United States Office of Air Quality EPA-450/4-81-030
Environmental Protection Planning and Standards September 1981
Agency Research Triangle Park NC 27711
Air
Questions and Responses
at the 1981 Workshops
on Procedures To
Demonstrate Attainment
of the NAAQS for Ozone
in the 1982 SIPs
-------
EPA-450/4-81-030
Questions and Responses
at the 1981 Workshops on Procedures
To Demonstrate Attainment of the NAAQS
for Ozone in the 1982 SIPs
by
Edwin L Meyer, Gerald L Gipson,
and Robert F. Kelly
U.S. Environmental Protection Agency
Office of Air Quality Planning and Standards
Research Triangle Park, North Carolina 27711
Prepared for
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Air, Noise and Radiation
Office of Air Quality Planning and Standards
Research Triangle Park, North Carolina 27711
September 1981
••'• r.-;.,-.^orn Street
Illinois 60604
-------
This report is issued by the Environmental Protection Agency to report
technical data of interest to a limited number of readers. Copies are
available free of charge to Federal employees, current EPA contractors
and grantees, and nonprofit organizations - in limited quantities -
from the Library Services Office (MD-35), U.S. Environmental Protection
Agency, Research Triangle Park, North Carolina 27711; or, for a fee,
from the National Technical Information Service, 5285 Port Royal Road,
Springfield, Virginia 22161.
This document has been reviewed by the Office of Air Quality Planning
and Standards, U.S. Environmental Protection Agency, and approved for
publication. Subject to clarification, the contents reflect current
Agency thinking.
Publication No. EPA-450/4-81-030
U 5 rnv'^
ii
-------
PREFACE
In April 1981, the Environmental Protection Agency (EPA) held workshops to
describe procedures for demonstrating attainment of the National Ambient Air
Quality Standard (NAAQS) for ozone in the 1982 State Implementation Plans
(SIPs) prepared by State or local agencies. Workshops were held in three
cities: Philadelphia (P), St. Louis (SL) and San Francisco (SF). A large
portion of these workshops was spent explaining and demonstrating the use of the
city-specific EKMA. A number of questions concerning EKMA and its use were
raised by workshop attendees. These have been compiled and responses have been
prepared by members of the staff of the Monitoring and Data Analysis Division,
Office of Air Quality Planning and Standards, U.S. EPA.
The April 1981 workshops were divided into six broad topics dealing with
city-specific EKMA:
I. The Conceptual Basis for EKMA and Relationship of the Modeling
Procedure to the Form of the Ozone Standard;
II. Air Quality and Meteorological Monitoring Support Needed to
Perform Analyses Using City-specific EKMA;
III. Generating Ozone Isopleth Diagrams for Use in City-specific EKMA;
IV. Application of Isopleth Diagrams in the EKMA Procedure;
V. Efforts to Validate EKMA;
VI. General Modeling Questions.
The questions and answers are included under the topic in which they were
asked.
A number of the questions cover points which have not been explicitly
covered in previous, written guidance concerning city-specific EKMA, but yet were
of concern to State and local air pollution, transportation and planning agencies.
Therefore, the purpose of this document is to disseminate the questions and
concerns of attendees raised at the workshops and the Agency's responses thereto.
-------
QUESTIONS AND RESPONSES AT THE 1981 WORKSHOPS ON
PROCEDURES TO DEMONSTRATE ATTAINMENT OF THE NAAQS
FOR OZONE IN THE 1982 SIPs
I. Ozone Standard and Conceptual Basis for EKMA
1. Question: What should a State do if it only has the HC/NO ratio for
one year but has ozone data for several years? (P, SF)
Replv: The median NMOC/NO ratio observed on mo^elpd days during the
year in which the precursor data are available should be used
for other years. In the event fewer than five days are modeled
•during the year with precursor data, use the median ratio for
all days with precursor data. See also pages 42-43 of the
March 1981 Guideline for Use of City-specific EKMA in Preparing
Ozone SIPs.
2. Question: If a 51% reduction is needed at one site, how can one be sure
that this is adequate for all sites? (SL)
Reply: One cannot be certain that a reduction which is sufficient to
meet the standard at one site is sufficient to meet the NAAQS at
all sites. Page 10 of the March 1981 Guideline suggests that
the five days with the highest daily maximum ozone concentration
at each site be modeled. This should provide an adequate safe-
guard that the upper portion of the control requirement distri-
bution has been defined for each site. Selection of the highest
site-specific control requirement is specified in the Guideline.
