.8951 0.9370 2.4105 1.2211 1 2772 1.34S) 1.4027 0.2612 i l» 8} ZnS 22 41 21 11 7 61 14.70 -999.90 0.40 0.166E»01 -999.900 1.00
OSOO 0.9370
.6500 0.9370
.9300 0.8016
.(000 0 BO 16
.7400 0.8016
.1400 0.7901
.0200 0.7904
.1400 0.7904
.7500 0.6B08
.6200 0.6801
.3100 0.6808
.5600 ,0.7157
.4700 0.71S7
.1400 0.7157
.1700 0.87S2
.1500 0.8751
.1000 0.8752
.0400 0.0144
0.1300 0.0102
0.5700 0.0142
0.0300 0.0148
0.1100 0.0162
0.0200 0.0093
0.2000 O.OOS9
0.0900 0.0058
0.0400 0.0051
0.0400 0.0120
0.2400 0.0150
0.3400 0.0124
0.0800 0.0191
0.1000 0.0134
0.0700 0.0076
0.1500 0.0079
O.OSOO 0.0065
0.0200 0.0080
0.1200 0.0061
0.0700 0.0124
0.2300 0.0071
0.1800 0.0053
0.4100 0.0070
0.0400 0.0070
0.0100 0.0048
0.2100 0.0037
0.2200 O.OOS1
0.2200 0.0043
0.0400 0.0073
0.0050 0.0051
0.2800 0.00)8
0.5100 0.0050
0.0400 0.0052
0.0200 0.0083
o.oBoo o.ooie
0.2500 0.0041
0.0100 0.0062
0.3800 0.0053
0.2500 0 0067
0.3500 0.0156
0.0100 0.0072
0.0400 0.0075
0.2100 0 0079
0.6100 0.0498
0.7200 0.0191
0.4400 0.0397
0.3300 0.0402
0.1200 0.0391
0 0050 0.0106
0.0050 0.0405
1.9000 1 6989
1.4000 1.3013
1 0000 0.9363
0.4500 0.5418
0.1500 0.2726
.1215 0.8954 0.9170
.1215 0.89S4 0.9170
.8596 0.7041 0.8016
.8596 0.7041 0.8016
.8596 0.7041 0.8016
.8774 0.7165 0.7904
).8774 0.7165 0.7904
>.8774 0.7165 0.7904
>.7706 0.6414 0.6808
3.7706 0.6414 0.6808
1.7706 0.6434 0.6808
).«<06 .7044 0.7357
).8606 .7048 0.7357
1.8606 .7048 0.7157
).9275 .7524 0.8752
1.1275 .7524 0.8752
.9275 .7524 0.8752
.0182 0.0560 0.0144
.0147 0.0510 0.0102
.0152 0.0195 0.0142
.0211 0.0773 0.0148
.0209 0.0402 0.0162
.0154 0.0238 0 0093
.0096 0.0231 0.0059
.0096 0.0230 O.OOS8
.0081 0.0228 0.0051
.0187 0.0419 0.0120
.0241 0.0418 0.0150
.0168 0.0424 0.0124
.0283 0.0404 0.0191
.0230 0.0194 0.0134
.0127 0.0176 0.0076
.0131 0.0180 0.0079
.0106 0.0172 0.0065
.0137 0.0188 0.0080
.0106 0.0172 0.0061
0187 .0201 0.0124
.0107 .0174 0.0071
.0085 .0178 0.0051
1.0110 .0198 0.0070
1.0122 .022] 0.0070
1.0081 0.021« 0.0048
1.0065 0.0184 0.0017
1.0081 0.0172 0.0051
1.0065 0.0178 0.0043
1.0113 0.0184 0.0073
1.0081 0.0172 0.0051
1.0082 0.0230 0.0048
1.0081 0.0229 0.0050
1.0084 0.0238 0.0052
.0135 0.0232 0.0083
1.0081 0.0111 0.0048
1.0061 0.0130 0.0041
1.0101 0.0136 0.0062
1.0084 0.0133 0.0053
1.0116 0.0127 0.0067
1.0238 0.0525 0.0156
1.0099 0,0166 0.0072
1.0096 0 0152 0.0075
1.0111 0 0466 0.0078
1.0501 0.0434 0 0498
1.0116 0.0369 0 0392
1.0416 0.0369 0.0397
1 0416 0.0369 0.0402
1.0404 0.0360 0.0391
1 0416 0 0369 0.0106
1.0115 0.0369 0.0405
L.7923 2.1277 1.6989
1 3966 1.7S08 1 3011
1 0180 1 3551 0 9361
1.5990 0 8560 0.5418
1 1016 0.4561 0.2726
.4305 1.2241 1.2772 1.1154 .4027 0.2612 5 18 8} ZnS 22 48 21 18 8.53 14.70 -999.90 0.40 0.166E401 -999.900 1.00
.4105 1.2241 1.2772 1.1451 .4027 0.2612 5 18 81 ZnS 22 48 21 18 9.41 14.70 -999.90 0.40 0.166E*03 -999.900 3.00
.9154 0.6852 0.7050 0.9731 .0057 0.1169 5 26 81 ZnS 21 24 2} 54 3.23 19.50 -999.90 0.26 0.440E<02 -999.900 1.00
.9154 0.6852 0.7050 0.9724 .0057 0.1169 5 26 81 ZnS 23 24 21 54 3.59 19.50 -999.90 0.26 0.440E»02 -999.900 3.00
.9151 0.6852 0.7050 0.9721 .0057 0.1169 5 26 83 ZnS 21 24 21 54 1.83 19.50 -999.90 0.26 0.410E402 -999.900 3.00
.0436 0.7185 0.7409 1.0131 .0496 0.1490 6 5 83 ZnS 22 10 22 40 4.74 17.50 -999.90 0.27 0.770E*02 -999.900 3.00
.0116 0.7185 0.7409 1.0131 .0496 0.1490 6 5 83 ZnS 22 10 22 40 5.40 17.50 -999 90 0.27 0.770E«02 -999.900 3.00
0136 0.7185 0.7409 1.0131 .0496 0.1490 6 5 11 ZnS 22 10 22 40 6.32 17.50 -999.90 0.27 0.770E.02 -999.900 3.00
.4911 0 5561 0.5661 0.8207 .8119 0.0822 6 12 83 ZnS 22 43 21 11 1.00 14.90 -999.90 0.20 0.110Ct02 -S99.100 3 00
.4911 0.5561 0.5661 0.8207 .8449 0.0822 6 12 83 ZnS 22 41 21 13 3.39 14.90 -999.90 0.20 0.340E*02 -999.900 1.00
.1911 0.5564 0.5661 0.8207 .8449 0.0822 6 12 81 ZnS 22 11 21 11 1.75 11.90 -999.90 0.20 0.340E<02 -999.900 3.00
.9248 0.6887 0.7089 0.9787 .0129 0.1114 < 24 81 ZnS 21 C 21 28 1.07 14 10 -999.90 0.26 0.590E<02 -999.900 1.00
9248 0.6887 0.7089 0.9787 .0129 0.1114 6 24 81 ZnS 21 6 21 28 3.24 14.10 -999.90 0.26 O.S90E«02 -999.900 3.00
.9248 0.6887 0.7089 0.9787 .0129 0.1314 6 24 81 ZnS 21 6 21 28 1.46 14.10 -999.90 0.26 0.590C<02 -999.900 1.00
.1818 0.8085 0.8171 1.0860 .1266 0.1601 6 27 81 ZnS 21 11 22 1 1.17 20.50 -999.90 0.30 0.710E*02 -999.900 3.00
.3838 0 8085 0 8371 1.0860 1.1266 0.1601 6 27 83 ZnS 21 31 22 1 3.80 20.50 -999.90 0.30 0.710E-02 -999.900 1.00
.1838 0.8085 0.8171 1.0860 1.1266 0.1601 6 27 81 ZnS 21 11 22 1 1.17 20.50 -999.90 0.10 0.710E.02 -999.900 1.00
.0066 0.0160 0.1216 0.0011 0.0118 0.0001 6 1 79 SOI 6000 1.71 14.20 -999.90 0.11 0.144E»02 -999.900 0.80 1
>.0060 0 0118 0 1121 0.0011 0.0094 0.0001 6 21 79 SO4 2 16 0 0 1.21 8.80 -999.90 0.10-0.183E*02 -999.900 0.40 1
1.0151 0.0996 0 2854 0.0075 0 0247 0.0001 6 21 79 SO4 0040 2.53 21.20 -999.90 0.26-0.217E*02 -999 900 1.00 1
1.0037 0.0286 0.0575 0.0020 0.0057 0.0001 6 21 79 SO 4 3 65 0 0 1.64 8.30 -999.90 0.06 O.S57E«01 -999.900 0.30 1
1.0168 0.1084 0.1105 0.0082 0.0271 0.0001 7 18 79 SOI 0 0 3 41 2.69 14.10 -999.90 0.29 0.900E»10 -999.900 2.20 1.
1.0098 0.0660 0.1141 0 0019 0.0092 0.0001 2 19 80 SOI 11 21 0 0 2.15 4.40 -999.90 0.17 0.573E»02 -999.900 0.90 1
1.0158 0.1021 0.1824 0.0077 0.0119 0.0001 2 20 80 SO4 001 40 3.73 5.90 -999.90 0.28-0 .101O01 -999.900 0.30 1
1.0114 0.0750 0.1295 0.0056 0.0107 0.0001 2 21 80 SOI 9 55 0 0 3.11 5 00 -999.90 0.20 0.185E<02 -999.900 0.10 1
1.0142 0.0919 0.1617 0.0069 0 0111 0.0001 2 22 80 SOI 004 20 3.81 6.00 -999.90 0.25 0.900E>10 -999 900 0.20 1
1.0191 0.1235 0.1512 0.0091 0.0311 0.0001 10 17 79 SOI 6 54 0 0 3.51 7.80 -999.90 0.14 0.245E>03 -999.900 1.50 1
1.0081 0.0559 0.1442 0.0041 0.0110 0.0001 10 18 79 SOI 7000 1.53 5.10 -999.90 0.14 0.204E>02 -999.900 2.10 1.
1.0200 0.1279 0.1664 0.0097 0.0124 0.0001 10 19 79 SOI 004 18 1.81 11.00 -999.90 0.15-O.lllEtOl -999.900 1.10 1'
1.0105 0.0708 0.1958 0.0052 0.0168 0.0001 10 19 79 SOI 6 58 0 0 1.50 8.80 -999.90 0.18 0.913E402 -999.900 4.30 1
1.0081 0.0565 0.0775 0.0011 0.0059 0.0001 11 21 79 SOI 10 50 0 0 1.48 2.90 -999.90 0.11 0.8S8E«02 -999.900 2.00 1
1.0109 0.0710 0.1011 0.0051 0.0078 0.0001 11 22 79 SOI 002 55 2.10 4.50 -999.90 0.19-0.167E»01 -999.900 0.60 1.
>.0097 0.0652 0.0892 0.0049 0 0070 0 0001 11 27 79 SOI 10 51 0 0 2 52 4.50 -999 90 0.17 0.241E*02 -999.900 0.60 1
1.0131 0.0852 0.1185 0.0064 0.0091 0 0001 11 29 79 S04 0047 2.31 5.20 -999.90 0.23 0.661E>02 -999.900 0.40 1.
1.0086 0.0583 0.0791 0.0041 0.0062 0.0001 11 29 79 SOI 11 1 0 0 2.41 3.10 -999.90 0.15 0.17SE»02 -999.900 0.60 1.
1.0110 0.0847 0.1176 0.0061 0 0091 0.0001 12 1 79 SO4 11 9 0 0 1.45 2.90 -999.90 0.21 0.416E*02 -999.900 0.40 1
1.0076 0.0527 0.0714 0.0019 0.0055 0 0001 12 6 79 SOI 11 12 0 0 1.78 7.60 -999.90 0.11 0.241E*02 -999.900 1.10 1
1.0115 0.0760 0.1052 0.0057 0.0082 0 0001 12 7 79 SOI 002 35 2.81 8.00 -999.90 0.20 O.UOEiOJ -999.900 0.40 1.
.0175 0.1112 0.1SS7 0.0085 0.0125 0.0001 12 12 79 SOI 11 20 0 0 4.72 6.70 -999.90 0.11 0.120E«01 -999.409 0.20 1'
.0231 0.1447 0.2045 0.0111 0.0161 0 0001 1 22 80 SOI 0 0 1 18 6.01 7.70 -999.90 0.41 0.245E*03 -999.900 0.30 1
.0064 0.0451 0.0599 0.0011 0.0047 0.0001 1 23 80 SOI 10 59 0 0 2.07 1.70 -999.90 0.11 0.1S2E<02 -999.900 0.30 V.
.0059 0.0417 0.0550 0.0010 0.0041 0.0001 1 21 tO SOI 11 1 0 0 2.19 1.40 -999.90 0.10 0.176E«02 -999.900 0.10 13
.0091 0.0615 0.0819 0.0016 0.0066 0.0001 1 28 80 SO4 0 0 5 4 2.75 1.20 -999.90 0.16 0.858E*02 -999.900 0.10 1
.0126 0.0821 0.1139 0.0062 0.0090 0.0001 1 29 80 SCI 10 12 0 0 3.47 7.30 -999 90 0.22 0.215E.03 -999.900 0.20 1.
.0175 0.1112 0.1556 0.0085 0.0125 0.0001 1 30 80 SOI 0047 5.02 8.20 -999.90 0.31 0.9002*10 -999.. 900 0.10 11
1.0091 0.0625 0.0855 0.0016 0.0067 0.0001 1 10 80 SO4 10 49 0 0 2.40 6 70 -999.90 0 16 0.50SE«02 -999.900 0.40 1
1.0125 0.0819 0 1115 0.0062 0.0090 0.0001 2 11 80 SOI 11 22 0 0 1.46 5.90 -999.90 0.22 0.162E*01 -999 900 0.20 1.
1.0111 00751 0.1301 0.0056 0.0107 0.0001 1 5 80 SOI 9 59 0 0 3.12 3. 20- -999. 90 0.20 0.194E«03 -999.900 0.2012
1.0120 0.0798 0.1171 0.0059 0.0112 0.0001 3 6 80 SOI 0 0 4 16 1.21 5.30 -999.90 0.21 0.900E.10 -999.900 0.20 13
.0098 0.0654 0.1120 0.0019 0.0091 0 0001 3 6 80 SOI 9 56 0 0 2.70 4.70 -999 90 0 17 0.779E*02 -999.900 0.20 1.
0109 0.0727 0.1267 0.0051 0.0101 0.0001 1 11 80 SOI 10 3 0 0 2.52 5.20 -999.90 0.19 0.952E*02 -999.900 0.70 12
1 0101 0 0691 0.0745 0.0051 0.0056 0.0001 1 25 80 £31 004 56 2.51 2.90 -999 90 0. 18-0 .630E«02 -999.900 0.20 12
i 0109 0.0718 0.0774 0 0054 0.0059 0.0001 2 IS 80 SOI 0 0 5 39 1.10 7 10 -999 90 0 19 0.120E*03 -999.900 0.10 1.
1.0092 0.0631 0.0680 0.0016 0.0050 0.0001 2 27 80 SOI 0 0 1 52 2.06 5.10 -999.90 0. 16-0 .418E>02 -999 900 0.10 1.
1.0098 0 0664 0 0715 0 0019 0.0053 0.0001 2 29 80 SOI 00142 30 5.90 -999.90 0. 17-0 .289E<02 -999.900 0.20 13
1.0136 0.0888 0 0961 0 0067 0 0073 0.0001 3 10 80 SOI 9 59 0 0 1 01 1 60 -999.90 0 24 0.24SE>03 -999.900 0.50 1
1.0059 0.0119 0.0976 0 0031 0.0091 0 0001 10 21 79 SOI < 52 0 0 1 77 6.60 -999 90 0.10 O.B86E»01 -999 900 1.30 1,
1.0099 0.0663 0.1925 0.0019 0.0178 0.0001 6 9 80 SOI 5 11 0 0 2.45 11.20 -999.90 0.17 0.162E<01 -999.900 0.10 1.
).0156 0.1001 0.3077 0.0076 0.0283 0.0001 6 10 80 SOI 0055 1.58 11.70 -999 90 0.27-0 .988E<02 -999 SO* 0 30 1"
1.0099 0.0665 0.1951 0.0049 0 0177 0.0001 6 12 80 SOI 5 12 0 0 2.28 10.00 -999 90 0 17 0.900E*10 -999 900 0.40 1
.0068 0.0197 0.1236 0.0015 0.0091 0.0001 9 17 79 5 15 5 15 30 1.83 23.60 -999.90 0. 11-0 .256E+01 51.000 9.40 1.
1 0091 0 0617 0 1615 0.0015 0.0126 0.0001 9 25 79 S 11 IS 12 0 1 28 21.50 -999.90 0. 15-0 ,171E*01 88 000 9.40 13
1.0091 0 0617 0 1611 0 0045 0.0126 0.0001 9 25 79 S 12 5 12 30 1 41 22 00 -999.90 0 15-0 470E*01 70.000 9.40 1
1.0091 0 0637 0.1632 0 0045 0 0126 0 0001 9 25 79 S 12 35 13 0 1 31 22 50 -999.90 0. 15-0 -471E.01 70 000 9 40 1.
1.0097 0 0670 0 1713 0.0048 0.0131 0.0001 9 25 79 S 13 5 11 30 1.29 22.90 -999 90 0. 16-0 .977E.01 11.000 9 40 12
> 0091 0 0637 0 1625 0 0045 0 0126 0 0001 9 25 79 S 14 15 15 0 1 57 22 90 -999.90 0. 15-0.622E>01 53 000 9.40 1,
1.0091 0 0633 0.1599 0 0015 0 0126 0 0001 9 25 79 S 15 15 16 0 1 53 22.90 -999.90 0 1S-0-1SOE.02 22.000 9.40 I.
> 1279 1 1775 1 6611 2 1045 2.4431 0 2520 0 0 0 Pb 000 0-999.90 20 00 -999.90 0.35 0 900E«10 -999.900 2.00 1
1.1767 0.6268 0 9671 1 6724 2 0915 0 115B 0 0 0 Pb 000 0-999 90 20.00 -999.90 0 35 O.SOOEtlO -999 900 2 00 1
> 1128 0.1031 0.4725 1.2507 1.7113 0.0809 0 0 0 Pb 000 0-999 90 20.00 -999 90 0.35 0.900E«10 -999.900 2.00
1 0800 0.1220 0.1851 0.7495 1.2183 0 0001 0 0 0 Pb 000 0-999.90 20 00 -999 90 0.3S 0 900E>10 -999 900 2.00
1.0146 0.0615 0 1017 0 1721 0 7257 0 0001 0 0 0 Pb 000 0-999.90 20 00 -999 90 0 35 0 900E«10 -999 900 2 00 1
6.00
6.00
(.00
6.00
6.00
(.00
i.OO
(.00
(.00
6.00
(.00
(.00
(.00
(.00
(.00
(.00
(.00
(.00
o.io
o.io
1 0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0 10
0.10
0.10
0.10
0.10
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
00
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
0.10 1.00
0 10 1.00
11.00 13.00
10.00 10.00
7.50 7.50
5.00 S 00
.00-999.90 0.
.00-999.90 0.
.00-999.90 0.
.00-999.90 0.
.00-999.90 0.
.00-999.90 0.
.00-999-90 0.
.00-999.90 0.
.00-999.90 0.
.00-999.90 0.
. 00-999.10 0.
.00-999.10 0.
.00-191.10 0.
.00-111.10 0.
.00-999.90 0.
.00-911.10 0.
.00-199.90 0.
.00-991.10 0.
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
.90
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00 J.
.00 1.
.00 2.
.00 2.
.00 2.
.00 1.
.00 1.
.00 1.
.00 1.
.00 2.
.00 2.
.00 2.
.00 2.
.00 0.
.00 0.
.00 0.
.00 0.
.00 0.
.00 o.
.00 o.
.00 0.
.00 0.
.00 0.
.00 0.
.00 0.
.00 0.
.00 0.
.00 0.
.00 0.
.00 0.
.00 1.
.00 1.
.00 1.
.00 1.
.00 0.
.00 0.
.00 o.
.00 0.
1.
1.
1.
1.
1.
1.
I.
1.
1.
t.
1.
1.
I.
1.
1.
1.
1.
1.
1.
1.
1.
1.
1.
1.
1.
1.
I.
1.
1.
1.
1.
1.
1.
1.
1.
1.
1.
1.
1.
t.
1.
1.
t.
1.
11 1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1

