United States
Environmental Protection
Agency
Office of Air Quality
Planning and Standards
Research Triangle Park NC 27711
E^A-450/4-87-003
February 1987
Air
Technical Discussion
Related To The Choice
Of Photolytic Rates
For Carbon Bond
Mechanisms In
OZIPM4/EKMA

-------
    Technical Discussion
 Related to the Choice of
    Photolytic  Rates for
Carbon Bond  Mechanisms
    in OZIPM4/EKMA
             H.E. Jeffries
          Kenneth G. Sexton

        Department of Environmental
          Sciences and Engineering
        University of North Carolina
          Chapel Hill, NC 27514
             Prepared For

     Monitoring and Data Analysis Division.
   Office of Air Quality Planning and Standards
     U.S. Environmental Protection Agency
         Office of Air and Radiation
   Office of Air Quality Planning and Standards
        Research Triangle Park, NC
          „...„-„.»'»:??"""""

-------
This report has been reviewed by the Office of Air Quality Planning and Standards, U.S. Environ-
mental Protection Agency and approved for publication as received from the Principle Investigator.
Approval does not signify that the contents necessarily reflect the views and policies of the Agency,
neither does mention of trade names or commercial products constitute endorsement or recommen-
dation for use.
                                              ii

-------
                                Contents
1   Introduction                                                           l

    Purpose	    1

    The Photolysis Process	  .    1
       Photon Flux or Radiant Energy Flux	    2
       Radiation Models	    9
       Kinetics Parameters	   10

    Example Calculation	   12

    OZIPM Methods	   12

    Spectral Distribution of Rates	   15


2   Review of Calculated Rates                                          18

    Effects of Changes hi Cross Section and Quantum Yield	   18

    Values Calculated For Carbon Bond m hi OZIPM2	   25
       Nitrogen Dioxide Photolysis in CB3/OZIPM2	   29
       Carbonyl Photolysis in CB3/OZIPM2	   29
       Osone Photolysis in CB3/OZIPM2	   31
       How CB3/OZIPM2 Rates Influenced CB4/OZIPM3/4 Rates	   32

    Affect of Altitude on Photolysis Rates	   32


3   Atmospheric Rates                                                    34

    Measurements of Nitrogen Dioxide Photolysis Rates  .	   34

    Comparison of Theory and Measurement for Irradianee	   34

    UV Albedo	   36
       Adjustment of Calculated Actinic Flux for Albedo	   41

                                      iii

-------
4   Recommended Rates                                                   45

    Rates for Carbon Bond IV in OZIPMS/4   .................   45
       Surface Rates	   '45
       Mixed Layer Rates	   45
       Recommended CB4 Photolytic Rates	   46

    Rates for Carbon Bond m in OZIPM3 ...................   46

    Example Predictions Using New Rates	   47
       Simulation Conditions	   48
       Simulation Results	   49
                                        IV

-------
                                     Figures
 1. Extraterrestrial and Surface Solar Spectrum	     4
 2. Schematic Illustration of Gas and Aerosol Light Scattering Processes	     5
 3. Example of Spectral Irradiance Components	     7
 4.  Sun's Position in Central NC For Five Months	     8
 5. Actinic or Spherical Irradiance		    11
 6. Calculation of Nitrogen Dioxide Photolysis Rate for Zenith Angle = 0°	    14
 7. Relative Spectral Photolysis Rates	    16
 8. Photolysis Rates Calculated From Peterson 1976 Actinic Flux	    27
 9. Quantum Yields For Nitrogen Dioxide Photolysis	,  ,	    28
10. Examples of Measured Atmospheric Rates of Nitrogen Dioxide Photolysis	    35
11. Comparison of Radiation Model Predictions With Spectroradiometer Measure-
    ments.	    37
12. Comparison of Horisontal and Spherical Irradiance	    38
IS. Comparisons of Theory with Measurements for Atmospheric  NO2 Actinometry	    44
14. Example Simulation Using CB4: Comparison of Surface and Mixed Layer Pho-
    tolytic Rates	    52
15. Example Simulation Using  CBS: Comparison of Mixed Layer and  1984 Pho-
    tolytic Rates	    53
16. Comparison of CBS and CB4 Using Mixed Layer Photolytic Rates	    54
17. Comparison of CARB (CBS) and HCHO (CB4) Using  Mixed  Layer Photolytic
    Rates	    55
18. Example Simulation Using  CB4: 40% HC  Control Using Mixed Layer Pho-
    tolytic Rates	    57
19. Example Simulation Using CBS: 40% HC Control Using 1984 Photolytic Rates  ....    58
20. Comparison of Oione-Hydrocarbon Relationships for CBS and CB4	    59

-------
                                      Tables
 1. Calculation of Nitrogen Dioxide Photolysis Rate for Zenith Angle = 0°	    13
 2. Definitions of Headings used in Photolytic Rate Tables  	    19
 8. Theoretical Rates Used in EPA Models	    20
 4. Cross Sections and Quantum Yields Used	:	    26
 5. Nitrogen Dioxide Photolysis Rate as a Function of Altitude	    33
 6. Albedo Values Assumed by Peterson.	    40
 7. Albedo Values Measured by Dickerson, et al. .	    40
 8. UV Albedo Values Measured by Doda and Green	    41
 9. Comparison of Higher Albedo  Calculated Photolysis Rates with Atmospheric
    Measurements	    43
10. Recommended CB4 Surface Photolytic Rates	    47
11. Altitude Adjustments Corrected For Albedo	    47
12. Recommended CB4 Mixed Layer Photolytic Rates   	    48
IS. Recommended CBS Surface Photolytic Rates	    49
14. Recommended CBS Mixed Layer Photolytic Rates   	    49
15. Initial Conditions and Emissions Fractions	    50
                                           VI

-------
                                                                    1
 Introduction
Purpose


The purpose of this report is to explain the methods and identify the sources of
information used to calculate photolytic rates for the System Applications, Inc.
(SAI) chemical models Carbon Bond III (CB3) and Carbon Bond IV (CB4) for
use in SAI's OZIPM3 and OZIPM4 oxidant modeling programs.1 Both OZIPM3
and OZIPM4 programs are very similar; the former allows larger mechanisms to
be used and has the Carbon Bond X mechanism built-in, while OZIPM4 is smaller
and has CB4 as the default mechanism. The mixed layer photolytic rates described
in Chapter 4 are also the default photolysis rates for CB4 in the OZIPM4 program.
OZIPM4 is expected to be used in new State Implementation Plan calculations.

   This report will also review previous recommendations for photolytic rates that
were used in the original program, OZIPP*, and in the revised program, OZIPM2.3
In the last nine years these rates have undergone significant change. In addition,
new measurements and analysis of irradiances in the UNC outdoor chamber suggest
some major changes may be needed in the photolytic rates of the SAI developed
models when these models are used in atmospheric simulations.

The Photolysis Process

Three factors determine the photolysis rate of a chemical species, S,  in the atmo-
sphere.

   The first factor is the number of photons of a given wavelength (energy) moving
through an  area of space as a function of time, that is, the photons-cm~2-sec~1-
nm"1. This is often called the actinic flux and will be designated here by the script
letter A\.  Actinic flux is a function of wavelength, A, the solar zenith angle, x> the
altitude above the earth's surface, y, and the ground reflectivity or albedo, A.

-------
Introduction	Photon Flux or Radiant Energy Flux

    The second factor is the ability of the species molecule to absorb photons, that
is, the species absorption cross section (cm2-molecule~1) in the wavelength interval,
designated here as o^(S).

    The third factor is the fate of the  molecule  after it has absorbed a photon.
That is,  the number of  species molecules  that follow a particular reaction path
after absorbing a photon, divided by the number of-species molecules that absorbed
photons in the wavelength interval. This fraction is called the quantum yield (a
dimensionless number between 0 and 1)  for a given process, and is designated here
as $A(S).

    The photolysis rate of species S at a given zenith angle x> at an altitude y, and
at an albedo A, will be designated as  J(x,y, A;S), and is calculated by:

                                                                          (1)

that is, the photolysis rate is the  area under the product curve of photon flux,
absorption cross section,  and quantum yield as a function of wavelength. This area
has units of time"1.  The radiant  energy,  or photon flux,  A\, is common  to  all
species; the kinetics parameters, a\ and $*, are unique for each species.

    The lowest wavelength at which a species will photolyze (Ai) is often determined
by the lower limit of photon flux reaching the earth's surface (290-300 nm), while
the highest wavelength (A]) is often determined by the cross section or quantum
yield approaching zero at that wavelength.

Photon  Flux  or Radiant Energy Flux
R&diometric Quantities and Units
Radiant  energy, Q\,  is normally measured in joules.   The relationship between
wavelength, A,  and energy  is given by Planck's law. The energy of a single photon
is given by
                                   o  -hc
                                   fiA~T
where h is Planck's constant. Because 1 watt = 1 joule/sec,
                        h = 6.626176 x I0~u  joule-sec
                          = 6.626176 x lO"34  watt-sec2
The speed of light,  c, is
                      c  = 2.99792458 x 108  meters-sec"1
                        = 2.99792458 x 1017  nm-sec'1

-------
Photon Flux or Radiant Energy Flux	Introduction


Thus the energy of a single photon of wavelength A is given by^

        "   fl>-1.9864775 xlO-»(i)   Joule-secnm-sec^
                                   \X/        photon

                = 1.9864775 x 10~16(y)   joule-photon-1
    Broadband, or spectrally integrated radiant energy is
/

                                     i,
and is measured in joules.
    Spectral radiant flux is dQ\/dt and is measured in watts-nm l.  Broadband
radiant flux, or just radiant flux, would be measured in watts.

    Spectral radiant flux density, £>, is radiant flux per area (a) or

                                  *  = dQ>
                                  X    dadt
and the units are usually watts-m~2-nm-1 or photons-cm-'-sec^-nm"1.

    Radiant flux density incident on a surface is called irradiance.

Components of Radiation at the Surface
The radiant flux density outside the earth's atmosphere is called the extraterrestrial
solar irradiance (#*)•  One of the most recent estimates of M\ was produced by the
World Radiation Center, Davos, Switzerland, and was reported in Iqbal.4  This
spectrum is shown as  the top line of Figure 1.

    On their way to the earth's surface, many of the photons incident on the atmo-
sphere undergo multiple scattering and  absorption processes (see  Figure 2). This
not only partially depletes the radiant energy before it reaches the surface, it also
affects the directional distribution of the photons. Notice that in Figure 2, the shape
of the scattering intensity around aerosol particles is different from that around gas
particles. This shape  is a complex function of the aerosol optical  properties. The
component of the surface flux that is not scattered or absorbed is called the direct
normal  spectral irradiance (that is, the amount of energy, or photons of a given
wavelength, per unit time incident on a unit surface that is perpendicular to the
direct solar beam and coming only from «5° view of sky containing the sun).

-------
Introduction
                                                       Photon Flux or Radiant Energy Flux
g
I
«J
   2.60
   2.40 -
   2J20-
   2.00-
    L80-
    L60-
    L40
    L20-
    LOO-
   0.80 -
   0.60-
   0.40-
   OJ20-
   0.00
                  World Radiation Center and LICOR 1800
        0.3
                          0*        .     0.7  ^ .
                                 _   (Thousands)
                                 Wavelenatn. nm
0.9
1.1
      Figure 1.  Extraterrestrial and Surface Solar Spectral Irradiance.
          Top: World Radiation Center recommended extraterrestrial solar
          irradiance, 2-nm resolution 300-700 nm, 10-20 nm resolution 700-
          1100 nm; Bottom:  Measured surface global irradiance with LiCor
          1800 spectroradiometer using 2-nm intervals and 4-n.m resolution.

-------
Photon Flux or Radiant Energy Flux
                                                                            Introduction
       RAYLEIGH - SCATTERED DIFFUSE
                                                             AEROSOL-SCATTERED DIFFUSE
                                   CIRCUMSOLAR
                                   RADIATION

      Figure 2.  Schematic Illustration of Gas and Aerosol Light Scatter-
          ing  Processes (top).  Bottom: Resulting direct and diffuse solar
          radiation components. (From Iqbal.4)

-------
Introduction	Photon Flux or Radiant Energy Flux

    Because many of the photons that are scattered from the direct beam undergo
multiple scattering from gas molecules and aerosol particles, the component that
is scattered can arrive at the surface from a full hemisphere (2?r steradians).  This
sky light is called the diffuse spectral spherical irradiance (that is, the amount
of energy  per unit time incident on a unit sphere with no view restrictions). The
diffuse spherical irradiance is often divided into downward and upward components.