3. Question: Is the data for calendar years or ozone seasons? (SL)
Reply: Data are for ozone seasons. See page 12 of the March 1981
Guideline.
4. Question: Original guidance suggested calculating 2nd highest value when
two years of data are available, why has this changed? (SL, SF)
Reply: The change was made because of the way in which the ozone NAAQS
is worded and because the data base is discrete rather than
continuous. The wording implies that the standard is met if, on
average, no more than one daily maximum ozone concentration per
year exceeds 0.12 ppm. If, after controls are implemented,
Q.12 ppra is exceeded on only two days over a two year period,
the NAAQS would be met. Hence, in calculating the control
requirement, the controls must be sufficient to insure daily
maximum ozone <_ 0.12 ppm on all days except two over a two year
period. With, a discrete data base, this will occur if the third
b-ighest control calculation is selected at a site.
-------
5. Question: If an area which has monitored for a long period of time changes
its monitoring method, is it necessary to go back and change any
values from earlier studies? (SF)
Reply: No.
6. Question: Sunlight simulation is a function of latitude and longitude.
Is a provision made for altitude? Isn't altitude considered by
entering atmospheric values? (SF)
Reply: No provision is made for altitude in the OZIPP computer program.
Altitude could only be considered by modifying the photolytic
rate constants. This would be difficult, because not all rate
constants would be affected similarly by the change in altitude.
7. Question: Does EKMA work well considering definite diurnal cyclic patterns?
(SF)
Reply: The validation tests in which predictions obtained with OZIPP
were compared with air quality observations and the EKMA pro-
cedure has been compared with changes in ozone predicted with
more sophisticated models or observed with trends have all
assumed diurnal cyclical meteorological patterns. For example,
sunlight intensity and mixing height were varied diurnally in
accordance with our best estimates.
Monitoring
8. Question: Is the "NMOC" monitorinq method proposed by EMSL fa_st_enouc[h_ to
be used In the field? CSF)
Reply: Yes. The method proposed by EMSL is described at length in a
report, Determination of Nonmetbane Organic Carbon (NMOC) by
Cryogenic Preconcentration and Flame lonization Detection
(Frank. McElroy, U.S. Environmental Protection Agency, MD-17,
Research Triangle Park, North Carolina 27711, project officer).
An air sample is taken directly from the ambient air at the
monitoring site or from a bag sample obtained previously. A
fixed-volume portion of the sample is drawn through a tubular
tra|D cold enough, to "freeze" all organic compounds except methane
(which, passes through the trap). Helium is then flushed through
the trap to remove any residual methane. The temperature of the
trap Is raised to 70-80 C so that the collected organic compounds
are reyolatilized and flushed into a conventional FID analyzer
modified to operate with a helium carrier qas. The area under
the resulting response peaks is then translated to concentration
units uslnq a calibration curve.
-------
9. Question:
Reply:
10. Question:
Reply:
11. Question:
Reply:
12. Question:
Reply:
The cryogenic trap simultaneously concentrates the nonmethane
organic compounds and separates and removes methane. Thus, the
technique is both direct reading for NMOC and more sensitive
than conventional NMOC analyzers. The method appears to be
suitable for both upwind measurements as well as measurements
taken in downtown areas.
Assuming a State has 30 days worth of data for HC and NO and
the five highest ozone days don't correspond to the 30 days,
what should be done? (SF)
Use the median NMOC and NO concentrations from the days with
data as a basis for computing post-8 a.m. emissions. To deter-
mine the NMOC/NO ratio for use in EKMA, calculate the 6-9 a.m.
ratio for each dSy with NMOC and NO data. Use the median
NMOC/NO ratio for all days with the necessary precursor data.
X ~--
If a State doesn't have any NMOC data at all, what should be
done? Can they use default values (9.5:1) previously recom-
mended in the city-specific isopleths? (SF)
Every effort should be made to obtain valid NMOC and NO data.