1<
1<
1<
11




1
11
1(
11
1<
1(
1(
1(
1<
11
11
11
1<
u
1<
u
11
1.0 1(
1.0 1C
1.0 1<
1.0 1<
1.0 1<
1.0 1(
1.0 1C
1.0 11
1.0
1.0
1.0
1.0
.00 0.1 1.0
.00 2.5 1.0 1
.00 3.0 1.0
.00 1.00 1.0 1.0
.00 1.00 1.0 1.0
.00 7.00 2.0 1.0
.00 7.00 2.0 1.0
00 7.00 2.0 1.0
.00 7.00 2.0 1.0
.00 7.00 2.0 1.0
.00 7.00 2.0 1.0
.00 7.00 2.0 1.0
.00-999.90 1.5 1.0
.00-999.90 1.5 1.0
.00-999.90 1.5 1.0
.00-999.90 1 S 1.0
1.20 1.20 1.00-999.90 1.5 1.0
0.0400 0 0886 0 0994 0 1517 0.0886 0.0122 0 0446 0.0967 0 1060 0 2376 0 0001 0 0 0 Pb 000 0-999.90 20 00 -999.90 0.3S 0.900EllO -999 900 2 00 1 1.60 1.60 1.00-999.90 1.5 1.0
I 1
1
L 1
L 1
1
1
L
L
L



L
L

L


>
1





1
1

>
)
I

)


































10
11
12
11
14
IS

-------
tz
tz
tz
tz
tz
tt
tt
It
It
tt
tt
tl
tt
tt
tt
tz
IZ
tt
It
It
11
tl
tl
tl
tt
tt
tt
tt
tz
tz
It
It
It
It
It
tt
It
tl
tt
tt
tl
tt
It
11
11
It
tl
tl
11
tl
tl
tl
tl
tt
It
tl
tl
tt
tt
tt
tt
tt
tt
tt
tt
tl
tl
tl
tl
tl
tt
tt
tt
zt
zt
ZI
zt
zt
zt
tt
11
zz
ZI
zt
zt
Tt
ot
ot
6T
61
Bl
tl
n
0'
0'
o;
0-
0-
0-
0-
0*
0'
0'
0'
0'
0'
o-
0"
o-
o-
o-
o-
0*
0'
o-
0'
0'
0'
0'
0'
0'
O'l
O'l
O'l
•1
•
•1
•1
•1
•
0'
0'
0'
0'
O'l
O'l
O'l
O'l
O'l
O'l
O'l
O'l
O'l
O'l
O'l
O'l
O'l
O'l
O'l
O'l
0'
O'l
O'l
O'l
O'l
O'l
O'l
O'l
O'l
0''
O'l
O'T
O'T
O'T
O'T
O'T
O'l
O'l
O'l
O'T
O'T
O'l
0'1
O'l
I 0
I 0
I 0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0'
0
0
0
I 0
0
0'
0'
0'
0'
0'
0'
»•
0'
0'
«•
0'
0'
0'
0'
0'
0'
0'
0'
0'
0'
0'
0'
• •
0'
0'
0'
0'
0'
0'
0'
0'
• o-
0'
0'
0'
0'
0'
0'
0'
o-
o-