    The combination  of direct normal and diffuse spectral spherical irradiance is
called  global spectral spherical irradiance (£A)>  When the units of $\ are
photons-cm~2-sec~l-nm~1, then  the term "actinic flux"  is often used.   Some re-
searchers have also used "actinic irradiance," "sealer irradiance,"  or "spherical irra-
diance." Without these qualifiers, the term irradiance generally means radiant flux
on a horizontal, or flat, surface.

    The spectral irradiance measured on a cosine response instrument, a so called
"flat" receiver, at solar noon in central North Carolina on a clear day is shown as
the bottom line in Figure 1. Figure 3 shows, in the ultraviolet range (300-400 nm),
an example of the various spectral spherical irradiance quantities described above.
Notice that, at wavelengths less than 370 nm, the diffuse component is larger than
the direct component.  This is  because Rayleigh scattering is  a strong function
(w A~4) of wavelength.

    As shown in Figures 2 and 3, not all of the radiant flux is downward. Part of
the diffuse spherical irradiance is caused by the reflection of some of the incident
radiation  from surfaces, When the source of the incident radiation is the sun, the
term "albedo"  is commonly used instead of reflectance.  In the meteorological lit-
erature, reflectance of the earth's atmosphere, as by clouds and air, is also called
albedo.4 In very general simple terms, the albedo (A) may be defined thus:
                         Radiation reflected from a surface
                     J\ *^~' •j-ujTuir	 -...	
                          Radiation incident on a surface
This definition includes the surfaces of aerosols and gas molecules so that, for exam-
ple, if one looks down from a location well above the ground, most of the reflected
radiation  is actually from the gas and aerosol below the viewpoint and only partly
from the generally low reflection from the ground below. Of course, the reflected
radiation  is itself reflected,  and thus multiple scattering produces additional flux
in all directions.  Albedo is usually wavelength dependent (spectral albedo, A\),
and the integral, over wavelength, of the flux-weighted spectral albedo is called the
integral albedo or just albedo (shown here without A subscript), that is,
                                 = # A,A, dX

-------
  Photon Flux or Radiant Energy Flux
                                                                       Introduction
                                 Zenith Angle»0
6
a
a
«*
*
      0.00
          300
320
340       360       380
  Wavelength, nm
                                                             400
420
        Figure 3. Example of Spectral Irradiance Components. Calculated
            from Schippnick and Green Fit to Dave Irradiance Model using
            World Radiation Center extraterrestial solar irradiance.

-------
Introduction
Photon Flux or Radiant Energy Flux
               300
                                                        60
        W 270
                                                          90  E
               240'
  120
                       210
                                                150
                                   180
                                    S
          Figure 4. Sun's Position in the sky at a location in central North
          Carolina For Five Months: a) June 22; b) August 24; c) Sept. 23;
          d) October 24,  e) Dec. 22.  Numbers along top arc are Eastern
          Daylight Time.
    Because of the sun's apparent motion through the sky, the path length traversed
by photons reaching the surface varies as a function of time  of day and location
on the earth's surface.  Longer atmospheric paths provide more opportunity for
scattering and absorption.  Thus, the relative contributions of diffuse and direct
radiation to global radiation varies as a function of local solar coordinates (zenith
angle). In addition, each "place" has a unique set of daily solar coordinates for each
day of each year. Figure 4 illustrates the approximate solar coordinates for central
NC for selected days in five months.

-------
Radiation Models	Introduction

Radiation Models
Radiative transfer models have been developed to predict the global spectral irra-
diance using measured values of M\ and descriptions of the atmospheric absorption
and scattering processes.  One such complex model developed by Dave5 was used
by  Peterson6 to predict the actinic irradiance at the earth's  surface, and at sev-
eral altitudes. This model assumed a cloudless, plane parallel, non-homogeneous
atmosphere. The model atmosphere was divided  into 40 layers, and each layer had
unique properties: concentrations of air molecules, water vapor, carbon dioxide, and
ozone; number density of aerosols; and temperature. A monochromatic unidirec-
tional solar flux was incident at the top of the layers, and the surface at the bottom
of the layers was an ideal diffuse reflector with a variable albedo. In each layer, the
attenuated direct solar flux and upward and downward diffuse components were
calculated.

    The primary  output  of the Peterson model was presented as the actinic flux
(units of photons-cm'^sec"1) over 5-nm and 10-nm intervals  from  295 nm to 700
nm at 10 zenith angles (x =0°, 10°, 20°, ... 86°) for albedos of 0.0 and 0.05.  In
addition, the changes in flux with altitude and albedo of 0.10  were  also  described.
The actinic flux was predicted to increase with  distance above the surface.  This
is due not so much to increased absorption of the downward component as it is
to a decrease in the upward  component when approaching the ground.  That  is,
when well above the surface there is a downward directed component, as well as a
significant upward directed component (about 1/3 of total flux) caused by reflection
from  the air molecules and aerosols below. At or near the  surface the downward
component is about the  same at higher up, but the upward  component is much
less (about 1/9 of the total flux) because the ground is not  as reflective as several
kilometers of air.  Near  the surface, then, the  total spherical flux is lower than
further away from the surface mostly because the upward component is lower near
the surface. Later we will show albedos measured at different altitudes. The higher
altitude albedos were much larger than the surface albedos as would be expected
based on the above description.

    Schippnick and Green7 have fitted a series of analytical  functions to  all five of
the Dave irradiance models and to Peterson's actinic flux model. These functions
use as input: extraterrestrial flux,  solar zenith  angle, altitude  (0  to 5 km), and
surface albedo (0-0.3). They approximate, to a high accuracy, the original computer
simulation outputs allowing for the simultaneous calculation  of global irradiance on
a horizontal surface (what meteorological instruments measure) and the actinic
flux, or global spherical irradiance (what a molecule absorbs). This Schippnick and
Green model is used throughout this chapter to produce the various irradiances

-------
Introduction _ Kinetics Parameters

shown.  An updating of the actinic flux calculation methods are needed, however,
because the World Radiation Center values for H\ are lower than the values used
by Peterson and newer information  is available on aerosol properties which will
affect the calculated actinic fluxes. Because there is no need to compute and match
global irradiance measurements in OZIPM model applications, the original Peterson
fluxes, with their older assumptions of H\ and other gas and aerosol distributions
were used in Chapters 3 and 4 photolytic rate calculations so as to be consistent
with other workers.

Albedo

In his 1976 calculations, Peterson assumed that 0.05 was the "best" albedo in the
UV region. It is important to distinguish between the "regional" albedo, which will
affect the area wide amount of light reflected upward to be scattered back downward,
and the "local"  albedo such as the surface under a spherical sensor. The present
literature is not conclusive as to what are reasonable values for albedos in the UV
region, and some satellite data suggests values as high as 0.15. Figure 5 illustrates
the calculated actinic irradiance in the ultraviolet region (295-400 nm) at 0° zenith
angle, at sea level for regional albedos of  0.05 and 0.15. These values probably
represent the range of available photon flux for a primary photolysis reaction at the
earth's surface in this wavelength region.

Kinetics Parameters
In effect, the absorption cross section is "how big an area" the molecule presents
to the photon flux at each wavelength.  At a given wavelength, a large cross section
means that the molecule is more likely to absorb a photon, and therefore react.

   Depending  upon the energy in  the absorbed photon, some molecules photo-
disassociate in more than one way.  For example, formaldehyde (HCHO) photolysis
can produce different products:
                             HCHO -**-* H2 + CO                           (2)
                             HCHO -^ H + HCO                          (3)

The quantum yield for a process is the ratio of the number of molecules that react
by a given path to the number of molecules absorbing photons in the wavelength
interval. For HCHO there would be two quantum yields, one for process (2) and one
for process (3). Of course, the quantum yield  for a given process or the combined
quantum yields for all processes for a particular species cannot be greater than 1.0.
A quantum yield of zero means that, although  the molecule may absorb photons at

                                     10

-------
Kinetic* Pmmtten
                                                                      Introduction
   8
   a
  7
   d
  CM
   6
   o
   s
  .8
0.30
OJ28
OJ86
024
0.22-
OJ80-
0.18 -
046-
                        WRC ET, ZA-0, Ait-0. Albgdo=0.05i 045
OJ2-
040-
0.08-
0.06 -
0.04-
0.02 -
0.00
            300        320       340       360       380       400
                                   Wavelength, nm
                                                                   420
      Figure 5. Theoretical Actinic or Spherical Irradiance at the Surface
          for Two Albedos.
                                      11

-------
Introduction	Example Calculation

the given wavelength, it does not undergo the given process. The quantum yield is
dimensioftless.

    The absorption cross sections and quantum yields for a species must be mea-
sured in carefully controlled laboratory situations; they cannot be calculated theo-
retically. Newer measurements using better instruments have improved the accuracy
and precision for these data and have resulted  in a series of changes in calculated
rates over the last 10 years. Evaluations of the experimental measurements have
been published by various groups, such as NASA,* which recommend values for use
primarily in stratospheric modeling.  Sometimes, where there is  some uncertainty,
the absolute values of absorption cross sections have been estimated by "fits" to
smog chamber experiments. That is, the measured values are used to define the
relative spectral response, but the absolute magnitudes are scaled up or down as
needed to fit model predictions to data.

Example Calculation

In this example, the actinic  flux values are based on the WRC M\ with a wavelength
interval of 2 nm, centered  at 300, 302, 304, ....  Thus the a\ and  $* values for
each species must be averaged over the same 2 nm interval so that the area under
the product curve will be correctly approximated.  When this averaging is done,
the integration operation in Equation 1  can be replaced by a summation over all
the 2 nm intervals in which the product of the three components is not zero.  For
nitrogen dioxide (NO]), for example:
                               420                      _
              J(x,y,^;No2)=  £ XA(x,y,A)AA y = 0, A = 0.05. Figure 6 shows the same process
in graphical form.

    Some type of solar location program (astronomical almanac) is used to predict
the zenith angle of the  sun as a function of time for  the particular location and
day of the year.  These predicted zenith angles are used to interpolate the species
photolytic rates.

OZIPM  Methods

Tables of Peterson's  surface "actinic  flux" as a function of wavelength and zenith
angle for an albedo of 0.05 were incorporated as built-in data in the original OZIPP

                                     12

-------
OZIPM Methods	Introduction

       Table 1. Calculation of Nitrogen Dioxide Photolysis Rate for Zenith
         ^-Angle = 0°, altitude = 0 m, and albedo = 0.05.
                                                          dJ/dX*
300
302
304
306
308
310
312
314
316
318
320
322
324
402
404
406
408
410
412
414
416
418
420


0.0016
0.0035
0.0063
0.0110
0.0162
0.0227
0.0315
0.0390
0.0459
0.0510
0.0575
0.0597
0.0590
0.2277
0.2182
0.2160
0.2213
0.2293
0.2351
0.2366
0.2394
0.2363
0.2415
2i

1.0685
1.1700
1.1700
1.4150
1.6600
1.7100
1.7600
1.7600
2.0050
2.2500
2.3950
2.5400
2.5400
6.7600
6.7600
6.5400
6.3200
6.0450
5.7700
5.7700
5.5850
5.4000
5.4500
imxl

* dJ/dX column has been
dj
dX
/ 1017 photon
1.0000
0.9976
0.9953
0.9929
0.9906
0.9882
0.9859
0.9835
0.9812
0.9788
0.9765
0.9741
0.9718
0.5650
0.4050
0.2875
0.1900
0.1375
0.0914
0.0743
0.0572
0.0401
0.0230
3&dJ/dX =

multiplied by
>\ /10-19cm2\ _.
0.0171
0.0409
0.0734
0.1546
0.2664
0.3836
0.5466
0.6751
0.9030
1.1232
1.3447
1.4771
1.4563
8.6968
5.9739
4.0613
2.6574
1.9059
1.2404
1.0146
0.7648
0.5114
0.3027
0.4419
min
1000

Vcm2-min-nmy V molecule ) """ ""*
                                       13

-------
Introduction
                                                                    OZIPM M.thodi
      CM
      O
      55

       C
       O
       O
       w
1   S  5   1
               I—•«•
 §  §   i

«—01
                                                                         '!
                                                  S 1  52  S S  S  I § I
                                   *

                                   I!