This is the most credible way of deriving NMOC/NO ratios and
considering the impact of "post-8 a.m." emissions. If it is
impossible for a State to obtain these data in a timely fashion,
alternatives need to be discussed on a case by case basis with
the appropriate EPA Regional Office. Examples of alternatives
which have been discussed thus far include use of current ambient
ratios from similar, nearby cities; comparison of emission
inventory information of another city with ambient data to
derive a relation between emissions and ambient ratios which can
be used in the city under review; and use of the 9.5:1 ratio
(which was based on nationwide data from the mid-to-late '70's).
What if a State has two different ratios in the urban area? (SF)
Procedures described on pages 42-43 of the March 1981 Guideline
should be followed.
In most cases precursor measurements are made near major road-
ways; isn't this inconsistent with saying don't use ratios from
industralized areas that would not give areawide representation?
(SF)
No. Roadways, unlike industrial areas (in most cities), are
ubiquitous. Thus, even if a relatively large fraction of the
NMOC and NO observed at a monitoring site is due to a nearby
roadway, the monitor measures concentrations and relative
amounts of NMOC and NO which are likely to be representative of
a larger area. x
-------
13. Question: What are the criteria for NMOC monitoring upwind of the city?
(P)
Reply: Criteria for NMOC monitoring upwind of a city are as follow:
(a) To estimate surface NMOC
(1) Sample >_ 40 km upwind of center city at least
200 m from any local source. Measurements should be repre-
sentative of a large area (e.g., 20 km x 20 km or more).
(2) Sample 6-9 a.m. concentration using bag or
cannister. If a bag is used, shelter from sunlight and
.extraneous organic sources-prior to analysis.
(3) Use sum of species or cryogenic preconcentration
technique to analyze contents of bag or cannister. If a bag has
been used, analyze within 24 hours of collection. If a cannister
is used, analyze within 1 week of collection.
(b) To estimate NMOC aloft
(1) Collect cannister samples upwind aloft via aircraft
or some other device during early morning (6-9 a.m.) hours.
(2) Analyze contents of cannister within one week
using sum of species or cryogenic preconcentration techniques.
14. Question: What is the relative sensitivity of the monitoring for ozone?
(P)
Reply: Within a range of ozone concentrations of 0-0.50 ppm, the
accuracy (.i.e. reproducibility) of a well maintained chemilumi-
nescent analyzer is within +_ .003 ppm.
15. Question: What effect does changing the mixing height have? (P)
Reply: The sensitivity of control requirements to any particular
city-specific EKMA input variable cannot be reported categori-
cally. In general, control requirements tend to be relatively
insensitive to changes in mixing height which are typically
found throughout the U.S. However, this is not to say that
specifying the mixing heights is unimportant. Control require-
ments could be sensitive to mixing height changes under some
atypical set of conditions. Also, the change in mixing height
also affects the importance of other factors, such as pollutants
transported aloft and post 0800 emissions. Thus, the most
practicable methodologies should be used to obtain best estimates.
-------
16. Question: Clarify what is meant by changing mixing height during
stagnation? (P)
Reply: During stagnation periods, the change of mixing height from
morning to the afternoon maximum is similar to the normal pattern
of mixing height change. On stagnation days, a strong inversion
aloft, due to sinking air in the center of high pressure systems
often causes lower than normal afternoon maximum mixing heights.
The mixing heights on any day (including stagnation days) can be
found by following the procedures in Appendix A of the EKMA
Guideline document.
17. Question: What do you do for upwind ozone surface measurements when cities
merge or the city extends up to 40 km from CBD? (SL)
Reply: The purpose of making upwind measurements is to gain an indication
of what is being transported into a particular city on a large
scale. Hence, concentrations which are representative of a large
(e.g., ^ 20 km x 20 km) area are necessary. If cities merge,
surface-based, morning upwind ozone measurements are still
appropriate for use, provided there are no local sources of NO
(e.g., within 200 ml which depress measured ozone levels. If it
is not possible to obtain upwind surface measurements representa-
tive of a large area, direct measurements of ozone aloft (e.g.,
by aircraft) are necessary.
18. Question: What is the relationship between helicopter measurements and
surface measurements obtained during the RAPS study in St. Louis?
CSL)
Reply: St. Louis RAPS helicopter temperature soundings, when compared
with nearby radiosonde data, show similar structures (e.g.,
inversions), although the measured temperatures may differ.