O'T
O'T
O't
O't
O'T
O'T
O'l
6 oe-
6 OB-
OB'
OB-
OB'
oe
08'
08
OB-
OB'
OB-
OB
OB-
OB'
08
08
os-
os-
OB-
OB'
oe
' OB-
OB'
08
OB'
OB-
OB'
OS-
OB.'
08
08
or
OS-
OB
OB-
OB'
Ofl*
OB*
OB-
OB'
« 08
OB-
OB'
• OB-
OB'
OB-
OB'
OB-
OB'
08*
OB-
OB'
OB'
OB-
OB'
00' 1
OO'I
OO'I
OO'I
OO'I
OO'I
OO'I
OO'I
OO'I
OO'I
OO'I
OO'I
oo- 1
t OO'I
I 00'6
t 00'6
t oo- 1
t OO'C
o oo't
0 00' 6
0 OO'C
0 OO'C
0 OO'C
0 OO'C
0 OO'C
0 OO'C
0 OO'C
0 OO'C
> 00'6
S 00*1 00 t 05*0
00 I 00*1 05*0
00*1 00*1 OS 0
00*1 00*1 OS'O
OO'l OO'l OS'O
OO'l OO'l OS'O
. OO'l OO'l OS'O
OO'l OO'I OS 0
OO't OO't 05*0
OO'l OO'l OS'O
OO'l OO'l OS'O
OO'l OO'l OS'O
OO'I OO'l OS'O
OO'I OO'I OS'O
OO'I OO'I OS 0
OO'l OO't 05*0
00*1 OO'l OS'O
OO'T OO'T OS'O
OO'T OO'T OS'O
OO'I OO'I OS'O
OO'T OO'T OS'O
OO'T 00*1 OS'O
•O'T 00*1 OS'O
OO'I OO'I OS'O
00'! OO'I OS'O
OO'T OO'T OS'O
OO'T OO'T OS'O
00 'I 00- 1 OS'O
OO'T OO'T OS'O
00 'I OO'T OS'O
OO'I OO'I OS'O
OO'T OO'I OS'O
OO'T OO'I OS'O
00 I OO't OS 0
OO'l OO't OS 0
OO'T OO'T 05*0
OO'I 00 I OS'O
OO'T 00 t OS'O
OO't OO'I OS'O
OO'T OO'I OS'O
OO'T OO'T OS'O
OO'T OO'T OS'O
OO'T OO'T OS'O
OO'I OO'l OS'O
OO't OO'I OS'O
OO'T OO'T OS'O
OO'I OO'I OS'O
OO'I OO't OS'O
OO't OO'T OS'O
OO'T 00 I OS'O
OO'T OO'T OS'O
OO'T OO'T OS'O
OO'T OO'T OS'O
00-t OO'T OS'O 9
OO'I OO'T OS'O 9
t OO'T OO't OT'O Z
t OO'T OO'I OI'O I
t OO'I OO'T OT'O t
t OO'I OO'T OI'O Z
t OO'I OO'T OI'O Z
t OO'T OO'T OI'O I
t OO'T OO'I OT'O I
t OO'I OO'T OT'O Z
t OO'T OO'T 01*0 Z
t OO'T OO'T OI'O Z
t OO'T OO'T OT'O Z
t OO'T OO'I OT'O Z
t OO'I OO-I OT'O I
t OO't 00-T OfO t
t OO'I OO'l OfO Z
t OO't OO'T OI'O I
t OO'I OO'T OI'O t
t OO'T OO'T OT'O t
t OO'T OI'O SO'O
t OO'T OT'O SO'O
t OO'T OT'O SO'O
t OO't OT'O SO'O
t OO'T OT'O SO'O
t OO't OfO SO'O
t OO't OI'O SO'O
t OO'I OT'O SO'O
t OO'T OT'O SO'O
t OO'I OfO SO'O
t 00 'I OfO SO'O
9 00 8Z 000 0 Ot*3006'0 tt 0 06 666- 08 tt 68 t 0000 iaYd 0 0
9 00 BZ 000 0 01*3006 0 tt'O 06 666- 08 tl 06'E 0 00 idYd 0 0
9 00 BZ 000 0 01*3006 0 St 0 06' 666- 08 tt ZO't 0 00 XdYd 0
9 00 BZ OOO'O 01*3006 0 tl 0 06 666- 08 It 8B C 0 00 XdYd 0 1
9 OO'flt 000 0 Ot*3006'0 tt 0 06'666- 08 tt 18 E 0 00 X8Yd 0
9 00 8Z OOO'O Ot«3006'0 ZfO 06'666- 08 t Et t 0 00 iaYd 0
J 00'8Z 000 0 01*3006 0 TI'O 06*666- Ofl't t9 t 0 00 iaYd 0
9 00 SZ 000 0 01*3006 0 6f 0 06'666- 08*t OS E 0 00 idYd 0
9 00 BZ 000 0 01*3006 0 BE'O 06 666- 08 t tt E 0 00 XdYd 0
9 OO'BZ 000 0 01*3006 0 tE 0 06'666- 08 t SZ t 0 00 XdYd 0
OO'SZ OOO'O Ol'3006'O SfO 06'666- 06'tt 01 t 0 00 X8Yd 0 0
00*et 000*0 01*3006*0 9E 0 06*666- 08 tl ZZ E 0 00 X8Yd 0 0
00 BZ 000*0 01*3006*0 9E*0 06*666- 08 tt Eft 0 00 iBYd 0 0
00 SZ 000 0 01*3006*0 tf 0 06'666- Ofl'tt 9Z t 0000 i8Yd 0 0
OO'SZ 000 0 OI*3006'0 9f 0 06 666- OB'tl OZ t 0000 iaYd 0 0
00 8Z 000 0 01*3006 0 SfO 06 666- OB'tl tt E 0000 XaYd 0 0
00 8Z 000 0 Ot*3006'0 SE'O 06*666- 00 tt Zft 0000 iaYd 0 0
OO'BZ 000*0 01*3006*0 tt'O 06'666- 00 tt t6'Z 0000 iaYd 0 0
00 BZ 000 0 Ol*3006'0 9Z 0 06'666- OB'tl Ot Z 0000 i8Yd 0 0
OO'BZ OOO'O OI*3006'0 ZfO 06'666- OB'tt OB Z 0000 X8Yd 0 0
OO'BZ OOO'O Ot«3006'0 tt'O 06'666- OB'tt Zf Z 0000 iaYd 0 0
OO'BZ OOO'O Ot*3006'0 OfO 06"666- Ofl'tt OfZ 0000 iavd 0 0
OO'flZ OOO'O Ol»3006'0 Ot 0 06'666- 08'tl t9'Z 0000 idYd 0 0
OO'eZ OOO'O 01*3006 0 6fO 06 666- 08 tl Z9 Z 0000 idYd 0 0
00-8Z OOO'O 01*3006 0 BfO 06'666- 08 tt 9fZ 0000 iaYd 0 0
00 BZ 000 0 Ot*3006'0 tt'O 06"666- OB'tt It t 0000 iaYd 0' 0
OO'BZ 000 0 01*3006-0 tZ'O 06'666- OB'tl 9t Z 0000 iaYd 0 0
OO'BZ 000 0 01*3006-0 9fO 06' 666- OB'tl tf Z 0000 iaYd 0 0
OO'Bt OOO'O Ot»3006'0 OfO 06'666- 08 tt IS't 0000 iaYd 0 0
OO'Bt OOO'O Ot*3006'0 tt'O 06 666- 08 tl tO't 0000 iaYd 0 0
OO'St OOO'O Ot*3006"0 tf 0 06'666- 08'tl 18't 0000 i8Vd 0 0
OO'St OOO'O Ol*3006'0 OfO 06 666- OB'tl 99't 0000 i8Yd 0 0
00 8t 000 0 01*3006 0 OfO 06'666- OB'tt 19'Z 0000 i8Yd 0 0
OO'eZ 000 0 01*3006*0 6Z*0 06*666- OB'tt 6S*Z 0000 iaYd 0 0
OO'BZ 000 0 01*3006*0 Zt 0 06 666-1 08*11 1B*Z 0000 i«Yd 0 0
OO'BZ OOO'O 01*3006' 0 OfO 06' 666- 08'tl t9'Z 0000 i8»d 0 0
OO'BZ OOO'O Ot*3006 0 61'0 06 666- 08'tt IS Z 0000 i8Yd 0 0
00 Bt OOO'O 01*3006 0 BZ'O 06*666- 08 tl tS*t 0000 i'dYd 0 0
OO'Bt OOO'O 01*3006 '0 BZ'O 06'666- OB'tl Bt Z 0000 i8YJ 0 0
OO'BZ OOO'O 01*3006 0 tt'O 06'666- OB'tt tft 0000 iaYd 0 0
OO'St 000 0 Ot*3006'0 9fO 06'666- OB'tt tft 0000 i8Yd 0 0
OO'BZ OOO'O OI»3006'0 9Z'0 06'666- OB'tl If Z 0000 i8Yd 0 0
OO'BZ OOO'O 01*3006' 0 9Z'0 06'666- OB'tt IZ'Z 0000 i8Yd 0 0
OO'BZ 000 0 01*3006 '0 tZ'O 06'666- OB'tl Bt t 0000 MYd 0 0
OO'BZ 000*0 01*3006 '0 IZ'O 06'666- OB'll 6fZ 0000 X8Yd 0 0
OO'BZ OOO'O 01*3006*0 9t'0 06*666- OB'tl 8t t 0000 iaYd 0 0
OO'St OOO'O 01*3006 0 tt'O 06'666- OB'tl tft 0000 i»Yd 0 0
OO'SZ OOO'O OI«3006'0 tfO 06*666- 08*11 9ft 0000 i»Yd 0 0
OO'St OOO'O 01*3006'0 tt'O 06'666- OB'tt OI'Z 0000 i8Yd 0 0
00'8Z 000*0 01*3006*0 SZ'O 06'666- OB'tt 9Z'Z 0000 iaYd 0 0
OO'BZ 000 0 01*3006' 0 tZ*0 06'666- 08'tl 9t'Z 0000 iaYd 0 0
OO'BZ OOO'O OI'3006'O EZ'O 06'666- 08 11 tO'Z 0000 iaYd 0 0
OO'flt OOO'O 01*3006 '0 9f 0 06'666- 08'tt Zft 0000 i»Yd 0 0
OO'Bt OOO'O 01*3006-0 5fO 06'666- OB'tl ZZ'Z 0000 iaYd 0 0
OO'BZ OOO'O 01*3006-0 tfO 06'C66- OB'tl tB'T 0000 i»Yd 0 0
I OO'I OOO'tt- 10«36ET'0 81*0 06*666- OO'Ot 06'666-BI OZ 8t 61 S 81 St
I OO'T OOO'Sf- tO*36tt'0 OfO 06'666- OO'Ot 06'666-Bt 61 81 6t S 8t St
T OO'I 000'9t- ZO»30I9'0 8Z'0 06'666- OO'Ot 06'666-flT 61 Bl Bt S Bt SZ
5 22'5 000*11- ZO»3tBt*0 tZ*0 06*666- OO'Ot 06*666-8t Bt 81 Bt S flt St
I 00 I 000*61- tO*36tS'0 SZ'O 06'666- OO'Ot 06*666-«t fll Bt tl S flt SZ
1 00*1 OOO'tl- tO*3ttl'0 ZfO 06'666- OO'Ot 06'666-flt tl 81 tt S 81 SZ
T OO'l OOO'Sl tO*3Z9Z*0-tt*0 06*666- OO'OE 06*666-BT 11 8t 91 S 81 SZ
I OO'l OOO'ZS ZO*3Z69'0-tf 0 06*666- OO'Ot 06'666-Bt 91 81 91 S 8t St
1 OO'I 000'99 tO»366fO-tZ-0 06'666- OO'Ot 06'666-Bt 91 81 St S Bt SZ
1 OO'I OOO'tOl ZO*39tfO-tZ-0 06*666- 00 Ot 06'666-Bt 51 81 '51 S Bt SZ
1 OO'l OOO'SOt tO»3Ett'0-tt'0 06'666- 00 Ot 06'666-Bt St flt II S flt St
I 00*1 000*51 tO»3tEZ*0-tt*0 06'666- OO'Ot 06'666-flt tl Bt tl S flt St
1 OO'l OOO'ttt lO*3fltfl'0-tZ'0 06 666- OO'Ot 06*666-81 tl St El S 81 SZ
I OO't OOO'ttl lO»3E9fO-OZ-0 06*666- OO'Ot 06'666-8t tt 81 tl S Bt St
1 OO'l 000*011 I0»3f6f 0-tf 0 06-666- OO'Ot 06'666-8t tl 81 ZI S Bt SZ
1 OO'l 000-9S1 tO*3tll-0-9t 0 06-666- OO'Ot 06'666-BI tl fll tt S flt St
I OO'I OOO'Stl I0*39lf0-Sf 0 06*666- 00 Of 06'666-BT Zt fll tl S flt SZ
t OO'I OOO'ttl 10+3196-0-lZ-O 06'666- 00' t 06 666-flt 11 Bt tl S flt SZ
OO'OOt OOO'SS ZO*38Zf 0-Sf 0 06'666- Of OS'l Ot St 0 51 Md •UIJ 18 BZ
OO'OOT OOO'ttZ Z0«30tf 0-tS*0 06*666- Of OB'Z Ot ft 0 tt Md •"U 18 BZ
OO'OOT 000 IfZ tO*360f 0-95*0 06*666- Of Oft tl Ot tl Md »uu Ifl flt
00*001 000*68 E0«39ll*0-Bt*0 06'666- OS 06'Z E ZI 0 Zt Md •"IJ 18 8Z
OO'OOT 000'60t ZO»3tZe'0-tS'0 06'666- Of OfZ tl Ot tt Md •»TJ te Bt
OO'OOT OOO'tt tO*38tf O-95'O 06'666- OS' 09' t II Ot 01 MJ •«TJ IB flt
00 001 000'9l tO*attS*0-St*0 06*666- OfST Ofl'I tt Ot ZT Md •"1J IB tt
OO'OOT 000 95 tO*3ZE6'0-8t*0 06'666- OO'ST 0»*t 0 IT Ot IT Vi ""IJ IB It
OO'OOT OOO'SB tO»30I9'0-8t'0 06'666- Oftt OS't 0 Tl Ot 01 Md •"1J 18 tt
00*001 000 Ot- tO*3IOl*0 8Z*0 06*666- 08*11 OS'Z Ot Bt 0 flt Md W1M 18 9Z
00*001 000*91- fO*3SOf 0 9Z*0 06*666- OfSl OfZ 0 BI Ot tl Md •ulj Tn or
l> OO'Ct OO'T OT'O SO'O 00*001 000*61 tO«366t*0-t9*0 06*666- OZ*6 OO't Ot Ot 6 Ot Md -raij IB 9Z
I OO'OT OO'I OB't OB't T OO'Z OOO'O OI»3006'0 SfO 06'666- OO'OZ Oft 0000 HCO3J 18 Ot
t 00-OT OO-I Ofl't 08-Z T OO't OOO'O Ot*3006'0 tt'O 06'666- OO'Ot OS'E 0000 HO03J IS tl
1 00*01 OO'T OB't Ofl't T OO'Z OOO'O 01*3006' 0 EZ'O 06 666- OO'Ot OS't 0000 HOO3J 18 I
: OO'Ot OO'l OB't Ofl't I OO't 000 0 OI*3006'0 61'0 06'666- OO'Ot OO't 0000 H0033 IB ft
OO'OT OO't OB't OB'l I OO't OOO'O Ot«3006'0 91'0 06' 666- OO'Ot OS'Z 0000 HC-33J IB IZ
I 06'666-00'T lO'O tO'O T OO'Z 006 666- OI*3006'0 SfO 06'666- OO'OZ 06'666-0 000 qd 0 0
I 0«'C6C-00'I OfO OfO T OO'l 006'666- Ol*3006'0 SfO 06'666- OO'Ot 06'666-0 000 qd 0 0 0
t 06'666-00'I St'O SfO t OO'Z 006'666- Ot*3006'0 SfO 06*666- OO'OZ 06*666-0 000 qd 0 0 0
0 tOOO 0 6E6Z 0 ZEEO 0 9190 0 ttlO 0 9010 0 E9tO*0 B9IZ 0 0590 0 ZtSO'O 0016*0
0 tOOO 0 9t6Z 0 tttO'O t!90*0 BttO'O tOtO'O I9t0'0 tltZ'O 1590'0 EtSO 0 0065'0
0 1000 0 t£0t 0 EtEO 0 1690 0 6ttO 0 6010 0 91tO*0 tIZZ'O Z990*0 ZSSO 0 0091*0
0 1000 0 IE6Z 0 tEEO 0 1190 0 ttlO 0 9010*0 Z910 0 t9IZ* 6t90*0 ttSO 0 009E*0
0 tOOO 0 t06Z 0 BZEO 0 8990 0 tltO'O SOtO 0 SStO'O 6tIZ' 9t90'0 8tSO 0 OOZt'O
0 1000 0 6I8Z 0 61EO 0 ZS90*0 SttO 0 tOtO'O 8ftO*0 OltZ* 9E90*0 OESO'O OOft'O
0 1000 0 IStZ 0 ZtEO 0 6E90 0 tttO 0 1010*0 0110*0 B10Z* 8Z90*0 tZSO 0 OOBfO
0 1000 0 9t9Z 0 tOtO'O 6190 0 ZIIO'O 8600*0 IZIO'O IEOZ 1190 0 StSO 0 OOZE'O
0 tOOP'O 9Z5Z*0 88ZO*0 9650 0 6910 0 1600'0 tttO'O 6161* 5090 0 5050 0 0081*0
0 1000*0 6SfZ*0 18ZO 0 £850*0 8910 0 Z600*0 tOtO'O tS6f 8650 0 OOSO'O 006f 0
0 tOOO 0 9ttZ 0 69ZO 0 ZSSO'O 9910 0 6800'0 t690'0 906f 0 BBSO'O Z6tO*0 OOtfO
0 tOOO 0 9EIZ 0 6tZO'0 6150 0 8910*0 Z600*0 5010*0 Zt6t*0 9650 0 8610 0 0016 0
0 1000*0 lllZ 0 6tZO 0 OBSO 0 6910 0 Z600*0 9010'0 5161 0 1650 0 66t0'0 0006 0
o tooo o tsiz'o tezo'o easo o 6910-0 teoo*o 0110*0 096fo 0090-0 zoso*o oost* o
0 TOOO 0 IZtZ'O IIZO'O 9150'0 19t0'0 1600'0 tOtO'O 9E61'0 5650*0 1610*0 OOC9*0
0 1000*0 69EZ 0 ttZO 0 99SO 0 9910*0 0600*0 6690*0 SI6f 0 06SO'0 E6tO*0 0019 0
0 1000*0 19EZ 0 ttZO 0 S9SO 0 9910*0 0600*0 B690*0 ZI6t 0 6850*0 £610*0 0095*0
0 1000*0 9ZtZ 0 9SZO 0 6tSO 0 1910 0 9800*0 t890 0 Z98fO 8150 0 1810 0 006fO
0 1000-0 9ttl 0 SOZO 0 ttfO 0 tSlO 0 ZtOO'O Z990*0 6ttl*0 ZSSO 0 I9t0 0 OOtf T
0 tOOO'O IZIZ'O StZO 0 6150 0 Z9tO 0 tSOO'O 6t90'0 SZBf 0 OtSO'O BttO'O OOtl'O
0 lOOO'O 190Z'0 6EZO 0 tOSO'O 0910 0 1600*0 5t90*0 90flt*0 9950*0 SttO 0 006f 0
0 tOOO'O 910Z'0 tEZO'O IOSO'0 0910 0 OBOO'O It90'0 1081 0 9950'0 IttO'O 006fO
2 5222-2 t661'0 "" 0 t6tO 0 65tO*0 6t00'0 tt90'0 SBtl'O Z950'0 ZtlO'O OOlt'O
0 1000 0 9B61 0 IEZO'0 E6t0'0 65t0'0 6100 0 It90'0 EBU'O Z9SO'0 ZtIO 0 OOZS'O
0 lOOO'O 99BfO StZO'O OttO 0 tSIO 0 5100 0 S990'0 6ttfO 9SSO'0 I9t0'0 006fO
0 tOOO'O BZBfO ttZO'O £910 0 9510 0 ttOO'O t990'0 6£tfO SSSO'O 9910'0 OOZS'O
2 5222-2 I6"'" OIZO'O 95t0'0 5510-0 ttOO'O t990'0 OftfO fSSO'O 5910*0 OOtfO
0 1000 0 1911 0 tOZO'O OSIO'O SStO'O ttOO'O t990 0 tZtfO tSSO'O S9f0'0 OOSfO
0 tOOO'O 9t9Z'0 IOEO 0 SZ90 0 ZtIO 0 6600'0 l£tO 0 ttOZ'O 0190'0 BISO'O OOOZ'O
0 lOOO'O tOEfO t9ZO 0 ESSO 0 5910 0 8800 0 f690'0 CeBl'O IBSO'O 68t0'0 OOSZ'O
0 1000-0 6ZtfO 91ZO 0 OZSO 0 t910'0 tSOO'O 6t90'0 SZBfO ItSO'O BttO'O OOtZ'O
0 tOOO 0 9lOf 0 tEZO'O 66tO 0 09t0'0 OSOO'O Z190'0 Z6tf 0 1950 0 fttO'O 009Z'0
0 tOOO'O tOOZ'O ZtZO'O 9610'0 6510*0 6100*0 1190*0 IBll'O f9SO*0 ZtlO'O OOtfO
0 1000 0 t96t 0 BZZO 0 6810*0 6510*0 BtOO'O Ot90'0 9ttf 0 1950'0 IttO'O OOSfO
0 lOOO'O ttlZ'O ISZO 0 6ZSO-0 E9t0'0 IBOO'O tB90'0 tlBI'O ItSO'O letO'O OOII'O
0 1000 0 IOOZ 0 ZEZO 0 96t0'0 6SIO'0 6100'0 tt90'0 tetf 0 E950'0 ZttO'O OOtl'O
0 tOOO'O 9Z6f 0 tZZO 0 1810 0 8510 0 ttOO'O 8990'0 S9tf 0 6550'0 69t0'0 OOSZ'O
0 1000 0 8161 0 EZZO 0 08tO 0 BSIO'O ttOO'O t990'0 t9tf 0 6SSO'0 69tO 0 OOZf 0
0 tOOO'O t88f 0 6tt0'0 fttO'O tStO 0 9t00'0 9990 0 tStl'O tSSO'O 89t0'0 OOSZ'O
0 tOOO 0 E181 0 StZO 0 9910 0 9510 0 StOO'O 5990 0 tttfO 5550*0 I9t0*0 OOlfO
0 1000*0 8911 0 10ZO 0 ZStO 0 SStO'O ZtOO'O t990'0 IZtf 0 tSSO'O S910'0 OOOf 0
0 tOOO'O EStl'O 90ZO'0 6110 0 SStO'O ZtOO'O Z990'0 IZtf 0 ZSSO'O 1910 0 OOfZ'O
! I222'2 tUT'° tOZO'O tltO 0 tSIO'O 1100*0 Z990*0 tltl'O ZSSO'O I9t0*0 OOtZ'O
0 1000 0 9081 0 IIZO 0 6510*0 9510 0 ttOO'O 1990'0 tftl'O tSSO'O 99t0'0 OOSl'O
0 lOOO'O EISI'O ZtZO'O 0910'0 9S10'0 ItOO 0 I990'0 Sttf 0 tSSO'O 99t0'0 OOtfO
0 lOOO'O Ittt 0 EOtO'O ttlO 0 ISIO'O 1100*0 Z990'0 9ltl'0 ZSSO'O I9t0'0 OOtl'O
0 1000 0 8191*0 5610*0 6ZIO*0 ESIO'O 6900*0 1990*0 669f 0 OSSO'O E910'0 0061'0
0 lOOO'O US-I'O t6IO*0 tttO'O tStO'O 6900'0 1990'0 869fO OSSO'O f9tO*0 OOZfO
0 1000*0 9651*0 6810*0 6ttO 0 tStO 0 t900*0 t990*0 0691'0 OSSO'O E9t0'0 OOtl'O
0 lOOO'O 9ttfO tOZO'O ZttO'O tSIO'O HOO'O Z990'0 EttfO ISSO'O I9t0'0 OOII'O
' 522" ° "9r° "lO'O IZtO 0 ZStO'O 6900'0 1990'0 B69fO OSSO'O t9t0'0 OOtfO
0 lOOO'O EtSt 0 tetO 0 tttO'O ISIO'O t900'0 t990'0 9B91'0 OSSO 0 1910 0 OOtfO
1 tOOO'O 9ElfO 6tZO*0 6150*0 6910 0 Z600*0 5010*0 tt6f 0 9650'0 86tO*0 OOtO'O
0 tOOO'O 9691*0 66t0'0 9ttO 0 tStO'O OtOO'O t990'0 90tl'0 tSSO'O I9t0'0 OOZO'O
5222-2 'fl''0 '910'0 tetO 0 etlO'O Z900'0 0190'0 t99l'0 ISSO'O B910'0 OOSl'O
tOOO'O OStO'O tSOO'O 609fO ZOtO'O 6010'0 IBIO'O tttO'O BStO'O IBtO'O OOtfO
lOOO'O 1910-0 0900-0 tfBfO tttO'O OZTO'O OBIO'O TttO'O tSIO'O OBtO'O OOtO'O
tOOO'O ZEZO 0 ZBOO'O S09Z*0 6501*0 t9lO*0 OBIO'O tltO'O tSlO'O OfltO'O OOSl'O
5222-2 JZf0'0 "OO'O "51 '0 ftOfO 1910*0 6110*0 TtfO'O tStO'O 6ttO*0 OOSfO
1000-0 BOZO-0 ttOO'O tttZ'O 9S60'0 OStO'O 8110-0 6tt0'0 ZStO-0 8110-0 OOOZ-0
tOOO'O 59ZO 0 teOO'O 900fO ZOZl'O 06t0'0 SBtO'O IStO'O BStO'O SBtO'O OOOZ'O
tOOO;0 ZBZO 0 B600 0 6ltf 0 91tf 0 1010*0 6810*0 I9t0'0 0910'0 6«t0'0 006Z'0
tOOO'O ttZO'O S600'0 6STE'0 9IZT"0 16TO'0 tfllO'O tStO'O 6S10'0 1810'C 0091'0
1000 0 SttO 0 6tOO*0 ttJZ 0 1101*0 1910*0 6110*0 tttO'O tStO'O SllO'O OOSfO
looo'o tozo'o itoo*o tett o zteo'o ttto'o eno'o etto'o tsto'o Btto o ooei'O
looo'o ttzo o seoo'o eetf o oszfo teto'o tflto'o tsto-o esto-o tBto-o oosf o
I222'2 "no ° S90° ° ttlz'e oteo*o ttto*o Btto-o etto'o tsto'o etio-o oosf o
1000 0 tttO'O £900'0 IIIZ'O OIBO'O tttO'O 6110'0 OlfO'O tSIO'O 6tIO'0 OOlfO
lOOO'O 6910 0 0900'0 IIOZ'O flOBO'O IZIO'O IBIO'O tttO'O SStO'O 1810'0 OOtfO
looo'o 99to'o teoo-0 tztt o eizfo leio-o SBIO'O isto'o ssto'o ssto'O oosfo
tOOO'O IIZO'O 9100'0 ZBSt 0 ftOf 0 9510*0 6tlO*0 OltO'O tStO'O 6ll0'0 OOlfO
tOOO'O 60ZO 0 EtOO'O 9BtZ'0 8160*0 OStO'O BttO'O 6tt0'0 ZStO'O SttO'O OOlfO
1000*0 IOZO'0 ttOO'O 66EZ*0 1160*0 5110*0 BtlO'O 6tt0'0 tSIO'O SttO'O OOZZ'O
1000*0 6tZO'0 19t0'0 tIBO 0 I6t0'0 OS90'0 0161'0 1910'0 060f 0 ZttO'O 0060'0
5222'2 221°'° "':o'° t'"'0 "">'•> >w« 609Z-0 9990-0 sosi-o BIOI-O oooz-o
TOOO o eeto o 99zo-o 6tzi o etto-o etof o TISZ'O 5590-0 zetfo eiof o ooss*o
1000*0 ttfO 0 BZZO'O ttOf 0 6990-0 BBBO'O Ottt'O IISO'O Ittf 0 6880'0 OOOB'O
looo-o toto'o otto-o tstfo teto'0 ssofo ttsfo 9990*0 sosi*o tiofo ootfo
5222'2 ?S52-2 5152'° "" * su" " "ei" ° ««'» 5590*0 teti*o iteo*o oos9-o
5222 2 I ° '"O'O t9EO'0 TttO-0 eSOt'O 6010-0 9860'0 1080 0 OOtfO
TOOO 0 ttZO'O tfltO-0 IBBO'O ttSO'O IIIO'O 60tfO SBtO'O ttll'O tt60'0 OOtO'O
5222'° 5'"'0 "I0'<> T8»0'» 0150*0 ZttO'O lltt'O SBtO-0 tttfO ItBO'O OOtl'O
5222-2 fS52 2 1552-2 J"0'0 ""'" ""'o "ot*o uto-o eooi-o tt8o-o ooti-o
1000 0 1510*0 lOEO'O B6tt Sil£(>.0 1810*0 ItOt'O tllO'O 1660'0 IISO'O OOtO'O
tooo'o imfrt c«rT-n BCIA 2 ""0^0 tBll 0 ttBt 0 0510 0 1191*0 ttll'O OOIO'O
tJo?:o s"t*S "":S lll°-l sU!!-0 "":* •'":• '"*:• ••«'• '"*••' •»"'•
5222-2 5SS5-2 Iti5'l) ""''•' ""''' "«•« sz^'o sttt*o 'T«-O sztt-S oo'i-S
5222 2 5?! "" ° lsso'0 tttO*0 UZO-O tB9I-0 ZI9fO Z98T'0 tB9I'0 OOSO'O
1000 0 ZIBO'O ZttO'O tttS'O fStl'O 19lf 0 8901*0 9151*0 1911*0 8901*0 0060*0
1000 0 ttZO 0 1600 0 OOOZ*0 1060-0 9IWO S910'0 ZtZO'O tSIO'O S910'0 OOtO'O
1000 0 E090 0 IStO 0 Otlf 0 1990'0 9800'0 16ZO'0 tOSO'O tttO'O tJtO'O OOZO'O