                                   ill
                                                         n*v( *M«*IM
       Figure 6.  Calculation of Nitrogen Dioxide Photolysis Rate Using
           World Radiation Center 2-nm Extraterrestrial Spectrum, Albedo
           of 0.05, and Zenith Angle = 0°.
                                        14

-------
Spectral Distribution of Rates	Introduction

and OZIPM2 computer code.  Similarly, species cross sections and quantum yields
were  alscrincluded as built-in data in OZIPP and OZIPM2.  At the beginning
of a simulation, these programs would calculate the species photolytic rates  as a
function of zenith angle and subsequently would interpolate these for the particular
zenith angle needed at the time in the simulation. In the OZIPM3/4 code, the first
step—calculation of species photolysis rate as a function of zenith angle—has been
eliminated along with the built-in data for actinic flux and species cross sections and
quantum yields. Instead, each species photolysis rate as a function of zenith angle
is stored as built-in data. These were computed external to OZIPM3/4 using the
method that was in OZIPM2, that is, using Peterson's actinic flux and up-to-date
species cross sections  and quantum yields. In OZIPM3/4, the user can also supply
species photolysis rates as a function of zenith angle as part of his input, and thus
override the built-in information.

Spectral Distribution  of Rates

In the Carbon  Bond  Mechanisms, the most important photoacceptor species are:
nitrogen dioxide (NO}), formaldehyde (HCHO),  ozone  (0$) photolyzing to atomic
oxygen  (Olr>), and acetaldehyde (RCHO).  As described above, HCHO has two path
ways: one to radicals  (HCHO,R) and one to stable products (HCHO,S).

    Using the actinic flux of Figure 5 (zenith angle of 0°) and currently recom-
mended species cross  sections and quantum yields, the relative, spectral photolysis
rates  (averaged over  10 nm intervals) of the four most important processes are
shown in Figure 7  (the area under each species curve is scaled to unity). Acetalde-
hyde's rate is much less than  HCHO, but its wavelength dependency is very similar
to HCHO,R. In addition, the relative spectral response of the Eppley  UV sensor is
included.  This  sensor is commonly used in field measurements and is used by the
University of North Carolina (UNC), SAI, and University of California at Riverside
(UCR) modeling groups to derive or adjust theoretical photolysis rates for simulat-
ing experiments in the UNC and outdoor UCR chambers. The figure shows that
the sensor only covers a part  of the total actinic spectrum. The sensor's response
is especially low at the short  wavelengths where most of the photolysis of O3 and
HCHO,R occurs.

    As shown in Figure 7, NOj  has a broad photolysis spectrum, extending into the
visible range (410 nm). O3  has a very sharp photolysis spectrum, with more than
half of the photolysis occurring in the interval 305-315 nm. This shorter end of the
spectrum is more affected by path length (more scattering) and thus the photolysis
rate of O3 will decrease faster with  increasing zenith angle than will that of NO2.

                                     15

-------
Introduction
                                                              Sp«ctr«l Distribution of R«t««
       CO
       o
       
 c

"o

 o
       Figure 7. Relative Spectral Photolysis Rates.



                                         16

-------
Spectral Distribution of Rates	Introduction

HCHO shows a broader photolysis spectrum than O3, and the rate is split between
the two processes that produce stable and radical products. The radical process for
HCHO is a very important source of radicals in all mechanisms.
                                      17

-------
                                                                   2
Review  of  Calculated   Rates
In this discussion, the photolytic rate data are presented as tables of rates vs. solar
zenith angle. Because Whitten also used the ratio of each rate to the NO} rate, these
ratio values are included (across the bottom of the tables) for comparison, but only
at a zenith angle of zero.  Table 2 explains the headings used in these tables. All
of the rates used  in all EPA models discussed in this report  appear in Table 3,
including the values recommended in this report. A few of these tables will appear
again in later chapters as their method of calculation is explained.

Effects of Changes  in Cross Section and Quantum Yield

The first set of theoretical rates used in EPA models were those of Schere and
Demerjian, 1977* (See Table 3, heading Schere77). These rates were based upon the
radiation model results of Peterson and the absorption cross sections and quantum
yields as they were known at the time. The absorption cross section for NOj changed
just as the report was being finished. An addendum was included giving the new
rates calculated with the new absorption cross sections. The old rates compared to
the new rates are shown to the right of the Schere77 table. This change subsequently
caused problems in the photolytic rates incorporated into the OZIPP program code,
and these will be discussed below.

   The Demerjian, Schere, and Peterson  1980 report,10 was a major revision of the
1977 work. In addition to using revised cross sections and quantum yield values,
this work examined the effect of altitude and albedo on rates.  The sea  level rates
are given in the Table 3, under the heading DemerSO. The rates  at various altitudes
will be discussed later.

   Carter's tt a/, final 1986 report11 on the new surrogate species mechanism used
reviewed information on cross sections and quantum yields as of February, 1986,

                                  18

-------
Effects of Changes in Cross Section and Quantum Yield
                                                           Review of Calculated Rates
         Table 2 Definitions of Headings used in Photolytic Rate Tables.

Sch«r«77       The rates calculated by  Schere and Demerjian in "Calculation of
               Selected Photolytic Rate Constants over a Diurnal Range," EPA-
               600/4-77-015, 1977.
DcaerSO        The rates calculated by Demerjian, Schere, and Peterson in "The-
               oretical Estimates of Actinic Flux and Photolytic  Rate Constants
               for Atmospheric Species in the Lower Troposphere,"  Adv.  Environ.
               Sci. TechnoL, 29, 1980.
CBS.84         The rates calculated by Killus and Whitten in "Technical  Discus-
               sions Relating To the Use of Carbon Bond Mechanism in OZIPM/
               EKMA," EPA-450/4-84-009,1984.
Cart«r86,8urf   The rates calculated by Carter, Lurmann, Atkinson, and Lloyd in
               "Development and Testing of a Surrogate Species Chemical Reac^
               tion Mechanism," Final Report, EPA Contract 68-02-4104, 1986.
VhittmiM.sur*  The rates calculated by Whitten in "Using CBM-X in EKMA with
               Computer Code OZIPM-3"; these surface rates were not explicitly
               described in the report but were determined  by back calculating
               from the altitude corrected values that  were used in the final re-
               port.
Vhitttn86,600B  The rates calculated by Whitten in "Using CBM-X in EKMA with
               Computer Code OZIPM-3" after altitude corrections were applied
               to the surface rates.
CB4.8urf .A-0.08 The rates recommended in this document.
CB4.640a.A-0.08 The rates recommended in this document.

ZA      Zenith angle.
J(N02)  Photolysis rate for NOj in units of min"1.
J(FR)    Photolysis rate for HCHO to radicals in units of 1 x 10~s min"1.
        Photolysis rate for HCHO to stable products in units of 1 x 10~3 min"1.
        Photolysis rate for CARB to radicals in units of 1  x 10~3 min"1. See text for
        CARB composition.
J(C8)    Photolysis rate for CARB to stable products in units of 1 x  10~3 min"1.
J(01D)  Photolysis rate for Os to OlD in units of 1 x 10~3 min"1.
J(CCHO)  Photolysis rate for acetaldehyde  to radicals in units of 1  x 10~3 min"1.
J(FS)
J(CR)
                                      19

-------
Review of Calculated Rates
Effects of Changes in Cross Section and Quantum Yield
Table 3. Theoretical Rates. Used in EPA Models.
"(all rates and ratios except NO2 have been mul
(all rates used A = 0.05 except those mark<
ZA
0
10
20
30
40
SO
60
70
78
86
ZA
0
10
20
30
40
BO
60
70
78
86
J(N02)
0.5129
0.5087
0.4963
0.4741
0.4395
0.3880
0.3168
0.2058
0.1014
0.0221
1.00
JCN02)
0.5326
0.5282
0.5159
0.4922
0.4562
0.4026
0.3287
0.2132
0.1050
0.0229
J(FR) J(FS)
2.14
2.11
2.01
1.85
1.61
1.30
0.92
0.49
0.20
0.04
5.60
5.53
5.33
4.98
4.45
3.73
2.84
1.62
0.72
0.15
4.17 10.92
J(Pt) J(F8)
2.23
2.20
2.09
1.90
1.62
1.27
0.86
0.43
0.16
0.03
2.96
2.92
2.81
2.61
2.32
1.92
1.41
0.80
0.34
0.07
J(010)
4.210
4.110
3.730
3.160
2.420
1.610
0.840
0.280
0.060
0.007
8.21
J(OID)
2.710
2.630
2.360
1.960
1.460
0.930
0.470
0.150
0.030
0.003
Itiplied by 1000)
;d otherwise)
Old
J(N02)
0.5784
0.5736
0.5595
0.5343
0.4950
0.4367
0.3562
0.2309
0.1109
0.0244










                                 1.00    4.19    5.66    5.09
                                                 20

-------
Effects of Changes in Cross Section and Quantum Yield
                        Review of Calculated Rates
Table 3, cont. Theoretical
"•(all rates and ratios except NO

ZA
0
10
20
30
40
50
60
70
T8
86


ZA
0
10
20
30
40
SO
60
70
78
86

J(N02)
0.4505
0.4559
0.4445
0.4245
0.3034
0.3472
0.2708
0.1840
0.0007
0.0104
1.00


J(N02)
0.5120
0.5087
0.4073
0.4741
0.4305
0.3880
0.3168
0.2058
0.1014
0.0221

J(FR)
1.81
1.70
1.70
1.54
1.32
1.04
0.71
0.36
0.14
0.02
3.05
1 Rates
2 have b
rter86 . £
J(FS)
2.78
2.75
2.65
2.47
2.21
1.84
1.37
0.70
0.35
0.07
6.05
-Vhitt«n86,8urf-
J(PR)
1.30
1.37
1.30
1.18
1.00
0.78
0.52
0.26
0.10
0.02
J(FS)
1.03
1.91
1.84
1.71
1.52
1.27
0.05
0.53
0.23
0.05
Used in EPA Models.
>een multiplied by 1000)
> ILEX
J(01D)
2.267
2.107
1.970
1.634
1.217
0.775
0.380
0.110
0.028
0.003
4.03


J(01D)
2.390
2.310
2.060
1.670
1.220
0.745
0.352
0.098
0.021
0.002

J(CCHO)
0.292
0.286
0.265
0.230
0.186
0.133
0.078
0.032
0.010
0.002
0.63












                                1.00    2.71
3.76
4.66
                                               21

-------
Review of Calculated Rates
Effects of Changes in Cross Section and Quantum Yield
Table 3, cont. Theoretical Rates Used in EPA Models.
"(all rates and ratios except NOj have been multiplied by 1000)

ZA
0
10
20
30
40
SO
60
70
80
86


ZA
0
10
20
30
40
50
60
70
80
86

J(N02)
O.S129
0.5087
0.4073
0.4741
0.4395
0.3880
0.3168
0.2058
0.1014
0.0221
1.00


J(N02)
0.6000
0.5952
0.5856
0.5642
0.5318
0.4851
0.4087
0.2882
0.1552
0.0348

J(FR)
1.74
1.71
1.64
1.49
1.30
1.05
0.74
0.40
0.16
0.03
3.39
,ot
J(FS)
2.06
2.05
1.97
1.83
1.65
1.38
1.05
0.60
0.26
0.06
4.02
-Vhitt«n86,600B-
J(FE)
1.79
1.79
1.68
1.54
1.33
1.07
0.74
0.40
0.16
0.03
J(FS)
2.39
2.39
2.29
2.15
1.95
1.68
1.30
0.79
0.38
0.08

J(01D)
2.390
2.310
2.060
1.670
1.220
0.745
0.352
0.098
0.021
0.002
4.66


J(01D)
3.240
3.150
2.820
2.310
1.710
1.080
0.530
0.160
0.040
0.005
                                 1.00    2.98
3.98
5.40
                                                 22