The ozone measurements taken above the surface-based mixed layer
show different amounts of ozone than surface (in the mixed
layer) ozone measurements. During morning hours, higher ozone
values are observed above the mixed layer. In the mixed layer,
some differences in ozone are seen due to the effect of wind
shear (also see Question #76). The ozone above the mixed layer
Is entrained into the mixed layer as the mixing height rises. A
method that uses surface monitors to estimate the effect of ozone
from aloft is given on pages 20, and 22-24 of the EKMA Guideline
document, and in more detail by Chan, A!lard and Tombach (1979).
This method gives reasonable agreement with helicopter measure-
ments of ozone aloft.
-------
19. Question: Does the NMOC cryogenic technique utilize GC? (SL)
Reply: No. The technique utilizes a preconcentration procedure and
then modified conventional FID devices. See the response to
Question 8 and Determination of Nonmethane Organic Carbon (NMQC)
by Cryogenic Preconcentration and Flame lonization Detection.
20. Question:
Reply:
21. Question:
Reply:
22. Question:
Reply:
23. Question:
Reply:
Why is there a stipulation that monitors must not be located
within 200 m of a major source when one is looking for the
highest value of precursors? (SL)
It takes several hours for peak ozone concentrations to be
formed as the result of chemical reactions among precursors.
-During this time, atmospheric dispersion also takes place.
Hence, one is looking for high precursor concentrations which
are representative of those within the city. We feel that
locating a site within 200 m of a large individual source will
not provide representative values unless there are many such
sources throughout the city. See also the response to Question 12.
Why does one look for the highest ozone and not the highest
precursors? (SL)
The highest ozone is of interest to determine whether the NAAQS
is currently being met and, in the event the standard is not
being met, to provide assurance that proposed controls are
sufficient to meet the NAAQS. The purposes for measuring ambient
precursors are to find relatively high concentrations which are
representative of the city so that they may be used to estimate
prevailing NMOC/NO ratios and as a basis of considering the
impact of post-8 a.m. emissions.
In estimating mixing heights, does one interpolate between two
NWS temperature soundings sites or is only one selected? (SL)
The nearest, most representative radiosonde sounding should be
selected as explained in Sections A.1.1 and A.1.2 of the EKMA
Guideline document. Interpolation is not recommended because
the rate of change of temperature between the two soundings may
not be linear, due to fronts or the effects of terrain on
temperature aloft.
What does one do if some of the required data are missing? (SL)
Missing data are of concern for two reasons: they are needed to
generate ozone isopleths with the OZIPP computer program, or
they are needed to estimate control requirements from the isopleth
diagramCs) using the EKMA procedure. There are a number of
built-in default procedures in the OZIPP computer program.
These procedures are utilized in the event the user does not
-------
24. Question:
Reply:
25. Question:
Reply:
26. Question:
Reply:
27. Question:
Reply:
28. Question:
Reply:
enter certain pieces of data. In other cases, the March 1981
Guideline document suggests procedures which can be followed in
the event not all the information desired is available. One
should familiarize oneself with the Guideline to learn what is
possible if certain pieces of data are unavailable. In the event
an appropriate course of action is still unclear, contact should
be made with the appropriate EPA Regional Office and dealt with
on an individual basis.
Can one equate NMOC with NMHC? (SL)
No. The distinction between NMOC and NMHC is that the latter
consists of hydrocarbon (i,e., organic molecules containing only
carbon and hydrogen atoms), whereas the former contains other
organic compounds as well. Since the smog chamber experiment
against which the chemical mechanism in EKMA is calibrated
included organic compounds other than hydrocarbons and the
atmosphere does also, NMOC is the proper terminology.
Since emission inventories include VOC which may exclude some
emissions that may be measured by a monitor, does this present a
problem? (SL)
Several nonreactive compounds are excluded from the emissions
inventory (e.g., methane, ethane and several halogenated com-
pounds). In contrast, these compounds (with the exception of
methane) may be present in ambient NMOC samples. However, the
concentrations of these nonreactive NMOC are a very small fraction
of the total NMOC measured in an urban area. Thus, any resulting
discrepancy between measured compounds and those included in the
inventory is likely to be very small.