-------
0.7200 0 0556 0 0666 0.2240 0 0763 0.0111 0-0180 0.0703 0 0347 0.3082 0.0001 000 PART 0000 4.00 17.80 -999.90 0.46 0.900E+10
0.6900 0.0582 0.06*5 0.2364 0.061? 0.011? 0.0185 0.0746 0.0372 0.3314 0 0001 000 PART 0000 4.39 1? 80 -999 90 0.49 0 900E+10
1.3200 O.OS39 0 064? 0.21S3 0.0759 0.0106 0.0177 0.0670 0.0329 0.2909 0 0001 000 PART 0000 3.85 17.80 -999.90 0.43 0.900C*10
0.9000 0.0583 0.0701 0.2372 0.0819 0.0118 0.0185 0.0751 0.0374 0 3329 0.0001 000 PART 0000 4.41 17.80 -999.90 0.50 0.900E*10
0.0050 0.0478 0.0571 0.1828 0.0679 0.0083 0.0162 0.0520 0.0246 0.2129 0.0001 0 0 0 PART 0000 2 81 17.80 -999.90 0.32 0.900E+10
A Month
B day
C year
D pollutant
beginning hour
ending hour
wind spaed
temperature
solar radiation
friction velocity
Honin- Gbukhov length
i nixing height
roughness length
M number of diameters
0 s«allest diameter
P largest dlajietar
} density
1 reference height
S leaf area index
F LAI correction exponent
J land use type
V sasiple set nu«bar
0.000 28. 00 6 0.50 .00 .00 .80 .0 1. 24
0.000 28.00 6 O.SO .00 .00 .80 .0 1. 24
0.000 28.00 6 O.SO .00 .00 .00 .0 1. 24
0.000 28.00 6 0.50 .00 .00 .00 .01. 24
0.000 28.00 £ 0.50 .00 .00 .00 .01. 24
0 000 28 . 00 6 0. 50 .00 . 00 .00 .0 1. 24

-------
241 	 Total
zas       18 11
Ooran t Hocst (1993),
DESERT GRASSES,  1-2 •
   3.0.   140.0
-999.9.     2.0
-994.9.  2
    1   4.
Mutter of Data Sets
ASQV, 19. 939-951.
HIGH SAGEBRUSH
       - tO  (em), id(cm)
       - us measurement ht.  (m).  temp.  meaa.
       - LAI (estimated). vegetation  state
       - no. of diamecara. density(om/cm"3>
       - partical diameter(microns I
                                                     hi.  (m)
MM-OO-YY a HR
(1st)
05-16-63 22:4*
05-18-63 22)41
05-18-63 22:46
05-26-63 23)26
05-26-63 23)24
05-26-63 23)24
06-05-63 22)10
06-05-63 22)10
06-05-63 22)10
06-12-63 22)43
06-12-43 22:43
06-12-63 22:43
06-24-43 23)06
06-24-43 23.06
06-24-83 23:06
06-27-43 21:31
06-27-93 21:31
06-27-93 21:31
ENDDATA
S04 1 2
E HR
(1st)
23)18
23)18
23)18
23)54
23)54
23)54
22)40
22)40
22)40
23)13
23)13
23:13
23)26
23)28
23:21
22:01
22:01
22:01


Nicholson and Oavies
BARLEY
-999.7, 12.0
1.0. 1.0
-999.9. -999.9
10 1.0
MM-OD-YY 3 Hit
(1st)
06-04-79 6:00
ENOOATA
304 6 10


W
(m/s)
7.61
.53
.43
.23
.59
.83
.74
.40
.32
3.00
3.39
3.75
3.07
3.24
3.46
3.17
3.80
4.37


(1967),


TEMP SH RAO USTAR MOM IN HEAT FLUX RA
(C) (N/m"2) lm/i) (ml (H/m>«2) (s/cm)
14.7 -999. 0.4 166. -999.
14.7 -999. 0.4 166. -994.
14.7 -494. 0.4 166. -944.
14.5 -999. 0.26 44. -994.
19.5 -499. 0.26 44. -999.
14.5 -994. 0.26 44. -999.
17.5 -944. 0.27 77. -944.
17.5 -999. 0.27 77. -949.
17.5 -999. 0.27 77. -944.
14.3 -994. 0.20 34. -444.
14.9 -999. 0.20 34. -999.
14.9 -999. 0.20 34. -999.
14.1 -999. 0.26 59. -999.
14.1 -994:9 0.26 59. -999.
14.1 -999.9 0.2< 59. -999.
20.5 -999.9 0.30 71. -999.
20.5 -999.9 0.30 71. -999.
20.3 -999.9 0.30 71. -999.