-------
Effects of Changes in Cro»« Section and Quantum Yield
                                                                          Review of Calculated Rates
Table 3, cont. Theoretical
"(all rates and ratios except NO

ZA
0
10
20
30
40
50
60
70
78
86


ZA
0
10
20
30
40
50
60
70
78
86

J(N02)
0.5358
0.5312
0.5166
0.4911
0.4524
0.3960
0.3184
0.2094
0.1025
0.0196
1.00


J0102)
0.5893
0.5851
0.5713
0.5470
0.5093
0.4537
0.3740
0.2578
0.1341
0.0242

J(FR)
1.89
1.85
1.76
1.59
1.36
1.06
0.71
0.36
0.14
0.04
3.53


J(FR)
2.18
2.14
2.04
1.86
1.60
1.27
0.87
0.45
0.17
0.04
1 Rate
2 have
.Surf.
J(FS)
2.98
2.94
2.83
2.63
2.34
1.95
1.44
0.83
0.37
0.10
5.56
.640m.
J(FS)
3.39
3.35
3.24
3.03
2.72
2.30
1.73
1.03
0.47
0.13
s Used in EPA Models.
j been multiplied by 1000)
»
A^U • UO
J(01D)
2.310
2.232
2.003
1.650
1.216
0.764
0.371
0.113
0.024
0.003
4.31


J(01D)
2.715
2.620
2.362
1.962
1.463
0.944
0.462
0.140
0.030
0.003

J(CCHO)
0.301
0.294
0.272
0.236
0.189
0.134
0.079
0.032
0.010
0.002
0.56


J(CCHO)
0.347
0.338
0.314
0.275
0.222
0.161
0.096
0.041
0.012
0.002
                                1.00    3.70    6.75    4.61
0.59
                                               23

-------
Review of Calculated Ratea	Effects of Changes in Cross Section and Quantum Yield

including corrections for errors that occurred in the printing of some previous re-
views of photolysis rates (personal communication from Carter, 1986). Photolytic
rates calculated from the cross sections and quantum yields used by Carter et al.
and  the Peterson 1976 actinic fluxes are shown in  the Table 3, under the heading
Carter86.  These rates were calculated by his program which assumes that all input
values were points connected by straight lines, even though HCHO data, for example,
were given as  IQ-nm interval averages centered at  the wavelength. The NO2 cross
sections and quantum yields  used by Carter are the same as those recommended
by Atkinson and  Lloyd  in their 1984 review.12  The NC>2 absorption cross sections
were the same as  those used in the Schere77 addendum (Bass et al.  data), and these
are still the recommended values in the latest NASA review.8 The quantum yields
used in the 1977 work, the 1980 work, and in Carter's work were different, however.

   The CB3.84 values were described in Killus and Whit ten's 1984  technical discus-
sion  document" on the OZIPM2 program.  The values in Table 3, however, were
extracted from the computer code and the CARB rates have been converted back to
HCHO rates. This  rates will  be described in detail in the next section.

   The Whitten86.Surf values in Table 3  were calculated by us using information in
the final report*  "Using CBM-X  in EKMA  with Computer Code OZIPM-3" and
in the computer code. These values do not actually appear in the  final report, but
are the basis of the altitude adjusted values that were recommended. The altitude
adjusted rates are given in the table headed Whitten86.600m in Table 3. The latter
two tables should be compared with the table headed CB3.84 in Table 3.

   A recent (November 1986) intercomparison of  photolytic rates was conducted
by Jeffries (UNC) among Schere (EPA), Carter (UCR), Lunnann  (ERT), Whitten
and  Gery (SAI) to establish a uniform treatment and to agree on sources of cross
sections and quantum yields for use in model testing.  NASA8 was the source for
NO}, HCHO, and Os. Carter  et al.11 was the source for CH3CHO. These cross sections
and  quantum yields were integrated  (depending upon whether the original values
were point values or interval averages) and re-interpolated to 5-nm interval averaged
values for use with Peterson's surface actinic flux data. The 5-nm interval averaged
cross sections and quantum yields are given in Table 4.

   The data in Table 4 were used with Peterson's surface flux, an albedo correc-
tion, and ratios of 640 meter photolysis rates to surface photolysis rates from the
Demerjian  et al.  1980 report to calculate the rates recommended in this report.
The  exact process will be described in Chapter 4.  The resulting rates are given in
Table 3 with the headings CB4.Surf.A=0.08 and CB4.640m.A=0.08.

                                     24

-------
Values Calculated For Carbon Bond III in OZIPM2	Review of Calculated Rates

    Figure 8 compares four species rates from these studies. Figure 9 compares the
quantum yields for NOj used in the OemerSO calculation and in the Carter86 calculation.
The different $> assumptions were responsible for the differences in the NO2 rates
shown in Figured.

Values Calculated For Carbon  Bond III  in OZIPM2

In the previous "Technical Discussions Relating To the Use of Carbon Bond Mech-
anism in OZIPM/EKMA,"13 Killus and Whitten describe the calculation of photol-
ysis rates for CBS in the OZIPM2 model. The method differed somewhat from the
procedure described in the  introduction above.

    Because OZIPM2 still had the Dodge Mechanism as the built-in default mech-
anism, the built-in photolysis data had to be applicable to that mechanism.  Thus
the photolytic data stored in OZIPM2 were the Peterson 1976 actinic flux and the
Schere and Demerjian 1977 cross sections and quantum yields. In effect the values
shown in Table 3, heading Schere77 were the rate constants  built  into the program.

    At the time Carbon  Bond III was implemented in OZIPM2, significant changes
had occurred in the cross sections and quantum yield values compared to the ones
used in the Dodge mechanism, and thus the program calculated photolysis rates
had to be corrected.  In addition, the Demerjian, et a/.,  1980 study had shown
that photolysis rates should increase with altitude.  Because the mixing height in
OZIPM2 often extended to  1500 m in typical applications, Killus and Whitten also
wanted to increase the calculated photolytic rates to account for this altitude effect.

    Finally, in the CBS  mechanism, carbonyls are combined  into a single species
CARB.  The characteristics of CARB were based on its approximating a mixture in
photochemical equilibrium with a constant composition:
Species Carbon Fraction
Formaldehyde
Higher Aldehydes
Glyoxal
Ketones
0.55
0.35
0.05
0.05
Thus, another correction factor had to be used to account for the CARB composition.
In Table 12 of the 1984 technical discussion report, the ratio of CARB photolysis to
HCHO photolysis was estimated to be 0.696 for radical product photolysis processes
and 0.65 for stable product photolysis processes.  Thus the "true"  HCHO radical
                                     25

-------
Review of Calculated Rate*
Value* Calculated For Carbon Bond III in OZIPM2
        Table 4. Absorption Cross Sections and  Quantum  Yields
                         (erou sections are 10~  cm -molecule~ , ba»« e)
NO2
nm
295-300
300-305
305-310
310-315
315-320
320-325
325-330
330-335
335-340
340-345
345-350
350-355
355-360
360-365
365-370
370-375
375-380
380-385
385-390
390-395
395-400
400-405
405-410
410-415
415-420
420-425
a
10.7
14.1
17.1
20.1
24.0
26.7
28.9
32.2
36.7
39.8
40.9
46.2
48.2
5L5
56.0
53.9
56.7
59.7
59.7
59.5
63.3
65.4
60.5
55.9
54.5
55.0
*
0.982
0.978
0.974
0.970
0.966
0.962
0.958
0.954
0.950
0.946
0.942
0.938
0.934
0.930
0.926
0.922
0.845
0.745
0.793
0.867
0.795
0.552
0.241
0.092
0.045
0.012
HCHO.R'
a
2.62
2.62
2.45
2.45
1.85
1.85
1.76
1.76
1.18
1.18
0.42
0.42
0.06
0.06












*
0.78
0.78
0.77
0.77
0.62
0.62
0.31
0.31
0
0
0
0
0
0












HCHO.S* CH3CHO
a
2.62
2.62
2.45
2.45
1.85
1.85
1.76
1.76
1.18
1.18
0.42
0.42
0.06
0.06












 a $
0.22 4.24 0.453
0.22 3.73 0.370
0.23 3.07 0.280
0.23 2.41 0.193
0.38 1.85 0.113
0.38 1.37 0.053
0.69 0.91 0.015
0.69 0.34 0
0.69
0.69
0.40
0.40
0.12
0.12












QlD
a $
55.81 0.900
28.84 0.898
14.98 0.753
7.56 0.311
3.87 0.043
2.02 0




















All a and <£ values were  averaged over 5-nm intervals except those marked  with  *, which were
averaged over 10-nm interval* (e.g., 295-305, 305-315), and therefore each of these values appears
twice for the 5-nm averaged actinic flux data (e.g., once for 295-300, once for 300-305).
                                           26

-------
Value* Calculated For Carbon Bond III in OZIPM2
                                                             Review of Calculated Rates
    <0
    OS,

    m
    "m
    51
    cu «
    09

    8s
    3^
    «d
    OS
O
K  5
                                              3  S  3  I  3  §  3 J  I  1
           
-------
Review of Calculated Ratea
                                             Values Calculated For Carbon Bond III in OZIPM2
      no
         290
~310   '   330   '   350      370   '   "'
               Wavelength, nxn
390      410   '   430
       Figure 9. Quantum Yields For Nitrogen Dioxide Photolysis.

                                       28

-------
Nitrogen Dioxide Photolysis in CB3/OZ1PM2	Review of Calculated Rates

photolysis  rate would  be multiplied by 0.696 to account for the composition of
CARB.    _

    So many corrections increased the likelihood of errors, and some were made in
the final specification of the photolysis rates for CBS used in OZIPM2.

    The  table headed  CB3.84  in Table 3  gives the photolysis rates  used by CBS
in OZIPM2.  These were computed by using built-in data in the OZIPM2 code.
For discussion purposes, the  CBS  CARB rate  has been converted back into HCHO
radical and stable rates in this table by "un-doing" the CARB composition correction
described above. For example the CARB to radicals rate at x = 0° was 1.21 min"1
in the OZIPM2 program.  The rate given  in the table headed CBS.84 in Table 3
is 1.74 = 1.21/0.696, which  would be the true HCHO,R  photolysis  rate assumed
by Killus and Whitten.  Likewise, the HCHO,S  rate in Table 3  was calculated as
2.06 = 1.34/0.65.

Nitrogen Dioxide Photolysis in CB3/OZIPM2
Killus and Whitten consulted the Demerjian tt a/., 1980 report to determine what
the J(NOj)  values should be for CBS when used in OZIPM2. The rates listed in the
1980 report at a zenith angle  of zero were
                         y,  meters J(0°, y, 0.05; NO2)
                                 0     0.5328
                              150     0.5652
                              640     0.6216

In Table 9 of the Technical Discussion report,15  Killus  and Whitten indicated that
OZIPM2 computed J(x, 0,0.05; NO]) according  to the older cross section data in
the Schere and Demerjian 1977 report which would give a value at x = 0° of 0.5784
min"1. Thus Killus and Whitten assumed that, with no changes, OZIPM2 would
calculate J(NOz) values that were comparable to those for an altitude somewhere
inbetween 150 and 640 meters using the newer data (which would have been about
a 9% increase compared to the ground value).  Therefore, no changes were made to
the 1977 NO2 photolysis rates  calculated by  OZIPM2 for use with CBS.
                                       •
   Inspection of the OZIPM2 computer code, however, shows that Killus and Whit-
ten were mistaken about the values built-in to OZIPM2; the values from the adden-
dum for the Schere and Demerjian report were the ones actually used in OZIPM2.
These gave a «7(NO2) value at x = 0° of  0.5129 min~x, less than the ground level
rate in the Demerjian et al., 1980 data and only  89% of the value that SAI thought
they were using.