Where does one site two wind monitors? (SL)
Table 1-1 of the EKMA Guideline document suggests that one site
be placed in the high precursor area. The other site should be
placed in an open area (but near the city) where the wind meas-
urements taken would be most representative of the general wind
flow through the urban area.
Where is the list of NWS stations applicable to specific
cities? (SL)
A list of NWS stations applicable to specific cities appears in
Table A-l on page A-3 of the March 1981 Guideline.
Are there major changes between the draft and the final guide-
lines for city-specific EKMA? (SL)
There are enough changes between the draft and final Guideline to
warrant obtaining a copy of the final version. Limited quantities
of the final Guideline may be obtained by writing:
-------
U. S. EPA Library
Mail Drop 35
Research Triangle Park, NC
27711
and requesting a copy of EPA-45Q/4-80-027, dated March 1981. The
document is also available from the National Technical Informa-
tion Service, Springfield, Virginia, as document PB81-118739 at
a cost of $11.00.
III. Generating Ozone Isopleths for Use in EKMA
29. Question: How is temperature considered in the model? (P)
Reply: Surface temperature is used to estimate mixing heights as described
in Table A-3 on page A-7 of the 1981 Guideline. The chemical
kinetics mechanism in OZIPP uses certain rate constants which are
most appropriate for temperatures of 80-90F.
30. Question: How can one get a better resolution for post 0800 emissions? (P)
Reply: It is possible to utilize a gridded, temporally-resolved inventory
to compute post-8 a.m. emissions more precisely. However, for
the reasons described on pages 31-32 of the March 1981 Guideline,
resolution finer than 10 km x 10 km grid squares should not be
used. It should be pointed out that fine spatial and temporal
detail in the inventory is not particularly meaningful unless the
wind field is accurately characterized. This is very difficult
to do even with a data base commensurate with the use of Level I
models. A depiction of the wind field which is accurate enough
for use with finely resolved inventories is not likely with a
Level III data base. Therefore, countywide inventories are
recommended for use with city-specific EKMA.
31. Question: Must one calculate new post 0800 emissions everytime? (P)
Reply: As described on pages 47-48 of the March 1981 Guideline, when one
applies the EKMA procedure to an isopleth diagram it is assumed
in drawing the diagram that both initial conditions and post-
8 a.m. emissions are reduced by the same proportion. Hence, for
a given site, it is not necessary to calculate new post-8 a.m.
emissions unless one is interested in seeing what happens if the
reduction in post-8 a.m. emissions is nojt proportional to the
reduction in initial concentration.
If one is concerned with modeling several different sites, the
decision of whether or not to recalculate post 8 a.m. emissions
becomes more subjective. Generally, if the new trajectory passes
over the same counties at about the same time, it should not be
necessary to recalculate post 8 a.m. emissions.
-------
32. Question: Where do you start the column? (P)
Reply: Using a Level III or city-specific EKMA analysis, the column is
assumed to start in center city near where the NMOC/NOX ratio is
measured.
33. Question: Assuming the parcel starts at center city, would one not recom-
mend use of data later in the morning? (P)
Reply: No. The 6-9 a.m. NMOC/NO ratio is used as an indicator of the
relative amount of NMOC and NO precursors available in the city
to form ozone. The measurements are more reliable than later
ones for two reasons. First, concentrations are higher and so
the measurements are more accurate. Second, by mid-morning some
of the photochemistry has proceeded. The result is that NO dis-
appears more rapidly than NMOC, yielding higher ratios. Hence,
it would not be appropriate to use such a ratio in the EKMA
procedure.
34. Question: If one has HC/NO ratios later in the morning, can one use
them? (P) x
Reply: No. See the response to Question 33.
35. Question: Can one use an emission factor of zero? (P)
Reply: Ordinarily no. However, there are exceptions. For example, a
trajectory which passed over a large body of water would have
zero post-8 a.m. emissions.
36. Question: What are the advantages of the OZIPP program? (P)
Reply: OZIPP is relatively easy to run, is well-documented, allows one
to consider local factors in assessing a city's ozone problem,
allows consideration of both NMOC and NO in strategy develop-
ment, provides isopleth diagrams which are very convenient for
assessing the impact of a variety of strategies, and depicts
relationships which are consistent with experimental data.