AEnv. 21. 1561-1571

- 10 (cm), id (en)
-999.
-999.
-994.
-494.
-999.
-999.
-994.
-999.
-999.
-999.
-999.
-999.
-999.
-999.3
-994.3
-999.9
-999.9
-999.9





RD RC
Is/em
(s/cm)
-999.9 -944.9
-999.
-999.
-999.
-999.
-999.
-999.
-999.
-999.
-994.
-999.
-999.
-999.
-999.
-999.
-444.3
-994.3
-944.9
-999.9
-999.9
-999.9
-994.3
-994.9
-999.9
-999.3
-999.3
-994.9
-999.9
-999.3
-999.9 -999.3
-999.9 -999.3
-999.9 -999.9










VD
(em/si
4.21
4.05
3.65
1.33
1.80
1.74
3.14
3.02
2.84
1.75
1.62
1.31
1.5«
1.47
1.14
1.17
1.15
1.10









































































































- v* measurement ht. (ml, tamp. oaas. ht. (m)


- LAI (estimated), vegetation state








- no. of diameters, density (om/cm"3)
E HR
(1st)



Nicr.olson and Davies
ROUGH PASTURE
-999.7, 11.0
1.0, 1.0
-999.9, -999.3
10 1.0
MM-OD-YY 3 HR
(1st)
06-21-79 2:36
06-21-79
06-21-79 3:53
07-19-79
07-19-79
07-19-79 4:07
ENDDATA
SO4 12 3


MS TEMP SH RAD USTAR MONIN HEAT FLUX RA
(B/S)
1.71


(1367).


(C) (H/m"2) (m/s) (ml (H/m"2)
(s/cm)
14.2 -999.9 0.11 -999e9 -999.3 1.49


AEnv. 21. 1561-1571

- tO (eal. zd(en)





RD RC
(s/cm)
Is/cm)
-999.3 -999.3










VD
(cm/si
0.04





ZO
(cal
0.8





Rl

0.051





phia

1.36





phih

1.36





- vs meaaurement ht* (m) , temp. m*as. ht. (m)


E HR
(1st)

4:00

3.43
2:34



Nicholson and Oavies
SHORT CRASS
-999.7, 9.0
1.0. 1.0
-999.9, -999.9
10 1.0
0.1 0.2 0.3
MM-OD-YY 3 HR
(1st)
10-08-79 «:59
10-10-79
10-17-79
2-04-80
2-06-90
2-19-80
2-19-90 11:23
2-20-90
2-20-60 11:22
2-21-60 9)35
2-22-80
2-26-60 9)56
ENDOATA





0.4
E HR
(1st)

5:48
5)20
4:32
4:50
4:40

4:40


4:20




MS
(n/s)
1.23
2.53
1.64
2.69
2.84
2.35


(1987),





0.5 0.
MS
(m/ 1 1
3.09
2.21
2.90
3.90
1.44
4.40
2.15
3.73
2.98
3.13
3.81
1.79

- LAI (esciaated) , vegetation state








- no. of diameters, density (ca/cm"3)
TEMP SH RAD USTAR MONIN HEAT FLUX RA
(C) (»/m"2) (m/s) (m) "(H/m"2)
(s/cm)
6.8 -999.9 0.10 -999*9 -999.9 1.19
21.2 -999. 3 0.26 -999*4 -999.9 0.38
6.3 -999.3 0.06 -499*9 -999.9 4.09
14.4 -999.9 0.29 -999*9 -999.3
-999.9 -999.9 0.27 -999*9 -999.9
9.9 -999.9 0.27 -999*9 -999.3


AEnv, 21, 1561-1571

- <0 (em), td(cm)
- we measurement ht. (m) . temp. meas. ht
- LAI (estimated) . vegetation state
- no. at diameters, density (dm/en*-3)
0.33
0.39
0.32





. (m)


RD RC
(a/cm)
(s/cm)
-999.3 -999.9
-999.9 -994.3
-999.9 -999.3
-999.9 -999.9
-999.3 -999.9
-999.3








-999.9








VD
(cm/s)
0.33
0.57
0.03
0.33
0.23
-0.01








ZO
(cal
0.4
1.0
0.3
2.2
1.3
3.2








Ri

-0.054
-0.042
0.093
0.000
-999.9
0.005








phia

0.86
0.88
1.93
1.00
-999.9
1.02








phih

0.73
0.77
1.33
1.00
-999.9
1.02








6 0.7 0.8 0.3 1.0- particle diamatazs
TEMP SH RAD USTAR MONIN HEAT FLUX
(C) (W/ra"2) (m/sl (m) (W/i»"2)
-999.9 -999.9 0.22 -999*9 -999.9
-999.3 -999.9 0.19 -999e9 -999.9
RA
(S'rcn)
0.61
0.63
- 9.4 -994.9 0.29 -999*9 -999.9 0.35
-949.9 -999.9 0.28 -999*9 -999.3
-999.9 -999.9 0.10 -999e9 -999.3
.5 -999.9 0.33 -999«9 -999.9
0.51
1.41
0.41
.4 -999.3 0.17 -999*9 -999.9 0.73
.9 -999.9 0.28 -999*9 -999.9 0.48
-99 .9 -999.3 0.22 -999*9 -999.3 0.61
.0 -999.9 0.20 -999*9 -999.9 0.80
.0 -999.9 0.25 -999*9 -999.3 0.62
.0 -999.9 0.16 -999*9 -999.9 0.72


RO
O/cn)
-999.9
RC
(i/cm)
-999.3
-999.9 -999.9
-994.9 -494.9
-944.9 -999.9
-999.9 -999.9
-999.9 -999.9
-999.9 -999.9
-999.9 -999.9
-999.9 -999.3
-999.9 -999.3
-999.9 -999.9
-999.9 -994.3


VD
(ca/s)
0.29
0.02
-0.19
0.28
0.13
-0.07
0.02
0.20
-0.10
0.09
0.04
-0.02

ZO
(cm)
0.4
0.8
1.4
0.3
0.3
0.4
0.9
0.3
0.4
0.3
0.2
1.0

Rl

-999.9
-999.9
-0.003
-999.3
-999.9
0.001
0.016
-0.010
-999.3
0.023
0.000
0.002

phia

-999.9
-999.9
• 0.37
-999.3
-999.9
1. 00
1.09
0.9«
-999.9
1.13
1.00
1.01

phih

-999.9
-999.3
. 0.93
-999.9
-999.9
1.00
1-.09
0.93
-999.9
1.13
1.00
1.01


-------
SO4
41 10
t

Hicaolaon and Oaviaa (1987),

A20V, 21.





1S61-IS71
ROUGH PA3TURS
-»99.7,
1.0,
-*»».*.
10 1.
wi-OD-rr

10-12-7*
10-12-7*
1O*1 at^'74
ill" n— ' *
10-17-7*
10-11-7*
10-18-7*
10-19-7*
10-19-7*
11-21-79
11-22-79
11-22-79
11-23-79
11-27-79
11-27-74
11-21-74
11-28-74
11-29-79
11-29-79
11-30-79
12-03-79
12-05-79
12-06-79
12-06-79
12-07-79
12-12-79
12-13-79
1-22-80
1-23-80
1-24-80
1-21-80
1-21-10
1-29-80
1-29-80
1-30-80
1-30-80
1-31-80
2-06-80
2-07-80
2-07-80
2-01-80
2-12-80
2-14-80
2-25-80
3-05-80
3-06-80
3-06-80
3-11-80
3-12-80
CNOOATA
9.0
1.0
-994.9
0
a HR
(lac)

,6,S»
6,54

7:00

6:58
10150

10:50


10:53

11:02

11:01

11:09


11:12

11:20
10:50

10:59
11:03

11:00

10:42

10:49

11:16

llilC


11:22

9:59

9:56
10:03




- 10 (Oil
. xd (oil
- va a*a.aur«unc
- LAI (*»tiaut*d) ,

S HR
Uat)
4:17
3,1*
* if
4:41

4:31


2:55

2:41
3:34

4:02

4:07

3:55

4:31
5:52

2:35


3llt


S;04

3l5l

4:07

6:25

4:00

3:41
5:03

5:16

4:36


7:03


MS
(•/*)
2.55
2.92
3.53
3.27
1.53
3.14
1.50
1.41
2.30
2.22
3.36
2.01
2.52
3.34
2.<2
2.41
2.41
4.30
3.45
4.05
1.47
1.71
2.11
4.72
3.29
6.03
2.07
2.19
2.7S
1.06
2.22
3.47
5.02
2.40
4.11
2.43
3.45
1. 11
4.31
1.33
3.46
1.96
3.12
3.23
2.70
2.52
4.34

- no. of
TSHF
1C) (M
-»*».
-»»».
7.
10.
5.
13.
8.
2.
4.
4.
6.
5.
4.5
6.5
3.1
5.2
3.1
7.0
2.9
-994.9
6.8
7.6
8.0
6.7
6.3
7.7
1.7
1.4
1.2
2.1
7.7
7.3
8.2
6.7
-499.9
-999.9
-999.9
-999.9
-944.9
-949.9
5.9
4.6
3.2
S.3
4.7
5.2
7.4


he. (•), t*ap- i
v*q*tation scat

i*aa. ht.
..

(Ml




diaawc*ra. d*naity (oja/ea**3l
3» RAO
/•"*)
-999.
-999.
-949.
-999.
-999.
-999.
-999.
-999.
-444.
-449.
-494.
-499.
-999.9
-999.9
-999.9
-999.9
-999.9
-999.9
-999.9
-999.9
-999.9
-999.9
-999.9
-999.3
-999.9
-999.9
-999.3
-999.9
-999.9
-994.9
-99*. 9
-999.9
-944.9
-999.9
-999.9
-999.9
-944.9
-994.9
-999.9
-999.9
-999.9
-999.9
-999.9
-999.9
-999.9
-999.9
-999.9

USTAR NOJUH HEAT flBX «A
(•/a) (•)
0.1* -499*9
0.24 -499*9
0.34 -999*9
0.33 -999*9
0.14 -944*4
0.3S -499*9
0.11 -999*9
0.14 -444*4
0.19 -499*9
0.17 -999*4
0.27 -994*9
0.15 -944*4
9.17 -499*9
9.25 -949*9
0.17 -999*9
9.23 -994.9
0.15 -999*9
0.39 -449*9
9.23 -949*9
0.32 -944*4
0.12 -499*9
9.13 -499*9
0.20 -944*9
0.31 -499*9
9.23 -944*9
0.41 -499*9
9.11 -999*9
9.10 -999*9
9.16 -999.9
9.07 -9*9*9
9.14 -999*9
0.22 -999*9
0.31 -444*9
9.16 -994*4
0.26 -444*9
0.11 -999*9
9.19 -999*9
0.15 -999*9
0.21 -999*9
9.14 -444*9
0.22 -44»*9
0.15 -999*9
.0.20 -999.9
9.21 -444*4
9.17 -994*9
0.19 -999*9
0.35 -999*9

M/B**2)
-»»».»
-»»».*
-9*4.9
-449.9
-4*9.9
-»»».*
'999.9
-999.9
-999.9
-999.9
-49*.*
-999.9
-999.4
-999.9
-999.9
-499.9
-999.9
-999.9
-999.9
-949. 3
-999.9
-999.9
-999.9
-999.9
-999.9
-999.9
-999.9
-999.9
-999.9
-999.9
-999.9
-999.9
-999.9
-99*. 9
-949.9
-999.9
-999.9
-999.9
-444.9
-944.9
-444.9
-994.9
-994.9
-444.9
-444.9
-444.9
-99*. 9

(a/ca»
0.67
0.52
0 40
0.31
0.2*
0.13
0.31
0.41
0.71
0.6*
0.74
0.47
0.»7
0.17
0.54
0.93
0.61
1.05
0.21
0.63
0.39
5.97
1.04
0.70
0.49
0.62
9.33
1.72
2.00
1.12
2.21
1.10
0.73
0.53
0.9S
9.63
0.76
0.91
9.86
0.54
1.91
0.72
0.92
9.76
9.73
9.94
9.67
9.41