                                     29

-------
Review of Calculated Rates _ Carbonyl Photolysis in CB3/OZIPM2


Carbonyl Photolysis in CB3/OZIPM2

Comparison of rates in Table 3 and Figure 8 shows that, compared to the rates
stored in the OZIPM2 program, the HCHO rates had decreased significantly by the
time Killus and Whitten were putting CBS into OZIPM2. They showed, in Table 10
of their Technical Discussion/5 that using what they thought was the Schere and
Demerjian 1977 data stored in OZIPM2, gives the ratios
J(0°, 0,0.05; HCHO, R)
  J(0°,0,0.05;N02)

J(0°, 0,0.05; HCHO, S)
  J(0°,0,0.05;N02)
                                                      ,
                                          = 9.68 x 10'
Based on their modeling of smog chamber experiments, however, Killus and Whitten
had to use lower rates to obtain good fits, and thus they concluded that better ratios
would be
                      J(0°,0,0.05; HCHO, R)             .
                        .7(0°, 0,0.05; N02)     *''*x™

                      J(0% 0,0.05; HCHO, S)             .
                        J(0°,0,0.05;N02)   = 3'76 X 10
Thus, Killus and Whitten stated that the built-in HCHO photolysis rates in OZIPM2
should be corrected by 2.71/3.7 = 0.73 for radicals and by 3.76/9.68 = 0.37 for sta-
ble products. Although the 0.37 "quantum yield update correction factor" appears
in Table 12 of the Killus  and Whitten report," the 0.73 factor appears  in the Ta-
ble 12 as 0.78. The actual "quantum yield update correction factor" used, however,
was 0.81 (shown by the final total correction factor being 0.564 which is 0.81 x 0.696,
the CARB to HCHO correction factor from above). Killus and Whitten argue (on page
42 of reference 13) that the difference between the 0.73  and the 0.81 accounts for
the increase in HCHO photolysis rate with height (which would have been about an
11% increase).

   The wrong assumption about which set of NO2 rates  were built-in to OZIPM2,
however, caused  these ratios to be wrong.  The actual Schere and  Demerjian 1977
data stored in OZIPM2 had the following ratios:
7(0°, 0,0.05; HCHO, R)
  J(0°,0,0.05;NO2)
                                          = 4.17 x 10~s
J(0°, 0,0.05; HCHO, 5)
  J(0°,0,0.05;N02)

                30
                                                 X 10
                                                      ,

-------
Ozone Photolysis in CB3/OZIPM2	Review of Calculated Rates

and the correct "correction factors" would have been 2.71/4.17 = 0.65 for radicals
and 3.76/10.92 = 0.35 for stable products.. Even using 0.81 as the correction factor
instead of the 0.65, however, did not compensate for the difference in the NC>2
photolysis rates, and thus the actual rates used were not at all what was intended.

    Comparison of the HCHO photolysis rates plots in Figure 8 shows that, m at-
mospheric applications, CBS in OZIPM2 had what are now known to be low values
of HCHO photolysis rates, and therefore most likely would have a lower atmospheric
reactivity than it  should have.  It would thus be quite sensitive to 'external' radical
sources, such as NMOC aloft as has been determined in several SIP-like cases.14 This
large sensitivity is probably not correct, because it is most likely  an artifact of the
anomalously low HCHO photolysis rates.

    A major reason for Killus and Whitten having chosen low ratios for
J(HC«O)/J(NO2) was that these values produced the best fits  to a large number of
UNC outdoor smog chamber experiments. Recent measurements of the irradiance
inside and outside the UNC chamber has shown, however, that the Teflon chamber
walls are  not 100% transparent in the 300-400 nm region.  Further, the interac-
tion between the film transmission and the reflective floor of the chamber results in
the NO] photolysis rate being slightly higher inside  the chamber than outside the
chamber,  while the  HCHO photolysis is lower inside than outside.  The best fits are
produced  by using the CB4.Surf.A=0.08 values for the atmospheric (outside chamber)
and using a chamber light model that accounts for both the film  transmission and
the floor albedo.  This model has been described  by Jeffries and Sexton.15  Thus,
while the lower HCHO photolysis  rates chosen by Killus and Whitten are appro-
priate when simulating the  chamber, they are inappropriate when modeling the
atmosphere.

Ozone Photolysis in CB3/OZIPM2
Killus and Whitten  applied a large correction factor to the O3 photolysis rate built-
in to OZIPM2. Part of the correction  was due  to changes in the cross section data
reported by Demerjian et a/., 1980, but a further correction was made to reduce the
photolysis rate based on some atmospheric measurements. The final correction was
0.53 times the Schere and Demerjian  1977 rates.  This resulted in rates that were
slightly higher (6%) than the surface rates calculated by Carter et a/.,  in 1986.
                                     31

-------
Review of Calculated Ratea	How CB3/OZIPM2 Rates Influenced CB4/OZIPM3/4 Rates

How CB3/OZIPM2 Rates Influenced CB4/OZIPM3/4 Rates
The base set of values used to calculate the CB4 photolytic rates recommended by
Whitten in his final report1 are listed under the heading Whitten86.Surf in Table 3. By
comparing these'with the CB3.84 rates and with the values of J(HCHO)/ J(NO2)  (e.g.,
2.71 and 3.76) described by Killus and Whitten in their Technical Discussion report,
you can see that the errors made in the earlier work were carried forward.  Thus
the values recommended by Whitten in his 1986 report depended upon the Schere77
NO; rates, upon the empirically determined value of HCHO-to-NO; photolysis  rates
for inside the UNC chamber, and upon the ratio-adjusted Schere77 ozone photolytic
rates. Because the ultimate recommendations, the values listed under Whitten86.600m
in Table 3, incorporated a new altitude correct not used in CB3/OZIPM2, the
actual sources of the recommended rates were hard to determine. The  final report
recommendations were simply the values listed under Whitten86.Surf corrected to a
higher altitude. So while the final recommended values were similar to the values
recommended in this report, the technical support for the recommendations were
lacking..

Affect of Altitude  on Photolysis Rates

As indicated above, the Demerjian it al. 1980 study computed rates as a function
of altitude and showed large increases (factors up  to 1.5 times the surface rate).
Because the mixing height often extends to 1500 meters in OZIPM3/4 applications,
the photolytic rates should be increased above the surface values.  We will assume
that the rates at 640 meters can- serve as approximate average rates for a 0-1500
meter mixing depth.  Table 5 gives the NO} rates  computed by Demerjian  et al.
1980 for 0, 640, and  1380 meters.  We will assume that the relative ratios of  these
values reflect the altitude effect  and that these ratios can be used to adjust surface
computed rates using somewhat different assumptions of o\, and $>. Thus Table 5
includes, at the bottom, the ratio of the 640 m rate to the surface (0 m) rate  when
the ground albedo is 0.05. Whitten used these ratios and the values in Whitten86.Surf
to compute Whitten86,600m, e.g., 0.5129 x 1.17 = 0.6001. We will perform a similar
adjustment for our recommended rates later in Chapter 4.
                                     32

-------
Affect of Altitude on Photolysis Rates	Review of Calculated Rates
Table 5 Nitrogen Dioxide I

ZA
0
10
20
30
40
so
60
70
78
86

J(N02)0
0.5326
0.5282
0.5159
0.4922
0.4562
0.4026
0.3287
0.2132
0.1050
0.0229
'hotolysis ]
Demer80-Al1
J(N02)640
0.6216
0.6180
0.6072
0.5868
0.5538
0.5028
0.4236
0.3000
0.1602
0.0312
Elate as a Function of Altitude.
L j 4»«dlA«i _«»_ _
b IbUaG* "" " ~~
J(N02)1380
0.6834
0.6678
0.6576
0.6384
0.6072
0.5568
0.4770
0.3486
0.1962
0.0372
                                 1.00       1.17      1.28
                             —Ratio  640 to 0  a. A-0.06-
                      ZA     N02      HCHOR   HCH08    010
0
10
20
30
40
50
60
70
78
86
1.17
1.17
1.20
1.20
1.21
1.25
1.29
1.40
1.53
1.37
1.28
1.28
1.29
1.31
1.33
1.37
1.42
1.49
1.52
1.50
1.24
1.24
1.25.
1.27
1.29
1.32
1.37
1.44
1.49
1.45
1.36
1.35
1.38
1.39
1.42
1.46
1.51
1.58
1.67
1.00
                                             33

-------
                                                                  3
Atmospheric   Rates
Measurements of Nitrogen Dioxide Photolysis Rates

A number of measurements of the atmospheric photolysis rate of NOj have been con-
ducted.16'17'18'19'20 Two sets of measurements, for example, are shown in Figure 10
from the Harvey, Steadman, Chameides work in 1977. The j\ refers to the mea-
sured NO] photolysis rates using a quartz flow reactor, and the Eppley UV values
were measured with the same type instrument as used at UNC. The points are the
measured values, the line were computed by Harvey et al. using 1977 theory and are
not relevant here. You should notice that the maximum measured ./(NO?) photolysis
rates are greater than the  theoretically calculated maximum NOj photolysis rate at a
zenith angle of (f given the presently accepted cross sections and quantum yields of
Table 4 and the Peterson  actinic flux at A = 0.05. That is, using the "best" values
of ^(NOj) and $A(NOJ) and the Peterson actinic irradiances for "best case" albedo,
the predicted maximum value is 0.503 min""1, compared with a measured value of
about 0.54-0.55 min"1. Thus measurements are about 1,1 times larger than theory.

   Parrish et al. conducted an NO} actinometry study in Colorado19 in 1983. They
could fit their clear sky values with an equation of the form
                    j(NOz)= 0.7830 e-OS60"ex   min-1                  (5)
and all clear sky measurements fell within ±7% of this equation. This equation
predicts a x = 0° value of 0.5463 min"1, in close agreement with that shown in
Figure 10 for Michigan, but also larger than values predicted by present theory.

Comparison of Theory and Measurement for Trradiance

There are only three items that can be adjusted in the theory to improve its agree-
ment with NO} actinometry measurements: actinic flux, absorption cross sections,
and quantum yields.

                                  34

-------
Companion of Theory and Measurement for Irradiance
                                Atmospheric Rates
                  E
                  o

                  s:   4
                 UJ
                                    0.2
 0.4

[mirr1]
                                                                             UJ
                      •r 0
                         s


                      + 30
                                                                       f 455
                                                                            UJ
                                                                            M
                                                                       .. so,
                                        0.6
                                                                         90
                                             l corr*l«t<«« MMM Ewl«r U»
                                            Jf for i elxr t«y. Oiu 
                              •r* ft«m ««ui uw M>f*» nolui 41151« MISI»K far cut
                       • SUNRISE TO  8<30 AJvL  clear

                       • 8>30 AM.  TO II-3O AM,  portly doudy

                      A ll<30 AM.  TO 1-30 RM.  ovwccnt
             I  3
             t£

             5  2
             a.
             o.
             UJ  I
at     0.2
                                       O.3     0.4    O.5
                                           jt [min-i]
            0.6
        Figure  10.   Examples of Measured Atmospheric Rates  of Nitrogen
            Dioxide Photolysis.
                                               35

-------
Atmospheric Ratea _ UV Albedo

    The NOj cross section values in current use were obtained by JBass et al.,21 with a
resolution-of 0.125 nm and an estimated absolute accuracy of ±10%.  These values
were generally confirmed in the range 375-420 nm by Harkner et al.,22 who also
established the quantum yield with a resolution of 1 nm in this same range.  The
low $> values  reported by Harkner et al. for  A  > 375 nm have been confirmed by
Davenport.23 Thus, while these factors  may  yet change with new measurements,
there is no evidence for making ad hoc adjustments in these kinetic values, and we
will continue to use the values as recommended by the NASA review.8

    We have recently  compared calculated global irradiance (300-420 nm region)
with computerized spectroradiometer global irradiance measurements. The calcu-
lated values were produced by the Schippnick and Green7 fits to the Dave spectral
irradiance model and to the Peterson actinic  flux model.  We updated  the ex-
traterrestrial flux to the World Radiation Center 2-nm flux. We found excellent
agreement over a zenith angle range of 14-80°. Examples of model predictions and
measurements from shortly after sunrise to solar noon are shown in Figure 11.