37. Question-: To get the program on a tape, can you specify parameters needed
for a specific system? (SL)
Reply: The OZIPP program is normally distributed on magnetic tape with
characteristics which are fairly common to most computer systems.
Upon special request, a specific tape characteristic will be
processed, contingent on the capabilities of EPA's UNIVAC computer.
38. Question: Is the light intensity assumed in OZIPP based on clear sky
values?
Reply: Yes.
-------
39. Question: Large changes In predicted max - 1 hour concentration with.
varying light intensity seem inconsistent with small control
requirement changes. (_SL)
Reply: The workbook cites a study in which peak ozone levels were found
to vary from 4 to 23% between two different light intensities,
whereas control requirements varied by only 1 to 2%. First, it
should be noted that peak ozone levels varied only between 4 and
7% for NMHC/NO ratios of 5:1 to 20:1 (the range most common in
urban areas). Thus, the apparent large difference in sensitivity
between peak ozone and control requirements is not that great
under usual conditions. Furthermore, peak ozone levels were
found to decrease proportionately with a decrease in light inten-
sity. Therefore, a set of isopleths generated with lower light
intensity should have the same relative shape and spacing as the
isopleths generated with the higher light intensity. As a con-
sequence, control requirements, which are a function of relative
distance between two isopleths, should be nearly the same.
Looking at this another way, light intensity affects the peak
level under existing conditions and the case of reduced emissions
corresponding to the standard. If changing the light intensity
affects both of these levels by roughly the same degree, then the
relative change between existing levels and the standard should
be very nearly the same for different light intensities, as borne
out by the results of the sensitivity analysis.
40. Question: Has work been done to input intermediate mixing heights (other
than early a.m. and max)? (SL)
Reply: Use of "intermediate" mixing heights has been evaluated by means
of a city-specific EKMA sensitivity test. Results of these tests
have shown that intermediate mixing heights are not important
when the relative change between the early morning mixing height
and the maximum afternoon level is small. The intermediate
levels (.i.e., the actual diurnal pattern) become more important
for large differences. Usually, however, insufficient data exist
to specify adequately the diurnal variation in mixing height,
especially during the breakup of the nocturnal inversion that
normally takes place in mid-morning. A characteristic curve
algorithm for depicting the diurnal variation in mixing height is
currently incorporated in Version II of OZIPP. This algorithm is
based on observation of intermediate levels, as well as theo-
retical considerations, and provides a more realistic representation
of the dilution process than was originally incorporated in
OZIPP. Unless sufficient data exist to specify more precisely
the diurnal pattern, use of the characteristic curve algorithm is
recommended.
10
-------
41. Question:
Reply:
42. Question:
Reply:
43. Question:
Reply:
44. Question:
Reply:
With new nationwide VOC controls, will transport levels decrease?
tSL)
Yes. Pages 52-60 and Appendix B in the March 1982 Guideline
provide specific suggestions on how to consider this in city-
specific EKMA.
Is the NMOC transport aloft example in the workbook based on
the new mixing height curve? (SL)
The sensitivity tests for precursor transport reported in the
Workbook and Appendix B of EPA-450/4-80-027, Guideline for
Use of City-specific EKMA in Preparing Ozone SIPs, were conducted
.prior to tne incorporation of the characteristic curve in OZIPP,
and thus the exponential algorithm originally in OZIPP was used.
However, spot checks of the calculations were performed using the
characteristic curve, and little, if any, difference was found.
Again, these test results provide a general guideline for deline-
ating regions in which transport levels may become important.
They are not intended to be used for specific estimates of the
effects of precursor transport for a given city.
Are the NMOC concentrations in the surface and aloft transport
additive for 03? (SL)
In general, the effects of precursor transport aloft and in the
surface layer are not additive. Apparently, this results from
the nonlinearities associated with the ozone formation process.
Has some type of analysis been done on ozone transport? (SL)
Yes. Over the past several years, efforts have been undertaken
to (a) document the extent of transported ozone, and (b) note the
impact of transported ozone levels on calculated control require-
ments. There are many references documenting the extent of
transported ozone. The most recent edition of the Criteria
Document for Ozone CEPA-450/8-78-004), as well as
EPA-6QQ/3-77-117 provide good starting points for those interested
in learning about the extent of transport. EPA-450/2-77-021b
provides one of the few published references on the sensitivity
of EKMA to transport. Much, as yet unpublished, work on the
sensitivity of EKMA results to transport has been done more
recently. Some results are printed on pages 105, 107 and 133-138
of the Workbook which was distributed at the recent Workshops on
Procedures to Demonstrate Attainment of the NAAQS for Ozone in
the 1982 SIP's.