DO RC VO
(j/ca» (a /can (em/at
-9*9.
-9*9.
~999
-»**.
-999.
-444.
-449.
-9*9.
-999.
-999.
-999.
-«»».
-449.
-9*9.
-999.
-4*9.
-94*.
-99*.
-9*4.
-4*9.
-44*.
-4**.
-499.
-9*9.
-944.
-4*9.
-441,
-999.
-999.
-999.
-999.
-999.
-949.
-994.
-449.
-944.
-449.
-99*.
-999.
-999.
-999.
-999.
-999.
-999.
-444.
-44*.
-44*.
-444.
-4**.
-9*4.
~999
-4**.
-»»*.
-44*.
-4*4.
-4»*.
'999.
'999.
-999.
-999.
-999.
-»»*.
-4»».
'999.
-999.
-4*4.
-44*.
-999.
-94*.
-4*».
-4*9.
-999.
-944.
-499.
-499.
-999.
-999.
-999.
-449.
-999.
-449.
-949.
-999.
-449.
-999.
0.20
0.0*
0 i6
0.04
-0.34
0.24
0.34
9.91
0.10
0.07
-0.96
-0.01
-0.12
0.15
-0.21
-0.03
0.05
0.02
-0.21
9.32
0.94
-0.11
9.07
0.23
0.18
-0.06
0.41
0.04
9.01
0.21
-0.03
-0.12
0.22
0.22
0.04
-9.36
9.92
-944.9 -0.18
-949.9 0.04
-99*.* 0.01
-944.9 0.10
-449.9 0.00
-»»».9 -0.09
-444.9 0.21
-444.9 9.51
-444.9 0.04
-444.9 9.02
-444.9 -0.53

3O4        68
Hicaolaon and Oavl*a
BARE SOIL
-994.7,    10.0
   1.0,     1.0
-4»».3,  -9»S.9
   10  1.0
  0.1  0.2
MM-00-Y1f

 1-2S-IO
 2-15-80
 2-27-39
 2-29-80
 3-10-60
 3-13-80
E.NDOATA
(1917),  AEnv,  21,  1561-1571
0.3
a HR
(lat)




9:59

9.4
e HR
(lat)
4:5C
5:39
4:52
4:04

4:31
9.5 9.6
MS
(a/al
2. S3
3.19
2.06
2.30
3.04
4.01
0.7
TEH?
1C)
2.S
7.4
5.1
5.9
3.6
4.3
9.1 9.9
SW RAD
(H/n**2)
-999.9
-944.9
-994.9
-999.9
-449.9
-999.9
1.
MI
(mi
t:
a.
0.
9
0
0
          ZO (Oil.  znfcml
          wa m*aaur*a*nc he. («), e*ap. m*a. ht.
          LAI (*atimat*d), «*q*tation ita,t*
          no. of 4iaaMt*ra, 4*naity (qm/cm**3)
                          1.0 - partial* dian*c*ri
                           TAR  MOHIN
                           /a)     In)
                           .11 -4*4*9
                          0.19 -999*9
                          0.16 -994*9
                          9.17 -444*4
                          0.24 -999*9
                          0.34 -999*9
Id
0
0
2
I
1
2
1
4
2
0
9
0
0
9
0
9
9
9
1
0
9
1
1
9
0
0
0
9
0
0
0
0
9
0
0
9
0
9
0
0
9
0
0
0
0
0
0
0
10
mi
Ri
-»»9.9
->*». 9
. -999.3
.
.
.3
.1
.3
.0
.6
.7
.4
.3
.6
.5
.6
.4
.6
.3
.4
.6
.4
.1
.4
.2
.5
.3
.3
.1
.1
.6
.4
.2
.1
.4
.1
.4
.1
.5
.2
.3
.2
.4
.2
.2
.2
.7
.3
0
-0
0
-0
0
0
-0
0
0
0
0
0
9
9
0
9
9
.
.

.
.
,
.
.
.
.



.


.
004
014
939
003
010
Oil
006
910
003
000
034
006
042
014
944
007
020
-999.9
9
9
0
9
9
a
9
9
0
9
0
9
a
0
.
.


.
,
.

.
.
.
.

,
023
034
008
008
015
004
049
044
Oil
035
023
004
000
018
-999.9
-999.9
-999.9
-449.9
-994.9
-444.3
0
-0
0
0
0
9
9
.
.
.
.
.


006
009
005
000
012
919
901
phia
-999.9
-999.9
-999
1.
9.
1.
0.
1.
I.
0.
1.
1.
1.
1.
1.
1.
1.
1.
1.
1.
-999
1.
1.
1.
1.
1.
1.
1.
1.
1.
1.
1.
1.
1.
1.
-999
-999
-999
-999
-999
-994
1.
0.
1.
1.
1.
1.
1.
.3
02
95
26
99
06
96
96
06
02
00
21
93
28
08
30
04
12
.9
14
22
04
04
09
92
34
29
96
23
14
02
00
10
.9
.9
.3
.3
.9
.3
03
97
03
00
07
05
01
pain
-949.9
-994.9
-999
1.
0.
1.
0.
1.
1.
0.
1.
1.
I.
1.
1.
1.
1.
1.
I.
1.
-999
1.
1.
1.
1.
1.
1.
1.
1.
1.
1.
I.
1.
1.
1.
-999
-999
-999
-999
-994
-999
1.
9.
I.
1.
1.
I.
1.
.9
02
90
26
91
06
06
96
06
02
00
21
03
21
01
30
04
12
.9
14
22
04
04
99
02
34
29
06
23
14
02
90
10
.9
.9
.3
.9
.3
.9
03
94
03
00
07
05
01
T runt

RA
RO RC
VO
/m**2) li/-=al (a/eat (a/cal (ca/a)
-999.9
-944.9
-944.9
-499.9
-999.9
-999.9
0.
9.
0.
0.
9.
9.
74
88
80
78
55
35
-999.
-944.
-444,
-444.
-49*.
-499.
-999.
-444.
-44*.
-4»».
-*»9.
-999.
9
9
0
9
0
-0
.01
.25
.01
.31
.25
.16
ZO
(on)
0.2
9.1
0.3
0.2
0.5
0.7


-0
9
-O
-0
0
-0
Ri

.016
.008
.024
.035
.004
.010
pnira

0.34
1.04
0.92
0.89
1.92
0.96
ptlih

0.89
1.34
3.85
9.80
1.02
9.93

-------
SO4 1 10
Mieholaon and Oaviaa
LONG GRASS
-9M. 7. «.0
1.0. 1.0
-***.». -999.9
10 1
MM-OO-YY

10-21-79
EVOOATA
3O4
.0
B HR
Hat)
«t52

4 2
(19*7).
AEav. 21. 15C1-1571
- zO (oi). zd(CB)
- wa muauraaanc ht. (a), tamp. neaa. ht
- LAX (aatimatad) , v*q*tation itata
. (a)






- no. of d-Umatara. danaicy (OB/OB"])
( HR
(lat)



Micholaoa and Daviaa
M .TEMP SN RAO USTAR MONIX HEAT FlOX
"(a/a)
1.77


(19*7).
(C) (W/B'-J) IB/I) IB) (K/B**2)
f.S -999.9 0.10 -999*9 -999.9


ACnv, 21. 15*1-1571
(a/c
I.



RA
Hi
«2



RO
(1/CBI
-999.9



RC
(a/cm)
-9»9.9




lea
0



VD
i/a)
.35



ZO Ri pniB phia
(CB)
1.3 0.071 1.59 1.59



TALI, BARLEY 77?
-999.7,
1.0,
-999.9.
10 1
0*1 0
MM-OO-YY

<-09-IO
6-10-«0
S-10-«0
6-12-40
ENOOATA
S
Hlcka *t
6.0
1.0
-999.9
.0
a HH
(lat)
5:11

5:10
5:12

8 3




£ HR
(lat)

5:05








MS
(B/a)
2.45
3.5«
1.67
2.21


al (19I«), BLM. 34,
- zO (oi), id (cm)
- wa a*ia*ur*B*nt ht. (a), t*ap. a*aa. ht
- LAI (aatimatad) . v*q*tation Jtata
- no. of dlaa*t*ri, danaity (oa/em**3)
TEMP SH RAO USTAR HONIN HEAT FLUX
(C) (W/B"2) (B/a) (B) (II/B"2)
11.2 -999.9 0.17 -999*9 -999.9
14.7 -999.9 0.27 -999*9 -999.9
10.9 -999.9 0.12 -999*9 -999.9
10.0 -999.9 0.17 -999*9 -999.9


103-121

. (a)



(a/e
0.
0.
1.
0.







RA
HI
>4
51
15
7«







RO
(i/oml
-999.9
-999.9
-999.9
-999.9







RC
(a/ca)
-999.9
-999.9
-999.9
-999.9












VD
(cm/.)
0
0
-0
0



.01
.04
.29
.21







ZO Ri phia phih
(CB)
0.3 0.006 1.03 1.03
0.3 -0.010 0.97 0,93
0.3 -0.010 0.97 0.93
0.4 0.000 1.00 1.00



GRASSLAND
9.4,
7.0.
-999.9.
10 1
01 fl
. -L U
MM-OD-YY

09-17-79
09-17-79
09-25-79
09-25-79
09-25-79
09-25-79
09-25-79
09-25-79
ENODATA
?b
Garland
-999.9
1.0
-999.9
.0
a HR
Hat)
15:05
15:35
11:35
12:05
12:35
13:05
14:35
15:35

1 3
(1912).




S HR
(lat)
15:30
16.00
12:00
12:30
13:00
13:30
15:00
UiOO






MS
(B/at
1.J3
1.40
1.21
1.44
1.34
1.29
1.57
1.53


- ZO (CBl, zdlcn)
- wa m*aauzaa*nt ht. (ml, taap. m*aa. ht
- LAIIaatiaatadl. v*q*tation itat*
- no. of diaancera, danaity (oa/cm**3)
TEMP S* RAO USTAR HONIN HEAT FLUX
(C) (»/«•• 2) (a/a) (a) (Vim--?,:
"23.S -999.9 0.11 -999*9 51.0
23.5 -999.9 0.11 -999*9 19.0
21.5 -999.9 0.15 -999*9 81.0
22.0 -999.9 0.15 -999*9 70.0
22.5 -999.9 0.15 -999*9 70.0
22.9 -999.9 0.16 -999*9 41.0
22.9 -999.9 0.15 -999*9 53.0
22.9 -999.9 0.15 -999*9 22.0



. (a)



(.'•1C
-999
-999
-999
-999
-999
-999
-999
-9«9






RA
al
.9
.9
.9
.3
.9
.9
.9
.9






RD
(a/cm)
-999.9
-999.9
-999.9
-999.9
-99*. 9
-999.9
-999.9
-999.9






RC
(a/cal
-999.9
-999.9
-999.9
-999.9
-999.9
-999.9
-999.9
-999.9











VD





(ca/a)
0
-0
0
0
0
0
0
0


.61
.01
.72
.44
.33
.12
.00
.00












Confarvnc* Proc**dinqa. 84.9-85*
GRASS - MIND TUNNEL
2.0,
-999.9.
-999. a.
-999.9
-999.9
-999.9






1 1.0
13.
MM-OO-YY


ENDOATA
Pb
Garland

3 HR
(lat)


1 3
(19<2).

E HR
(1st)




MS
(a/ II
-999. 3


- :0 aattaatadleal, zd(ca)
- wa mcaauraaent ht. (a), taap. raaaa. ht
- LAI Katiaatad) , v*q*tation icata
- no. at dlaa*t*ri, danaity (aa/ca"3)
- diaactar (aicrona)
TEMP SW RAO USTAR MOM IN HEAT FLUX
(C) (M/B"2) (m/J) (Bl (M/B"2)
20. -999.9 0.35 9.0*9 -999.9



. (a)









RA
'(s/cal
-999


.9







RO
(a/cat
-999.9







RC
(a/cal
-999.9













VO






(ca/a)
1.90






Confaranc* Proc*«dinqa. 849-aSi
GRASS - MIND TUNNEL
2.0.
-999.9.
-999.8,
-999.9
-999.9
-999.9






1 1.0
10.
MM-DD-tY



a HR
(1st)


E HR
(lit)


ws
(a/31
-999.9
- zO aatiaatad(ca), zd(ca)
- wa auauraaanc ht. la), tamp. maaa. ht,
- LAI (aitimacadl , v«q*tacion itace
- no. at dlaa*c*ri. density (qm/cm"3l
- diaa*t*r (aicronal
TEMP S« RAD USTAR MONIN HEAT FLUX
(C) (W/a"2) (m/ii (al iw/«"2)
20. -999.9 0.35 9.0*9 -999.9

. (al








RA
(a/a
-999
al
.9





no
(a/cai
-999.9





RC
la/cnl
-999.9











VD






(cm/JI
1.40
ENOOATA

-------
?»          13
Oarland  (1912).  Conference proewdinqa.  144-85*
GRASS) - MIMD TUMNU,
   2. a.  -999.9               - JO estimated (cml, id (cm)
-999.9.  -999.9               - ws BMSureB*nt ht. (Bl, teeta. ma*, tit.  (B)
-999.1.  -»99.9               - IAI (estimated) .  v*o;*tation state
    1  1.9                    - no. of diameters. density (qa/cm"3)
  7.S                         - ataaweur (micron*)
MH-OO-YY    B MB   E HR     W   TEMf   SH RAO  USTAR  HONIH  HEAT FLUX     RA      RO     RC     VD
        *  (1st)  (lie)   (m/»)     (CT e»/B"2)   (m/sl     (•)    (»/»"2)  U/CBI            I   3
Garland  (1912),  Conference Proceedings.  149-451
CRASS - WIND TUNNEL
   2.0,  -999.9               - 10 estimated (cm), zd(cn)
-999.9.  -999.9               - u* measurement. ht. (B) , temp. neas. M.  (»)
-999.1.  -999.9               - IAI (**tl]ut*4) ,  v*q*catlon itat*
1 1.0
s.o
MM-OO-YY


ENDOATA
no. off diameters, density (dB/ea**3)
- diam*t*r (microns)
a HR
(1st)


E HR
(1st)