    As was shown in  Figure 3, however, actinic flux, or spherical irradiance, has
a significant upward component that is not sensed by an upward  looking, cosine
response (flat)  receiver such  as the spectroradiometer or Eppley UV meter.  The
downward scattering of this upward component  has little effect on the flat,  upward
looking sensor, but as shown in Figure 12, an increase in surface albedo from 0.05
to 0.15 can have a large effect (22% increase) on global actinic irradiance  while it
only effects the global horizontal irradiance a small amount (< 2% increase). So the
fact that the horizontal irradiance model can fit horizontal irradiance measurements
still leaves the possibility that the regional albedo, or its impact on actinic flux may
not be correctly represented in the Peterson "best estimate surface albedo" actinic
flux data set.

UV Albedo

The present literature on ultraviolet albedo is  not conclusive as to what are reason-
able values for the "regional" albedo that should be used in the radiation transfer
model calculations. Few spectrally resolved albedo values are available. Peterson
used measured data from  Coulson and Reynolds74 to arrive at  the spectral albe-
dos shown in Table 6. Much of the literature  only discusses the integral  albedo,
that is, the  spectrally-weighted or "broadband" values. Using the solar spectrum
in Figure 1, and the equation
                                     36

-------
UV Albedo
                                                              Atmospheric Rate*
33
                332333
532233333-33533333:
        ssgssssssssss
                       s
                       ?
     Figure H. Comparison of Schippnick and Green Fit to Dave Radi-
         ation Model With LiCor 1800 Spectroradiometer Measurements,
         2-nm intervals, 4-nm resolution.
                                   37

-------
 Atmospheric Rate*
         UV Albedo
a
a
a
8
     2.40
                               Zenith Angle«0
                              340       360       380
                               Wavelength, nxn
400
420
       Figure 12.  Comparison of Horizontal and Spherical Irradiance at
           Two Albedos.

                                      38

-------
UV Albedo	Atmospheric Rates

we  can calculate the integral albedo that matches  Peterson's assumptions.  By
keeping Peterson's distribution assumptions as given  in Table 6, but increasing the
magnitudes uniformly, we can estimate both "broadband" or spectrally averaged
albedo(A) and UV albedo (e.g., A^o)- For example,
                          increase     A       A&Q
                               1.0     0.11     0.05
                               1.4     0.15     0.07
                               1.6     0.17     0.08
                               2.0     0.22     0.10

    Dickerson et al.18 reported albedos measured with the Eppley UV radiometer
for  various surfaces and from aircraft flights equipped with an upward and down-
ward looking Eppley UV radiometer system. These are reported in Table 7.  Doda
and Green25'26 also measured UV reflectances from  airplanes.  These researchers
extrapolated their altitude data to the surface and estimated surface reflectivity.
Some of their measurements are presented in Table 8. The Doda and Green values
for  pine forest are a factor of three lower than Dickerson's values for rain forest at
the same altitude.

    King and Herman27  and King28 calculated ground albedo from  a  statistical
fit using empirically determined  ratios of direct to  diffuse  light at a number of
wavelengths. They determined a mean ground albedo of 0.279 at wavelengths of
521  and 670 nm for an  area  near  the center of the city of Tucson  AZ.  This is
significantly higher than  the 0.10 value  at 520 nm and the  0.15 value at 670  nm
assumed in the Demerjian et al. 1980 study (see Table 6) suggesting that  the values
should be increased by nearly a factor of two.

    Satellite data29 indicate that the minimum integral albedo (200-4000 nm) over
the U.S. is  about 1.4  times higher  than  the  0.11 calculated from Demerjian et
al.  assumed values.  The whole U.S. (except for  the  Western deserts which have
minimum albedos of >  0.20) has a minimum measured integral albedo of  0.15-0.20.
Otterman and Fraser*0 used models of the scattering  properties of the  atmospheric
aerosols and LANDSAT data for the southwest U.S. to calculate surface  albedos of
0.264 in the region 500-600 nm with little sensitivity to the aerosol model assumed.
This value is consistent with the values found by King, and reported by the NIMBUS
3 data29 again suggesting that larger albedo values should be used in the photolytic
rate calculations.

    Using data from another satellite, the Atmosphere Explorer-E, Frederick and
Abfams31 also used  a  numerical model of radiative transfer, including multiple

                                     39

-------
Atmospheric Rate«	UV Albedo


                Table 6. Albedo Values Assumed by Peterson.

                                   A    albedo
290-400
400-450
450-500
500-550
550-600
600-640
640-660
660-700
0.05
0.06
0.08
0.10
0.11
0.12
0.13
0.15
                Table 7, a. Surface Albedo Values Measured by
             Dickerson et a/., using Eppley UV Radiometer.
surface
grass
black cloth
white plywood
road surface
cement
albedo
0.01
0.02
0.07
0.07-0.09
0.17
                Table 7, b. Near Surface Albedo Values Measured
             by Dickerson et al. Using Modified Eppley UV Radiome-
             ter.
                    surface     altitude, km.      albedo

                 rain forest         0.05-0.30      0.06-0.18
                high plains               0.3      0.12
                                          2.9      0.33
                                     40

-------
Adjustment of Calculated Actinic Flux for Albedo                              Atmospheric Rates


           Table 8. UV Albedo Values Measured by Doda and Green.

                    surface      altitude, km.       albedo
- pine forest
desert sand,
w. Texas
gypsum sand,
White Sands NM
0.0-0.33
0.0-0.33
0.0-0.33
0.02-0.04
0.05-0.07
0.16-0.19 (300 nm)
0.53-0.60 (400 nm)
scattering and surface reflection to calculate the distribution of albedo at two wave-
lengths, 331.2 and 339.8 nm. There were 7421 measurements; 69% produced albedo
values less than 0.30 and higher values were interpreted to include reflection from
clouds. Eleven percent of the albedo values were between 0.0 and 0.10, 29% were
between 0.10 and 0.20, and 20% were between 0.20 and 0.30. As expected, there
was no systematic variation of albedo with zenith angle.

    Based on such recent evidence, other newer radiation models have used regional
UV albedos of » 0.15 (see for  example Logan et al.32).

    We will therefore multiply the values in Table 6 by 1.6 to obtain integral or
"broadband" albedo of 0.17 and a UV albedo of 0.08. The value 0.08 will be used
to adjust the actinic flux before photolytic rate calculations are performed in the
rest of this report.

Adjustment of Calculated  Actinic Flux for Albedo
Information is supplied in the  Demerjian et al. 1980 study  to estimate the effect of
albedo changes up to 0.20.  Demerjian et al. give a table of increases in actinic flux
as a function of wavelength and zenith angle as albedo was increased from 0 to 0.10.
They state that the changes were approximately linear with albedo. We have used
this table to produce an albedo enhancement factor as a function of A and \. These
factors are used to adjust the spectral actinic irradiance when computing rates.

    If we apply a "regional" albedo correction to the Peterson actinic flux (that is,
increase the albedo from 0.05 to  0.08), then surface NO2 photolysis rates computed
from a\ and $\ in Table 4 are  in much better agreement with  Parrish et al. at-
mospheric measurements.  Computed values are given in Table 9 and are shown in
Figure 13.  The values in Table 9 that are under the heading 'Fit to Observations'
are predicted from Equation 5. The bottom line in Figure 13 shows (and the first
column of Table 9 gives) the values predicted from theory using the Peterson actinic

                                     41

-------
Atmospheric Rates	Adjustment of Calculated Actinic Flux for Albedo

fluxes and current cross sections and quantum yields. The second column (and the
next higher line in Figure 13) is the result of assuming a "regional albedo" of 0.08
rather than 0.05 in the standard flux calculations.  The third column is the result
of applying the "site specific"  corrections calculated by Parrish for his particular
location (e.g., elevation, aerosols, and total column 0$). The last column shows that
the fitted observations and the theory agree to within about 2% for small zenith
angles and within about  5% for large zenith angle where  the  site corrections are
quite large.
                                       42

-------
Adjustment of Calculated Actinic Flux for Albedo
Atmospheric Rates
Table 9. Comparison of Higher Albedo Calculated Photolysis Rates
^_with Atmospheric Measurements. .
	 Effect of Albedo on J(N02) —
ZA
0
10
20
30
40
50
60
70
78
86
A-0.05 A-0.15
0.5031 0.5358
0.4991 0.5312
0.4866 0.5166
0.4643 0.4911
0.4300 0.4524
0.3794 0.3969
0.3062 0.3184
0.2023 0.2094
0.0990 0.1025
0.0188 0.0196
SiteAdj FitObs Obs/Theory
0.5498
0.5461
0.5342
0.5127
0.4802
0.4326
0.3629
0.2584
0.1383
0.0240
0.5463 0.99
0.5433 0.99
0.5338 1.00
0.5167 1.01
0.4894 1.02
0.4472 1.03
0.3811 1.05
0.2733 1.06
0.1386 1.00
0.0045
	 Adjustment Factors 	
ZA
0
10
20
. 30
40
50
60
70
78
86
Albedo
1.06
1.06
1.06
1.06
1.05
1.05
1.04
1.03
1.04
1.04
Site
1.03
1.03
1.03
1.04
1.06
1.09
1.14
1.23
1.35
1.23
Total
1.09
1.09
1.10
1.10
1.12
1.14
1.19
1.28
1.40
1.28
                                                  43

-------
   Atmospheric Rate*
Adjustment of Calculated Actinic Flux for Albedo
 I

T
*
1
       0.60
       0.50
       0.40 -
       050 -
       0.20 -
       0.10 -
       0.00
                          Comparison of Theory with Measurement
          Figure 13.  Comparisons of Adjusted Theory with  Parrish et al.
             Measurements of Atmospheric NOj Photolysis Rates.19
                                           44

-------
                                                                  4
 Recommended   Rates
In this chapter, the actinic flux, kinetics parameters, and altitude effects from Chap-
ter 2 will be combined with the re-estimated best albedo value of 0.08 from Chapter 3
to estimate atmospheric rates at the surface and at 640 meters for use in OZIPM3/4
model program.

Rates for Carbon Bond IV in OZIPM3/4

Surface Rates
To compute surface rates, the actinic flux of Peterson, the albedo effect adjustment,
and the cross sections and quantum yields given in Table 4 were used.

   The resulting rates, which would be the values to use for CB4 in a ground level
application, are shown in Table 10.

Mixed Layer Rates
As described near the end  of Chapter 2, there  is a strong effect of altitude on
photolytic rates. This is because, as stated in Chapter 1, the air below a given
altitude well above the ground has a greater reflectivity upward than the absorbing,
low albedo ground. Therefore, the spherical flux at an altitude is higher than at
the surface. Table 5 gave the ratio of photolysis rate at 640 m to that at 0 m for
the "best, albedo" assumptions of Demerjian et al. (0.05). If the albedo assumption
of 0.05 is too low, then the ratios in Table 5 are too high because the surface values
are too low. Thus the ratios in Table 5 are not appropriate to adjust the rates of
Table 10.

   As an approximation to  the correct ratios for an albedo of 0.08, we will use the
ratio of photolysis rates at 640 m to those at 150 m for an albedo of 0.05. That is,

                                  45

-------
Recommended Rates	Recommended CB4 Photolytic Rates

we assume that
                     J(0°, 640,0.05; S) ^ J(0°, 640,0.08; S)
                     J(0°, 150,0.05;S) *  J(0°,0,0.08;S)
since at the higher albedo, there will be less difference between the higher altitude
rates and the surface rates because the surface is more reflective. These ratios are
shown in Table 11.

   Applying the ratios of Table  11 to the surface values of Table 10, gives the
recommended values for the mixed layer.  These are shown in Table 12.

Recommended CB4 Photolytic  Rates
The values in  Table 12 are the photolytic rates that should be used in OZIPM3/4
with the Carbon Bond Four Mechanism when performing control strategy calcula-
tions.

Rates for  Carbon Bond III in  OZIPM3

Because of previous use of CBS in making SIP calculations and because there were
a number of errors made in calculating the photolytic rates in OZIPM2,  we have
made new recommendations for photolytic rates for use with CBS in OZIPM3/4.