11
-------
45. Question: Are direct measurements of precursors NMHC or NMOC, reactive or
nonreactive? (SL)
Reply: There are very few volatile organic compounds classified as "non-
reactive." With the exception of methane, these are generally
present in small concentrations in the urban atmosphere. Hence,
reported NMOC levels consist almost entirely of "reactive"
compounds.
46. Question: Explain distinction between control points and control require-
ments? (SL)
Reply: Control requirements refer to the percent reduction in VOC
emissions necessary to achieve the ozone standard. In sensi-
tivity analyses, the importance of a particular variable is
evaluated by its effect on the VOC control requirements, which
necessitates speaking in terms of differences in control require-
ments. For example, the difference between 50% control and 45%
control is 5%. Note that this does not refer to a percent dif-
ference. In one Workshop, the absolute difference (i.e., 5%) was
referred to as control points, in order to distinguish between an
absolute difference and a relative, or percent difference.
47. Question: Why simulate initial concentrations when one has measured data?
ISL}
Reply:
48. Question:
Reply:
49. Question:
Reply:
50. Question:
Reply:
51. Question:
Reply:
A variety of initial conditions are simulated in order to generate
an isopleth diagram. Such a diagram provides a convenient short
hand way of depicting changes in peak ozone levels which might
accompany a variety of changes in NMOC and/or NO .
A
Can one assume uniform emissions during the day? (SL)
This is an acceptable procedure if there is no good basis for
making better estimates.
Does one consider major sources along the trajectory line? (SL)
Yes, unless there is reason to believe that the effective plume
height is greater than the afternoon mixing height.
Can one input a city-specific propylene/butane fraction? (SL)
No. The equivalent propylene/butane split has not been estab-
lished for any atmospheric mix other than the automotive mix used
in the original Bureau of Mines Smog Chamber Studies. Thus,
there is no basis for making other assumptions.
Can one monitor in an area heavily impacted by traffic to adjust
ratio? (SL)
No.
12
-------
52. Question:
Reply:
53. Question:
Reply:
54. Question:
Reply:
For N02/N0 ratio, would one average results for all sites? (SL)
/\
Yes.
How will fractions change when emissions from stationary sources
are reduced? (SL)
Ordinarily, the emission fractions would not change. The default
assumption in the EKMA isopleth diagram is that both initial
concentrations and post-8 a.m. emissions are reduced in the same
proportion.
Should data collected in an area dominated by industrial
sources be used? (SL)
If the precursor data are dominated by individual sources or if
there is reason to believe that the ambient chemical composition
of the air is radically different from that of the urban area,
these data should not be used in Level III EKMA analysis. This
is consistent with the propylene/butane split derived from the
Bureau of Mines Smog Chamber data.
55. Question:
Can the model be adjusted to give N02 levels?
predictions? (SL)
How good are the
Reply: A modified version of the OZIPP program is needed to generate
peak hourly N02 isopleths. The model's N02 predictions agree
well with smog chamber data. However, there has not been an
extensive effort underway to evaluate the performance of the
model in predicting N02 using other criteria.
56. Question: What is the basis for the statement that N02/NOV is not important?
(SL) X
Reply: Neither predicted peak ozone concentrations nor estimated control
requirements appear sensitive to the initial N02/N0 ratio.
A
57. Question: Does free format only apply to PLACE?
Reply: A free format card foilows the PLACE option card, and one also
follows the TITLE option card. Several of the other option cards
may require subsequent fixed format cards in some cases. Input
formats for the OZIPP option cards are summarized on pages 57-67
of EPA-600/8-78-014a, User's Manual for Kinetics Model and
Ozone Isopleth Plotting Package.
58. Question: Why use CALCOMP plotter? (SL)
Reply: The CALCOMP plotter can be used to generate an isopleth diagram.
A diagram generated in this fashion is neater, more precise and
usually easier to work with than the standard line printer gen-
erated diagram.