US
(a/a)
-999.9

TEMP S« RAO
(Cl (W/B"«1
20. -999.9

USTAR
(B/sl
0.3S

MONIN
(Bl
9.0*9

HEAT runt
(«/«"2)
-999.9

RA
(I/CM)
-999.9

RO
(s/ern)
-999.9

RC
(S/CB)
-999.3

VD
(CB/S)
0.45

?b          13
Airland  (1912),  Conf*r«nc* Proc««dinqi,  149-«5I
SKASS -  KIND  TUNNEL
   2.0.  -999.9               - tfl estimated (cmi, zd(eB|
-999.9,  -999.9               - «• B*a*uxuMat lit. (Bt, C*ap. aus. ht.  (Bl
-999. (,  -999.9               - UKutiaatM), v*q*tatloa (tat*
    1  1.0.                   - oe. at dlaB*t*r«, d*n*ity (q»/ca"3)
  3.2                         - dlaB*t« (Kleran*)
MK-OD-n    a  HR   B HR     OS   TEMP   SH RAO  VSTAR  HONIH  HEAT FLDX     RA      RO      RC     VD
           (1st)  (1st)   (m/i)    1C) (W/B*«2)   (•/*)     la)     (»/«••*>  (J/on(  (I/CBI  (a/a)  ICB/I)
                       -999.9    20.   -999.9   O.J5  9.0*9     -999.9  -999.9  -999.3  -999.9   0.15
ENDDATA

P6          13-
Garland  (19*2),  Coal*r*nc* Proc«*4inq».  «49-tS»
GRASS -  WIND  TUNNEL
   2.0,  -999.9               - zO Mtiaatcddax. zd(ca)
-999.9,  -999.9               - tr* awa
-------
Pb
Garland
GRAM -
2.0,
-999.9,
-999.1.
1 1
0.04
f 3
(19*2), Confer
NINO TUNHCL
-99».»
-999.9
-999.9
.0

MH-OD-YY S OH


ENOOATA
FEOOH
Garland
GRASS
2.0,
10.0,
-999.9,
1 1
2.1
MH-OD-YY

05-21-11
09-23-81
ENOOATA
FEOOH
Garland
GRASS -
2.0.
10.0,
-999.8,
1 1
2.*
MM-OO-YY

06-91-11
9*-17-ll
SNODATA
FEOOH
Garland
GRASS
2.9.
10.9.
-999.9,
1 1
3.8
MM-OO-YY

09-30-12
ENOOATA
Fin* Prt
Weaely •
LEAFLESS
100.,
39.0.
-999.9.
6 1
0.95 0.
MM-OD-YY

01-24-11
01-26-11
01-2«-ll
01-27-J1
01-27-81
01-27-11
01-21-11
01-28-31
01-21-91
01-2S-81
91-28-11
01-21-11
SNDOATA
(lit)


2 3
(1312).

-999.9
-999.9
-999.9
.0

S HR
(1st)



2 3
(19*2).

E HR
(lat)



Confer



enesi Proceedings, 149-15*
- zO estimated Ccml, id (cm)
- v* measurement ht. (ml, temp. mas. ht
- LAI (estimated) , vegetation atate
- oo. of diameters, density (gm/em"3>

m
(m/sl
-999.9


- diameter (microns)
TEM? SW RAO USTAR NONIM
1C) («/••**) (m/s) (ml
20. -999.9 0.35 9.0*9



HEAT

rune
<»/m"2)
-999.9




(ml

RA
(a/eml
-99».9







RO
(a/em»
-999.9






RC
(a/em)
-999.9




VD
(em/a)
9.09


ence Proceedings, 149-151




- ifl estimated (cm), zd(cm)
- •stlmated us neas. ht. (ml.


temp.


meas










ht. (m)
- LAI (estimated) , v*q*tation atat*


E HR
(1st)






WS
(m/al
2.5
3.0


Conference Pi
- no. of diametera. density
- diameter (microns)
TEKP S* RAO VSTAR MONIN
(C) (W/m"2) (m/sl (ml
20.9 -999.9 -999.9 9.0e9
20.0 -999.9 -999.9 9.0.9


oceedings. 849-15*


HEAT
(N/n







FLUX
l"2)
9.0
0.0





RA
(s/cml
-999.9
-999.9






(a/e


RO
•1
-999.9
-999.9








RC
(s/cml
-999.9
-999.9





VD
(em/a)
0.05
0.12



HIND TUNNEL
-999.9
-999.9
-999.9
.0

B HR
(lat)



1 3
(19*2),

-999.9
-999.9
-999.9
.0

3 HR
(lat)


12 4




- zO estimated (cm), zd(cml
- estimated us meas. ht. (ml.

temp.

meaa.





ht. (ml
- LAI (estimated) , vegetation state


E HR
(1st)






ws
(m/sl
3.5
3.5


-no. of diametara, density
- diameter (microns)
TEMP SW RAO USTAR MONIN
(C) (H/m"2) (m/i) (ml
20.0 -999.9 -999.9 9.0.9
20.9 -999.9 -999.9 9.0.9




HEAT
(W/n






FLUX
l««2)
0.0
0.0




RA
(s/cml
-999.9
-999.9





(1/c


RD
•1
-999.9
-999.9






RC
(a/eml
-999.9
-999.9




VO
(em/a)
0.18
0.12


Conference Proceedings. 149-158







- zO estimated (en), zd(em)
- estimated us neas. ht. (m) ,


temp.












meas. ht. (ml
- LAI (estimated) , vegetation atate


E HR
(lat)



t »1 (19*3), a


MS
(m/sl
2.4


U4. 27,
DECIDUOUS FOREST IN
-999.9
42.0
3.0
.0
0< 0.07
9 HR
(lat)
10:09
17:30
11:00
10:30
11:30
12:30
10:30
11:30
12:00
12:30
13:90
15:00





- no. of diametera, denaity
- diameter (microna)
TEMP SW RAO VSTAR MONIN
(C) (W/m"2) (m/a) (ml
20.9 -999.9 -999.9 9.0e9


237-255.
WINTER (North Carolina)
- zO estimated (en), zd(cm)
- us measurement ht. (m) , temp


HEAT
(W/a






. meaa


FLUX
"2)
0.0





. ht.


RA
(s/eml
-999.9





(m)



(a/e
-999








RO
ml
.9








RC
(a/en)
-99».9








VD
(cm/a)
9.42






- LAI (estimated I, vegetation state
- no. of diametera. density (gm/em**
9.01 0
E HR
(lat)
19:30
11:90
11:30
11:00
12:90
13:00
11:99
12:90
12:30
13:09
13:30
L5:30

.09 0.
MS
(m/al
4.0
2.1
2.5
2.5
1.6
1.8
3.6
2.7
2.9
3.3
2.1
i.:

1 - particle diametera
TEMP SW RAO USTAR MONIN
(C) l»/m*'2) {m/al (m)
9.2 -999.9 9.S4 -999.9
15.4 -999.9 0.26 -999*9
14.1 -999.9 0.21 -999*9
13.2 -999.9 0.31 -999*9
15.0 -999.9 0.31 -999*9
15.2 -999.9 0.25 -999*9
7.S -999.9 O.S6 -999*9
1.4 -999.9 0.57 -999*9
1.5 -999.9 0.41 -999*9
9.2 -999.9 0.5< -999*9
9.7 -999.9 0.57 -999*9
9.7 -999.9 0.35 -999*9


HEAT
(K/m

3)

FLUX
•*2)
4 9-. 9
-16.0
-20.0




15.0
S6.0
26.0
43.0
209.0

C.3.0
231.0
237.9


55.0



RA
(a/eml
-999.9
-999.9
-999.9
-999.9
-999.9
-999.9
-999.9
-999.9
-•99.9
-999.9
-999.9
-999.9




(s/e
-999
-999
-999
-999
-999
-999
-999
-999
-999
-999
-999
-999



RO
ml
.9
.9
.9
.9
.9
.9
.9
.9
.9
.9
.9
.9



RC
(a/em)
-999.
-999.
-999.
-999.
-999.
-999.
-999.
-999.9
-999.9
-999.9
-999.9
-999.9



VO
•(em/a)
0.919
0.920
9.120
0.110
0.070
9.140
0.610
0.320
0.800
9.550
0.200
0.090


-------
          II  3
       •c «1  (1S«2), Cent*
SHORT GRASS. TEXAS
   1.0.  -999.3
  19.3.  '  42.0
-944.9.  -9**.*
   13  1.4
mne* PrecMdlnq*. 443-9S2

   - <0 M«*JMt*d(cal. id (c»)
   - *• BMiuruwit ht.  IB), tup. »•••.  ht.  (•!
   - LAI(MtlBat*dl, vcqvtation stat*
   - no. at 4i*B*t*ri. dmuity  (q»/eB««3)

MH-OD-YY

01-23-71
01-25-71
01-23-71
01-25-71
01-25-71
01-25-71
01-23-71
01-25-71
01-25-71
01-25-71
01-25-78
01-25-71
08-25-78
08-23-78
08-25-18
88-25-78
08-25-78
38-25-78
BNDDATA
PART
Loc«nr «

a HR
(lit)
lllll
11:48
12:18
12141
13:18
13i48
Mill
14141
15111
15t4l
16:11
16i4>
17:11
17:41
UlL8
18:48
19:18
19:48

61 5

C HR
U*t)
11:48
12:11
12l 41
13 ill
13l4l
14:11
14:41
15:18
15l48
16:11
16:41
17:11
17:48
Ulll
18148
19111
19:48
20:18


MurphyU989l,


MS TENT SH RAO
(B/«
-99*.
-999.
-999.
-9*1.
-991.
-999.
-999.
-499.
-999.
-999.
-999.
-999.
-999.
-911.
-919.
-999.
1C) (N/B"2)
30. -999.9
30. -999.9
30. -999.9
30. -994.9
30. -99*. 9
30. -999.9
30. -999.9
30. -999.9
30. -999.9
30. -999.9
30. -999.9
30. -999.9
30. -999.9
30. -999.9
30. -999.9
30. -999.9
-999.9 30. -999.9
-9*9.9 30. -999.9


BtM.


4«, 353-366.
1.0 - partial
USTAR MOMIM
(»/») (a)
0.24 -944*9
0.25 -999*9
0.2C -999*9
0.32 -94»*»
0.20 -999*9
0.21 -99**9
0.22 -999*9
0.33 -999*9
0.24 -999*9
0.27 -999*9
0.33 -999*9
0.34 -999*9
0.32 -999*9
0.25 -999*9
0.27 -999*9
0.21 -999*9
0.20 -999*9
0.18 -999*9



• dlamt*r«
HEAT rune
(K/B--Z1
144.
133.
15«.
no.
173.
112.
45.
105.
102.
««.
52.
IS.
-14.
-29.
-41.
-36.
-35.
-42.



RA
(1/CBl
-999.9
-999.9
-99*.*
-911.9
-991.9
-999.9
-999.9
-991.9
-991.9
-919.9
-999.9
-999.9
-999.9
-999.9
-9*9.9
-999.9
-999.9
-999.9



RO Re VD
I«/CB
-»»*.
-*9».
-99*.
-9*1.
-91*.
-9*1.
-9*9.
-99*.
-9M.
-419.
-4*1.
-41*.
-411.
-411.
-9*1.
-9*9.
-9*4.
-4*».



(•/smi
-999.
-999.
-191.
-911.
-911.
-111.
-199.
-99*.
-»»*.
-999.
-911.
-911.
-111.
-991.
-111.
-911.
-991.
-999.



(cm/tl
• 0.220
0.140
0.110
.150
.230
.240
.150
.150
.190
0.150
a. no
0.294
0.200
0.200
0.250
3.150
0.070
0.140



PINE PLANTATION
28.0,
9.1.
9.0
6
MM-00-«














































790.0
-999.9
-999.9
1.
a HR
(lit)

















































- zO (oil, zdtca
- w« nuiuruunt
)
ht. (n|, tup.

m*«». Ht.

(B)






- IAI. v*q*tation Jtae*
- no. of dlan*t*n, cUniity (
-------
          T T i T T T
          i i i i »
0000,0000,3000000
T i i * i i i i i i i i t i i i
H ssssssss sHssi
I 1 I I I I I I I I I I t I I  I

-------
                  Appendix D



Predicted Deposition Velocities vs Particle Diameter

-------
uu
en
 _
a

LU
CTJ
a
Q_
Lu
a
       1.0E+02
       1.0E+01   -
1.0E+00  -
£     1.0E-01   -
      1.0E-02   -
1.0E-03  -
      1.0E-04
      1.0E-05
              0.1
                          1.0               10.0

                     PflRTICLE  DIflMETER (MICRONS)
100.0
 Figure D-1.   Predicted deposition velocity for the CARB-based models for u. = 10 cm/s,

              za = 10 cm, LAI = 1.0, p = 1.0 g/cm3, and neutral stability. The gravitational

              settling velocity is vg.
                                        D-1

-------
      100
          DEPOSITION VELOCITY (CM/SEC)
i.OOOE-05
         0.1
         1                    10
PARTICLE  DIAMETER (MICRONS)
                                                                      100
                 ADOM  1
          ADOM 2
ADOM 3
V(sed>
Figure D-2.  Predicted deposition velocity for the ADOM-based models for u. = 10 cm/s,
           20 = 10 cm, LAI = 1.0, p = 1.0 g/cm3, and neutral stability. The gravitational
           settling velocity is vf
                                   D-2

-------
o
LU
CO
u


>-
I—
I—I

o

LU
CO
o
a.
LU
a
      1.0E+02
      1.0E+01   -
      1.0E*00  -
       1.0E-01   -
       1.0E-02   -
       1.0E-03   -
       1.0E-04
       1.0E-05
               0.1
                                 1.0              10.0


                            PflRTICLE DIflMETER  (MICRONS)
100.0
  Figure D-3.  Predicted deposition velocity for the UAM-based models for u. = 10 cm/s,

              Zo « 10 cm, LAI = 1.0, p = 1.0 g/cm3, and neutral stability. The gravitational

              settling velocity is vf
                                         D-3

-------
          APPENDIXE

Implementation of the Modified Source
     Depletion Method in ISC2

-------
E.1    Overview of Method

       Horst (1983) describes a method for incorporating the effects of deposition on the
vertical distribution of material in a plume, without resorting to a full surface depletion
treatment (see the discussion of plume depletion techniques in Section 3). The method includes
a vertical profile adjustment factor as well as a source depletion factor, so that concentrations in
the lower portion of the plume approximate those produced by the surface depletion model
This adjustment factor is important because deposition is proportional to near-surface
concentrations.  The source depletion method by itself overestimates near-surface
concentrations, thereby overpredicting deposition rates which hastens removal of material from
the plume.
       Let V(xAh) denote the vertical distribution of plume material in the absence of
deposition.  In terms of the notation used by Horst (1983),
u
                                                                                     (E-l)
where D is the crosswind-integrated concentration distribution for a plume at a height h above
the ground, released from a source of unit source strength.  Then the vertical distribution factor
that is modified to account for the depletion resulting from deposition, Vd (x^h), is defined to
be:

                                                                                     (E-2)

FQ(x) is the fraction of material that remains in the plume at the downwind distance x (i.e., the
mass that has not yet been deposited on the surface).  This factor may be thought of as a source
depletion factor, a ratio of the "current" mass emission rate to the original mass emission rate.
      is the vertical profile adjustment factor.
       The effect of Equation (E-2) is illustrated in Figures E-l and E-2. Figure (E-l) displays
a depletion factor FQ, and the corresponding profile correction factor P(x^z) for a distance at
which oz is U times the plume height. This assures that the plume has been in contact with the
ground for a long enough time that significant deposition has occurred. The depletion factor is
constant with height, whereas the profile correction shows that most of the material is lost from
the lower portion of the plume.  Figure (E-2) compares the vertical profile of concentration both
with and without deposition and the corresponding depletion of material from the plume.  The
depleted plume profile is computed using Equation (E-2).
                                           E-l

-------
       2.0 n
     N 1.5 -
     D
       •1.0 -
                           Depletion  Factor
Profile Correction
FIGURE E-1. ILLUSTRATION OF THE DEPLETION FACTOR FQ AND THE

           CORRESPONDING PROFILE CORRECTION FACTOR
                                 E-2

-------
       2.0 -i
     N 1.5
     O
     E
       •1.0 -
     N
       0.0
                                                Original Profile
                       Depleted  Profile
   I I I i i I  I I i | I i i  I I i i i i |*Ti i I i i i  i i | i f i i  i i i i i |
0.0         0.5         1.0          1.5         2.0
                        Concentration
FIGURE E-2. VERTICAL PROFILE OF CONCENTRATION BEFORE AND AFTER
          APPLYING FQ AND P(x^) SHOWN IN FIGURE E-l.
                              E-3

-------
       FQ(x) is a function of the total deposition velocity (vj, V(x^h), and
                          EXP
                        (E-3)
where zd is a height near the surface at which the deposition flux is calculated.  This equation
reflects the fact that the material removed from the plume by deposition is just the integral of
the deposition flux over the distance that the plume has traveled.  For general forms
and P(x^:), Equation (E-3) is evaluated numerically.
       The deposition velocity for particles generally contains a component related to the
settling velocity, vr A "tilted plume" is used to simulate the effect of gravitational settling on the
plume as a whole. This approximation entails replacing the plume height, h, in Equations (E-l)
and (E-2) with

                                A, =» h - 1 vt                                        (E-4)
                                         u

For large travel-times,, h, can become less than zero.  However, the tilted plume approximation
is not a valid approach in this  region. Therefore, a minimum value of zero must be imposed on
h,.  In effect, this limits the settling of the plume, although the deposition velocity continues to
account for gravitational settling near the surface.
       The profile correction factor P(x^) is developed by Horst (1983) for the case in which
reflection of material from the mixing lid is not important.  He finds
1-1
                                                                                     (E-5)
where R(z^d) is an atmospheric resistance to vertical transport.  When the product vsR(z^a) is
•of order 0.1 or less, the exponential function is approximated (for small argument) to simplify
                                                                                      (E-6)
                                            E-4

-------
 This simplification is important, since the integral in Equation (H-6) can be computed using
 analytical approximations for many forms of R(z^J that are consistent with the Briggs' formulas
 for oz (Gifford, 1976). Typically, only the largest particles may have a settling velocity vs large
 enough to require the numerical integration of Equation (E-5).

       •The atmospheric resistance is defined as
where K(z) is the vertical eddy diffusivity. Because we will be using empirical expressions for av
K(z) should be consistent with these.  Horst (1983) points out that

                                K - u oz ^                                        (E-8)
                                          ax

and that
                                z = V37* az                                        (E-9)

for a Gaussian plume from a ground-level source, where z  is the mean height of the

distribution of mass in the plume.  Using Equation (E-9) to map z to av Equation (E-7) is
represented by
                                          o
                                           * dx

This allows R(z^J to be evaluated for particular forms of oz. Horst provides solutions for

                             
-------
is derived in Section E-2.  (Solutions to Equations (E-7) and (E-6) for each form of oz are listed
in Section E-2).

E2    Extension of Solutions for Urban Classes A and B

       The Briggs' curves for oz for urban locations during stability classes A and B have the
form of Equation (E-12).  Therefore, Equation (E-10) becomes
               *«*L at? yrr&p
             jlAin(^|^
                                                                                (E-13)
The limits are implicit functions obtained from Equation (E-9):
                                                                                (E-14)
That is, x(z) is the distance at which oz equals z/V2/ii .  If both sides are squared, x(z) can be

expressed as the root of a cubic equation. In developing the FORTRAN code to implement
Horst's method, we use an iterative method to solve for the root of Equation (E-14).
       Adding Equation (E-13) to the solutions given by Horst for the other forms of ov we
have:
  a. = OK
Jl
\ Tt
                   -Li.
                   au
                                                                                (E-15)
                                                                                (E-16)

-------
  o, - afll * ftt);
                                                                                     (E-17)
              fee)
                *
               Jl
I in
         au
                y/1
- 1
                          + 1
                +  1
                -  1
(E-18)
       The profile correction factor, P(^z), requires the integral of the product of R^J and
the vertical distribution factor V(x,z,h), which is dominated by a series of exponential functions
of height. An analytic solution to this integral is possible for the R(z^,j) terms involving ln(z/zd),
(z-z,j) and (z2^2); but the complex form of Equation (E-18), coupled with the supplementary
relation in Equation (E-14), precludes such a solution. After trying several approximation
techniques, the solution for P(x^zd) with R(z^d) given by Equation (E-18), was approximated as
follows.
       First, a program was developed that solves the integral in Equation (E-6) numerically.
This not only allows us to test various approximate results, it can also serve as a numerical
solver if no analytical approximations are found to be adequate.  Then, we developed an
approximate  expression for R(z^:<1) for small z, which is facilitated by the fact that the constant
b = 0.001 for urban classes A and B:
                           or
                                                                                    (E-19)
where k
2b  [T
 a N|  2  '
                                            E-7

-------
This allows Equation (E-18) to be written as
                           cat
          (1 + to)1**  +  1   (1
                                                          - 1
                                                                                   (E-20)
Further expanding (1 + kz)^4 as 1 + kz/4, the natural log expression in Equation (E-18)
becomes
                              -l-In
                                             *  J
                                                           (E-21)
This gives a leading tenn that is the same as Equation (E-15), for oz = ax, and the approximate
result for P is
                                           v?  JL/
                                            1C  OU
                                                   -1
                                                          (E-22a)
where
                                                 .  2
                                                          (E-22b)
       Comparing Equation (E-22) with the numerical solution, the approximate form worked
well for small z, but diverged from the correct solution at larger z. Several empirical
adjustments produced a good fit for a full range of heights.  The resulting analytic expressions
for P(x^d) for each of the oz functions are:
 - v.
ua
                                  In
                                                                                   (E-23)
  o? = axf( 1 + bx)lf2:
                       ua
                                         t\   1    o i
                                        -1 -I + - I at -
                                   Jt
                                   2
                                                                                   (E-24)
                                           E-8

-------
      axl(I * &e):
                                                                               (E-25)
                 vf<°'-"1
             •» 1 +
                    ua
              ^
                                                                               (E-26)
For the last form,  k = —   — , and
                     a  N 2
at(l - .0006 aj2
0.6724 a.
                                      300m
                                      300/n
and
                                      1000m

                                      1000m
The approximation to the integral in P(x^J for az = ax(l + bx)1/2 matches the numerical
solution to within 1% for zd = 0.03 m and oz ^ 4000 m, and it matches to within about 2% for
zd »  1 m over the same height range.

E3    Mixing-Lid Treatment

       The results presented above do not include the presence of a mixing lid.  With such a lid,
the profile correction factor must operate only within the mixed layer, so that the upper limit in
the integral for P(x^d) is z,, not infinity, in Equation (E-6). Furthermore, the standard
formulation of V(xAh) is also a function of z^ since the distribution of material in the plume is
                                         E-9

-------
"reflected" from z = z^  In principle, the additional reflections in the Vfr^Azj just add to the
number of terms in P(x^), since the form of the integral in Equation (E-6) remains the same.
However, most of the emphasis in obtaining P(xrzd) is placed near the ground, because this is
where the depletion correction is most important Therefore, in the interest of streamlining  the
implementation of the method, we have adopted an alternate strategy. We solve for P(x^d) for
the case of well-mixed plumes (oz > z,), and compare the results with Equations (E-23) through
(E-26).

       In the well-mixed limit,
so that P(x^d) involves the integral of terms involving just ln(z/z,j), (z-zj, and (z2^2), since the
exponentials are not present in Equation (E-27).  Performing the integrations yields functions
that are equivalent to Equations (E-23) through (E-25), except oz is replaced by a constant times
z,:
                                                                                   (E-28)
                                                                                   (E-29)

Therefore, the effect of the lid is to limit the size of oz in evaluating P(x£
-------
also performed well when compared to the numerical integration, so this more compact result
was adopted

E.4    Numerical Integration For P(x^d)

       Because Equation (E-3), involves the numerical evaluation of an integral over the
distance from the source to each receptor, an analytic representation of P(x^d) is preferred in
order to streamline the computations. As discussed in Section E.I, P(x^d) can be represented
by simple analytic functions so long as v, R(z^,j) £ 0.1.  However, larger and denser particles
(greater than 10 urn in diameter) have settling velocities great enough to violate this condition
at times.  For these situations, the full expression for P(xrzd) (Equation E-5) must be solved
numerically as well  This means that each point evaluated in Equation (E-3) involves a
numerical evaluation of Equation (E-5). This can be time-consuming.

       The subroutines developed make use of a general integration routine.  It subdivides the
interval into more and more equally-spaced segments until the value of the integral converges to
within an  imposed tolerance. Further study is recommended to optimize these integration
procedures. For  example, the tolerance level might be too restrictive.  Or, an integration
technique might be  specifically designed for the integrand. We know the form of V(xAh) and
R(z^,j), and may  be able  to increase the efficiency of the integration by designing an algorithm
that "knows" where the integrand changes most rapidly, and least rapidly. Fewer points are
needed to integrate across regions in which the variation of the integrand is nearly linear.

       There is also a possibility that the solution to Equation (E-5) can be approximated
simply enough to  avoid its numerical solution.  We have been  able to recast the integral in
Equation  (E-5) to one of the form:

                                  '"'A e-^D^ * "D*? dz                             (E-31)
                          *<

for R(z^d) of the  form

                 *(z^)  = A ]n(zlzd) * B(z - zd) * c(z2  - 2<2)                        (E-32)

With a suitable definition of a new variable of integration, it appears that Equation (E-5) could
be solved  to yield a  representation made up of the product of  a Gamma function, an exponential
function, and a parabolic  cylinder function.  Such a solution has not been completely worked
out,  since  it is not clear that such a representation would lead  to a more efficient evaluation of
the integrals.

                                           E-ll

-------
                                        References

Giffbrd, RA, Jrn 1976:  Turbulent diffusion - Typing schemes:  A review.  NucL Saf., 17, 68-86.

Horst, T.W.,  1983: A correction to the Gaussian source-depletion model  In Precipitation
       Scavenging, Dry Deposition and Resuspension, H.R. Pruppacher, R.G. Semonin, W.G.N.
       Slinn, eds^ Elsevier, NY.
                                          E-12

-------
                                     TECHNICAL REPORT DATA
                   (Please read Instructions on reverse  before completing)
 I. RETORT NO.
    EPA-454/R-94-015
                                                               3. RECIPfENrS ACCESSION NO.
 4. TITUS AND SUBTITLE
    Development and Testing of Dry  Deposition
 Algorithms
                 5. REPORT DATE
                   April  1994
                 6. PERFORMING ORGANIZATION CODE
 7. AUTHOR®
    Joseph  S.  Scire and Gary E. Moore,  Sigma Research
    Corporation	
                                                               8. PERFORMING ORGANIZATION REPORT NO.
 9. PERFORMING ORGANIZATION NAME AND ADDRESS

    Sigma Research Corporation
    196 Baker Avenue
    Concord,  MA  01742
                                                               10. PROGRAM ELEMENT NO.
                 11. CONTRACT/GRANT NO.
                                                                  68-D9007,  Work Assignment
                                                                 3-1
 12. SPONSORING AGENCY NAME AND ADDRESS
    U.S. Environmental  Protection Agency
    Office  of Air Quality Planning  and Standards,  TSD
    Research Triangle Park,  NC  27711
                                                               13. TYPE OF REPORT AND PERIOD COVERED
                    Final Report
                 14. SPONSORING AGENCY CODE
 13. SUPPLEMENTARY NOTES
 This document replaces  EPA-454/R-92-017.
EPA Work Assignment Manager:   Jawad S. Touma
 16. ABSTRACT

      This  study was designed to identify dry deposition models suitable for routine
 use, evaluate and intercompare several techniques,  and select the most  appropriate
 approach for use in the  Industrial  Source Complex (ISC2) model.  Reviews were conducted
 of methods  for computing dry deposition velocity,  plume depletion, and  certain
 micrometeorological parameters from routinely-available observations.   Several
 observational data bases were identified from the literature and used in testing  and
 evaluating  ten particle  deposition  velocity models.   Recommendations for computing
 particle deposition velocity, plume depletion,  and micrometeorological  variables  were
 made.  These techniques  have been incorporated  into a revised version of the ISC2 model
 and related processor programs.
 17.
                                     KEY WORDS AND DOCUMENT ANALYSIS
                   DESCRIPTORS
                                              b. IDENTIFIERS/OPEN ENDED TERMS
                                                                                a. COSATI Field/Group
    Air Pollution
    Dry Deposition
    Air Quality Dispersion Modeling
    Meteorology	
 18. DISTRIBUTION STATEMENT

    Release Unlimited
19. SECURITY. CLASS (Rtport)
    Unclassified
21. NO. OF PAGES

    128
                                              20. SECURITY CLASS (Page)
                                                  Unclassified
                                                                                22. PRICE
EPA form 1220-1 (Rev. 4-77)  PREVIOUS EDITION IS OBSOLETE

-------