   The calculation of these new CBS rates follows the same procedure as that for
CB4 except that the carbonyl weighting factor for radical processes and for stable
process (see discussion in Chapter 2) has been recalculated and applied to the HCHO
photolysis rates to produce the CARB photolysis rates.  The original weighting factor
for production of radicals (wr) was

       wr = (0.55 + 0.005 + 0.0175) J(HCHO, R) + 0.35(J(CH3CHO)/J(HCHO))

where the numerical coefficients reflect the CARB composition as described in Chap-
ter 2. The coefficient of the last term, J(CH3CHO)/J(HCHO), changed from 0.35 to
0.16  for the new photolytic rates, thus lowering wr to 0.602  compared to its old
value of 0.695. The factor relating HCHO,S rates to CARB,S rates remains the same
as before, 0.65.

   The recommended  rates for CBS in OZIPM3/4 are given  in Table 13  and Ta-
ble 14.
                                     46

-------
Rates for Carbon Bond III in OZIPM3	Recommended Rates
Table 10.
Recommended CB4 Surface Photolytic Jlates.
(all rates and ratios except NO2 have been multiplied by 1000)
n-o A o._ ^ A f\ r\a
ZA
0
10
20
30
40
50
60
70
78
86
J(M02)
0.
0.
0.
0,
0.
0.
0.
0,
0.
0.
5358
5312
5166
4911
4524
3969
3184
2094
1025
0196
J(FR) J(FS)
1.
1.
1.
1.
1.
1.
0.
0.
0.
0.
891
859
759
593
359
062
714
360
138
036
2
2
2
2
2
1
1
0
0
0
.980
.943
.828
.631
.341
.947
.436
.830
.367
.103
. \JQ 	
J(03)
2
2
2
1
1
0
0
0
0
0
.310
.232
.003
.650
.216
.764
.371
.113
.024
.003
J(CCHO)
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
301
294
2?2
236
189
134
079
032
010
002
                         1.00    3.53    5.56    4.31    0.56
             Table 11. Altitude Adjustments Corrected For Albedo.

                        Ratio 640 to 150m, to approx A-0.08
                   ZA   N02     HCHOR   HCHOS   010     CCHO
0
10
20
30
40
50
60
70
78
86
1.
1.
1.
1.
1.
1.
1.
1.
1.
1.
10
10
11
11
13
14
17
23
31
24
1
1
1
1
1
1
1
1
1
1
.15
.15
.16
.17
.18
.20
.22
.24
.24
.20
1
1
1
1
1
1
1
1
1
1
.14
.14
.14
.15
.16
.18
.21
.25
.27
.23
1
1
1
1
1
1
1
1
1
1
.17
.17
.18
.19
.20
.24
.25
.24
.25
.00
1
1
1
1
1
1
1
1
1
1
.15
.15
.16
.17
.17
.20
.22
.26
.26
.17
                                        47

-------
Recommended Rates                                     Example Predictions Using New Rates


         Table 12 Recommended CB4 Mixed Layer Photolytic Rates.

              (all rates except N02 have been multiplied by 1000)


                      	--CB4,640m,A=0.08	
                  ZA   J(N02)   J(FR)   J(FS)    J(03)   J(CCHO)
0
10
20
30
40
50
60
70
78
86
0 . 5893
0.5851
0.5713
0.5470
0.5093
0.4537
0.3740
0.2578
0.1341
0.0242
2.179
2.141
2.036
1.858
1.599
1.272
0.873
0.448
0.172
0.043
3.391
3.353
3.237
3.032
2.724
2.304
1.732
1.034
0.466
0.127
2". 715
2.620
2.362
1.962
1.463
0.944
0.462
0.140
0.030
0.003
0.347
0.338
0.314
0.275
0.222
0.161
0.096
0.041
0.012
0.002
                       1.00   3.70   5.75    4.61    0.59
Example Predictions Using New Rates

Simulation Conditions
To illustrate the effect of the new recommendations, some example simulations will
be compared.  Example 1  (CB4) and Example 2 (CBS), which were used in the
OZIPM3 guidance document1, were used as sources of conditions. The only differ-
ence between Example  1 and Example 2 was the change in mechanisms.  Table 15
gives the simulation conditions.  In all examples, the Characteristic Curve for Mixing
Height was used. All simulations began at 0800 LDT and ran for 10 hours.

    In the control cases, the initial  HC was reduced by various amounts and no
reduction occurred for NOX.
                                    48

-------
Simulation Conditions  	Recommended Rates

            Table 13.  Recommended CBS Surface Photolytic Rates.
               (all rates except NO2 have been multiplied by  1000)
ZA
0
10
20
30
40
50
60
70
78
86
J(N02)
0.5358
0.5312
0.5166
0.4911
0.4524
0 . 3969
0.3184
0.2094
0 . 1025
0.0196
	 I/DO , Ol
J(CR)
1.139 .
1.119
1.061
0.964
0.827
0.653
0.447
0.237
0.104
0.040
in , A=U . i
J(CS)
1.937
1.913
1.838
1.710
1.522
1.265
0.933
0.539
0.239
0.067
JO 	
J(03)
2". 3 10
2.232
2.003
1.650
1.216
0.764
0.371
0.113
0.024
0.003
                         1.00    2.13    3.62    4.31
         Table 14. Recommended CB3 Mixed Layer Photoly tic Rates.
               (all rates except NO2 have been multiplied by 1000)
. ZA
0
10
20
30
40
50
60
70
78
86
J(N02)
0.5893
0.5851
0.5713
0.5470
0.5093
0.4537
0.3740
0.2578
0.1341
0.0242
	 UDG ,*>
J(CR)
1.312
1.289
1.228
1.124
0.972
0.781
0.547
0.295
0.129
0.048
*um , A~W . i
J(CS)
2.204
2.180
2.104
1.971
1.770
1.498
1.126
0.672
0.303
0.083
JO 	
J(03)
2.715
2.620
2.362
1.962
1.463
0.944
0.462
0.140
0.030
0.003
                         1.00    2.23    3.74    4.61
                                     49

-------
Recommended Rates
                                                              Simulation Conditions
Table 15, a.
Item
Latitude
Longitude
Date
Initial Mixing Ht.
Final Mixing Ht.
Initial NMOC
Initial NOX
Surface NMOC
Surface Os
Aloft NMOC
Aloft 03
Temperature
Water Vapor
Initial Conditions.
Value Units
39.93
75.10
6/24/80
250
1235
1.100
0.120
0.038
0.010
0.040
0.070
303
25000
degrees
degrees
mm/dd/yy
.. meter
meter
ppmC
ppm
ppmC
ppm
ppmC
ppm
K
ppm
                        Table 15, b. NMOC Fractions.
         Source    PAR    ETH   OLE   TOL    XYL   RCHO  HCHO    NR
Emissions
Surface
Aloft
0.51
0.53
0.61
0.04
0.05
0.06
Table
Hour
NMOC
NOX
1
0.17
0.35
2
0.17
0.35
0.03
0.03
0.03
15, c.
3
0.17
0.35
0.12
0.07
0.08
0.10
0.06
0.07
0.03
0.16
0.09
0.02
0.10
0.06
0.16
0.00
0.00
Emissions Fractions.
4
0.10
0.19
5
0.02 0
0.03 0
6 7
.02 0.02
.03 0.03
8
0.02
0.07



                                      50

-------
Simulation Results	Recommended Rates

Simulation Results
Figure 14 shows the NO, NO2, and O3 profiles for CB4 using the surface photolytic
and 640 m photolytic rates from Tables 10 and 12 for the Example 1 case. The higher
photolytic rates (mixed layer rates) resulted in about a 5% increase  in predicted
maximum O3 for these conditions.

    Figure 15 shows the NO, NO2, and O3 profiles for CBS using the 640 m photolytic
rates from Table 14 and the older 1984 photolytic rates from Table 3 for the Example
2 case.  The-higher photolytic rates (new rates) also resulted in a 5% increase in
predicted maximum O3 in this N0x-limited  situation. That the increase was  not
greater was caused by the fact that the much larger increase in CARB,S rates offset
the increase in CARB,R rates and less CARB photolyzed to radicals in the simulation
using the higher 640 m rates than did in the simulation using the 1984 rates. That
is, the total CARB consumption by photolysis to stable products was  181% of the
amount consumed with the 1984 rates,  while the consumption by photolysis to
radical products was 97% of the  amount consumed by this process with the 1984
rates. Therefore, most of the increase using the new rates was due to the increased
NO2 photolytic rates as opposed to the higher CARB rates.

    Figure 16 shows the NO, NO2, and O3  profiles for both CBS and CB4  using
the mixed layer photolytic rates.  CBS was initially more reactive than CB4 when
using the same basic photolytic data. This was  in  part because of the constant
composition CARB in CBS.  Figure 17 shows the  CARB profile from CBS and  the
"calculated CARB" profile from CB4 (i.e., HCHO+ 2 RCHO). CBS maintained a high
level of CARB in the middle of the simulation compared to CB4.  This higher CARB
level occured at the maximum photolysis rate and produced many more radicals
in CBS  than in CB4, thus increasing  its  overall reactivity. The larger amount of
radicals also resulted in a more rapid removal of the NOX causing the 03  profile
to show a sharp decrease in rate at about 400 minutes.  CB4 clearly has a more
realistic treatment of the carbonyl chemistry than CBS.

    Figure 18 shows the NO, NO2, and O3 profiles for CB4 using the  mixed layer
photolytic rates for: a) the Example 1  case, and b) the same conditions with a 40%
HC reduction. The 40% HC reduction (i.e., 1.10 to 0.66 ppmC)  resulted in a 32% O3
reduction (i.e., 0.260 to 0.178 ppm).

    Figure 19 shows the NO, NO2, and 03 profiles for CBS using the new mixed layer
photolytic rates for a) the Example 1 case, and b) the same conditions with a 40%
HC reduction. The 40% HC reduction (i.e., 1.10 to 0.66 ppmC)  resulted in a 25% 03
reduction, that is, less reduction than occurred with CB4. A partial  explanation for

                                     51

-------
Recommended Rates
                                                                    Simulation Results
         0.30
                          Carbon Bond Four. Two Photolytic Rate Sets
  a
  a
  o"
  0
  U
                                                                              600
       Figure 14. Example Simulation Using CB4: Comparison of Surface
          and Mixed Layer Photolytic Rates.  (Example 1  from Guidelines
          for OZIPM31).
                                        52

-------
  Simulation Results
                                                              Recommended Rates
                  Carbon Bond III Two Photolytic Rate Seia
a
a

o
d
                             200                 400

                                 time, mins
600
        Figure 15.  Example Simulation Using CB3: Comparison of Mixed

           Layer and 1984 Photolytic Rates. (Example 2 from Guidelines for

           OZIPM31).
                                      53

-------
Recommended Rates
Simulation Results
        0.26
                             Carbon Bond HI and IV. 64O Meters
  c_
  o"
 3
                                                                              6OO
      Figure 16.  Comparison of CBS and CB4  Using Mixed Layer Pho-
          tolytic  Rates.
                                        54

-------
Simulation Results
                                                                  Recommended Rates
u
E
a
o"
3
      0.080
                              CB3 CARB vs CB4 HCHO-i-RCHO
      0.070 -
      0.060-
      0.050 -
0.040-
      0.030 -
      0.020 -
      0.010-
      O.OOO-
                                ,CARB,CB3
                          HCHO +
                           2*RCHO, CB4
                                2OO                   4OO
                                time, mins after O8OOIIDT
                                                                     6OO
      Figure 17.  Comparison of CARB (CBS) and HCHO + RCHO (CB4) Using
          Mixed Layer Photolvtic Rates.
                                       55

-------
Recommended Rates	Simulation Results

the difference between the two mechanisms can be obtained from Figure 20.  This
figure shows the O3-HC relationship for CB4, CBS with older 1984 rates, and CBS
with the new mixed layer rates. All three profiles show a limiting of O3 production
with increasing HC, however, the CB4 profile shows a slightly "steeper" dependence
of O3 on HC. That is, CB4 made slightly more O3 at higher HC and less O3 at lower
HC than did CBS. In these simulations, it is the relative difference in the CARB vs.
HCHO profiles during the middle of the simulations"that is the primary cause of the
differences between  CBS and CB4 in this case.

    Figure 20  also shows  a  typical control calculation.  The base case O3 is as-
sumed  to be 0.18 ppm.  Each mechanism predicts a different HC level would be
needed  to produce this amount of O3:  0.667 ppmC for CB4; 0.648 ppmC for CBS,
1984 rates; and 0.591 ppmC for CBS, new rates. The slopes of the three curves
are similar  between the 0.12 and 0.18 ppm O3 lines; that is AO3/AHC values are:
0.335 ppm/ppmC for CB4; 0.320 ppm/ppmC for CBS, 1984; and 0.357 ppm/ppmC
for CB3, new rates. Since the  absolute reactivity of the three mechanisms is very
close at the 0.18 ppm O3  level, very similar control  requirements are predicted for
the different mechanisms:
      /O
= 100(a
0.179\
  6687
0.187\
                                          forCB4
                                 (0 187\
                                 ^-—)   for CB3,84
                                 0.649 /

                      28% = 100^    for CB3,new
    Although the three mechanisms gave similar results in this example, the im-
proved treatment of carbonyl chemistry in the Carbon Bond Four mechanism makes
it a superior mechanism compared to Carbon Bond Three and its predictions should
more accurately reflect the effects  of emissions and photolytic rate changes  than
those of Carbon Bond Three.
                                     56

-------
                                                                     References
I
u"

J
       0.30
                             Carbon Bond Four, 64O Meters
                               200
40O
                     6(X)
                                    time, mins
     Figure 18. Example Simulation Using CB4:  40% HC Control Using
        Mixed Layer Photolytic Rates.
                                    57

-------
   Reference*
                         Carbon Bond III, 640m Rates
o.
a

o

o
u
                             200                 400


                                 time, mina
600
         Figure 19. Example Simulation Using CB3: 40% HC Control Using

            1984 Photolytic Rates.
                                       58

-------
                                                                        Refe
                            Examples 1 and 2, Guide-lines Doc.
|
s
o
^
a
o
                                    ._-_< 0.59| (28%)




                                      ]«-A°iLv! 0.667 (27%)
                                                                           1 20
     Figure  20.  Comparison of Ozone-Hydrocarbon Relationships for

         CB3 and C134.
                                      59

-------
References
                                    References
 1    H. Hogo,  G.Z. Whitten,"Guidelines for using  OZIPM3  with  CBM-X or  Optional  Mecha-
      nisms," EPA/600/3-86/004, U.S. Environmental Protection Agency, Research Triangle Park,
      North Carolina, Jan  1986.

 2    Whitten, G.Z., H. Hogo, "User's Manual for Kinetics Model and  Ozone Isopleth Plotting Pack-
      age," EPA-600/8-78-014a, U.S. Environmental Protection Agency, Research Triangle Park,
      NC,  1978.

 3    Gipson, G.L. "User's Manual for OZIPM-2: Ozone  Isopleth Plotting With Optional Mecha-
      nisms/Version 2," Office of Air Quality Planning and Standards, U.S. Environmental Protec-
      tion  Agency, Research Triangle Park, NC,  1984

 4    Iqbal, M., "An Introduction to Solar Radiation," Academic Press, NY, 1983.

 5    Dave, J.V.,"Development of Programs for Computing Characteristics of Ultraviolet  Radia-
      tion," Final Report, NAS5-21680, NASA Report CR-139134, National Aeronautics and Space
      Administration, Goddard Space Fit. Ctr., Greenbelt, MD, 1971.

 6    Peterson,  J.T., "Calculated Actinic Fluxes for Air Pollution Photochemistry Applications,"
      EPA-600/4-76-002, U.S.  Environmental Protection Agency,  Research Triangle Park,  NC,
      1976.

 7    Schippnick, P.F.,  A.E.S. Green, "Analytical Characterization  of Spectral  Actinic  Flux and
      Spectral Irradiance in the Middle Ultraviolet," Photochemistry and Photobiology, 35, pp 89-
      101,  1982.

 8    "Chemical Kinetics and Photochemical  Data for Use in Stratospheric Modeling," Evaluation
      #7,  Jet Propulsion Laboratory Publication 85-37, 1985.

 9    Schere, K.L. and  K.L. Demerjian, "Calculation of Selected Photolytic Rate Constants over a
      Diurnal Range," EPA-600/4-77-015, U.S. Environmental Protection Agency, Research Trian-
      gle Park, NC,  1977.

10    Demerjian, K.L., K.L. Schere, and J.T. Peterson,  "Theoretical Estimates of Actinic Flux and
      Photolytic Rate Constants for Atmospheric Species in the Lower Troposphere,"  Adv. Environ.
      Set.  TechnoL,  29, 1980.

11    Carter, William P.L., F.W. Lurmann, R. Atkinson, and A. Lloyd, "Development and Testing
      of a  Surrogate Species  Chemical Reaction  Mechanism," Final Report, EPA Contract 68-02-
      4104, U.S. Environmental Protection  Agency, Research Triangle Park, North Carolina, 1986

12    Atkinson, R. and  Allen C. Lloyd, "Evaluation of Kinetic and Mechanistic Data for Modeling
      Photochemical Smog," J. Chem.  Ref. Data. 13, No. 2, p315-444, 1984.

13    Killus, J.  and G.Z. Whitten, "Technical Discussions  Relating To the Use of Carbon Bond

                                            60

-------
                                                                                 References
      Mechanism in OZIPM/EKMA," EPA-450/4-84-009, U.S. Environmental Protection Agency,
      Research Triangle Park, NC, 1984.

14   Jeffries, H.E., K.G. Sexton, "Sensitivity of Carbon Bond Mechanism to NMOC aloft." Progress
      Report, Sensitivity of EKMA-type Control Estimates to Model Inputs, EPA Co-operative
      agreement CR812315, 1986

15   Jeffries, H.E. and K.G. Sexton, "UNC Chamber Photolytic Rate Calculation Procedure,"
      Technical Guidance  Report, Department of Environmental Science, UNC, Chapel Hill, NC,
      1986.

16   Zafonte, L., P.L, Rieger, J.R. Holmes, "Nitrogen Dioxide Photolysis in the Los Angeles At-
      mosphere," Envir. Sci. Techno. 11 5, pp 483-487, 1977.

17   Harvey R.B., D.H. Stedman, W. Chameides, "Determination of the Absolute Rate of Solar
      Photolysis of NO2,"  J.A.P.C.A. 27, 7 pp 663-665, 1977

18   Dickerson R.R., D.H. Steadman, A.C. Delany, "Direct Measurements of Ozone and Nitrogen
      Dioxide Photolysis Rates in the Troposphere," J,  Geo. Res. 87, pp 4933-4966, 1982.

19   Parrish, D.D., P.C. Murphy, D.L. Albritton, F.C. Fehsenfeld, "The Measurement of the Pho-
      todissociation Rate of N02 in the Atmosphere," Atmos. Environ. 17 pp 1365-1378, 1983.

20   Madronich S., D.R. Hastie, B.A. Ridley, H.I. Schiff, "Measurements of the Photodissociation
      Coefficient of NO2 in the Atmosphere: I. Method and  Surface Measurements," Journal Aim.
      Chem. 1  pp 3-25, 1983

21   Bass, A.M, A.E. Ledford, A.H. Laufer, "Extinction Coefficients of NO2 and  N204" J.  Res.
      NBS, 80A, No.  2, pp 143-166, 1976.

22   Harkner, A.B., W. Ho, J.J. Ratto, "Photodissociation Quantum Yield  of NO2 in the Region
      375 to 420 nm," Chem. Phy. Lettr. 50, No. 3, pp 394-397, 1977.

23   Davenport, J.E., "Determination of NO2 Photolysis Parameters of Stratospheric Modeling,"
      Report No. FAA-EQ-78-14,  1978.

24   Coulson,  K.L. and D.W. Reynolds, J. Appl. Meteorol.,10, pp 1285, 1971.

25   Doda, D.D., A.E.S. Green, "Surface Reflectance Measurements in the  UV from an  Airborn
      Platform, Part 1," Applied Optics 19 pp 2140- 2145, 1980.

26   Doda, D.D., A.E.S. Green, "Surface Reflectance Measurements in the  UV from an  Airborn
      Platform, Part 2," Applied Optics 20 pp 636-642, 1981.

27   King, M.D., B.M. Herman, "Determination of the Ground Albedo and the Index of Absorption
      of Atmospheric Particnlates by  Remote Sensing. Part  I: Theory.,"  J. Atmos. Sciences 56 pp
      163-173,  1979.

28   King, M.D., "Determination of the Ground Albedo and the Index of  Absorption of Atmo-
      spheric Particulates  by Remote Sensing.  Part II: Application," J. Atmos. Sciences 86 pp

                                           61

-------
References
      1072-1083, 1979.

29    Raschke, E., T.H. Vander Hair, W.R. Bande«n, M. Pasternak, "The Annual Radiation Bal-
      ance of the Earth-Atmosphere System During 1969-70 from Nimbus 3 Measurements," J.
      Atmoa.  Sciences SO, pp 341-364, 1973.

SO    Otterman, J., R.S. Fraser, "Earth-atmosphere System and Surface Reflectivities in Arid Re-
      gions from LANDSAT Data,"  Remote Sena. Environ. 5, pp 247-266, 1976.

SI    Frederick, I.E., R.B. Abrams, "The Surface Albedo of the Earth in the Near Ultraviolet
      (330-340 nm)," Remote Sens.  Environ. 11, pp 337-347, 1981.

82    Logan, J.A, M.J.  Prather, S.C. Wofsy, M.B. McElory, "TVopospheric  Chemistry: A Global
      Perspective," J. Geo. Res. 86  pp 7210-7254, 1981.

S3    Jeffries, H.E., K.G. Sexton, C.N. Salmi, "The Effect of Chemistry and Meteorology on Ocone
      Control Calculations Using Simple Trajectory Models and the EKMA Procedure," EPA-
      450/4-81-034, U.S. Environmental Protection Agency, Research Triangle Park, NC,  1981.
                                           62

-------
                                    TECHNICAL REPORT DATA
                            /Please read Instructions on the re\ ersc before completing
1 REPORT NO.
  EPA-450/4-87-003
                                                            3. RECIPIENT'S ACCESSION NO. .
4. TITLE AND SUBTITLE
  Technical Discussion  Related To The Choice Of Photolyti
    Rates  For Carbon Bond  Mechanisms In OZIPM4/EKMA
                                5. REPORT DATE
                                c    February 1987
                                6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
  H.  E.  Jeffries  and Kenneth G. Sexton
                                                            8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
  Dept.  Of Environmental  Sciences And Engineering
  University Of North Carolina
  Chapel Hill, NC
                                                             10. PROGRAM ELEMENT NO.
                                11. CONTRACT/GRANT NO.
12. SPONSORING AGENCY NAME AND ADDRESS
   Air Management Technology Branch  (MD 14)
   Monitoring And Data Analysis Division
   Office Of Air Quality  Planning And Standards
   U.  S.  Environmental Protection Agency
                                                             13. TYPE OF REPORT AND PERIOD COVERED
                                14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
   EPA Project Officer:   Edwin L. Meyer,  Jr.
16. ABSTRACT
        This report proposes to explain  the  methods, and to  identify the sources  of
   information used to  calculate photolytic  rates, for the chemical models Carbon
   Bond III (CB3) and Carbon Bond IV  (CB4) for use in the OZIPM3 and OZIPM4 oxidant
   modeling programs.   The mixed layer photolytic rates described in Chapter 4  are alsc
   the default rates for  CB4 in the OZIPM4 program, expected to be used in new  State
   Implementation Plan  calculations.
        This report also  reviews previous recommendations for photolytic rates  used in
   the original OZIPP program and in  the revised program OZIPM2.
 7.
                                KEY WORDS AND DOCUMENT ANALYSIS
                  DESCRIPTORS
                                               b.IDENTIFIERS/OPEN ENDED TERMS
                                              c.  COSATl Field/Group
  Photolytic Rates
  Chemical Bond
  Photochemical Model
  OZIPM
  Carbon Bond Model
Oxidant Model
State Implementat
  Plan
ion
18. DISTRIBUTION STATEMENT
                                               19. SECURITY CLASS (This Report)
                                              21. NO. OF PAGES
                                                 70
                                               20 SECURITY CLASS (This page)
                                                                          22. PRICE
EP'A Form 2220-1 (Rev. 4-77)   PREVIOUS EDITION is OBSOLETE

-------