13
-------
59. Question: If there are two options for control needed, does one use PLOT
or CALCULATE option? (SL)
Reply: The CALCULATE option is used to perform a single simulation -
i.e., make a prediction of peak ozone from a given set of NMOC
and NO initial concentrations. The ISOPLETH option is used to
generate an isopleth diagram, i.e., produce a diagram showing
peak ozone concentrations as a function of variable initial NMOC
and NO concentrations. The isopleth diagrams are used to calcu-
late control requirements. The PLOT option is used with the
ISOPLETH option to generate a plotter version of the diagram.
The PLOT option does not apply to the CALCULATE option.
60. Question: What is UNIVAC capacity?
model? (SL)
What size computer is needed for the
Reply: The OZIPP program requires about 50k word storage on the UNIVAC
computer. This corresponds to approximately 200k bytes of
storage.
61. Question: Can data only be input on cards? (SL)
Reply: The OZIPP program is a batch oriented program, reading input data
from cards. On some computer systems, input data can be stored
in files in card-image format and input to the program. That
capability is completely system dependent, and must be determined
on a case by case basis.
62. Question: Are there procedures to check-out program on a State's computer?
(SL)
Reply: An Addendum to the OZIPP User's Manual explains Version II of
OZIPP, and contains two example OZIPP runs. These examples may
be used as benchmarks to insure proper operation on individual
computing systems.
63. Question: Can one compare results against standard EKMA curves? (SL)
Reply: One can make such comparisons, but city-specific EKMA is likely
to give somewhat different answers. Because of the conservative
assumptions in standard EKMA, city-specific EKMA will likely show
lower control requirements in most, but not necessarily all,
situations.
64. Question: Is the 250 m height just an example or is this the value to be
used as a minimum? (SL)
Reply: 250 meters is to be used for the mixing height at 0800 LCT
if 1) the mixing height estimated by the method in Appendix A of
the Guideline document is less than 250 meters, or 2) if no
morning mixing height value is available.
14
-------
65. Question:
Reply:
66. Question:
Reply:
67. Question:
Reply:
68. Question:
Reply:
69. Question:
Reply:
Is 25Q ra a default value for OZIPP? (SL).
No. If one fails to specify a morning mixing height, the OZIPP
program assumes a value of 510 m. This is based on the climato-
logical mean value for Los Angeles. Note, however, that 250 m is
the recommended minimum morning mixing height.
Does one also use 250 m when calculated value is < 250 m? (SL)
Yes. This is generally a good idea when utilizing Level III
EKMA (with the trajectory beginning in the urban area). It is
based on observations of urban vertical profiles which suggest
significant levels of precursors can exist above the calculated
morning mixed layer in urban areas. Hence, if one used a morning
mixing height substantially less than 250 m, one would overesti-
mate the amount of dilution of initial precursor concentrations
occurring later in the day.
Is there a maximum height for the program? (SL)
The computer program used to estimate mixing heights (published
as EPA-45Q/4-81-022) uses a value of twice the normal daily
summer maximum mixing height to warn the user of a possible high
value for mixing heights. The value is entered by the user and
can be found in Table A-l of the EKMA Guideline document or in
AP-101. This feature allows the user to check the high mixing
height for accuracy before entering it into EKMA/OZIPP.
Is diagram presented in the workbook typical for cities other
than Philadelphia? (SL)
The diagram of pollutant and temperature vertical profiles
for Philadelphia is similar to profiles measured by helicopter on
other days in Philadelphia and during the St. Louis RAPS. The
inversion above the urban mixed layer (see Questions 74 and 91)
will trap most pollutants emitted near the surface. This causes
a layer of pollutants to be trapped below the cleaner air above
the inversion. The ozone concentration below the inversion is
less than the concentration above due to scavenging of the ozone
by nitric oxide emissions.
In coastal situations, if one has two measurements of mixing
height, one on the coast and one further inland, which one should
be used in OZIPP? (SF)
The mixing height that is most representative of the urban area
(where the EKMA/OZIPP trajectory is assumed to start) should be
used for the 0800 LCT mixing height.
15
-------
70. Question:
Reply:
71. Question:
Reply:
72. Question:
Reply:
73. Question;
Reply:
The choice for maximum mixing height depends on where the down-
wind ozone peak occurs. If the site observing peak ozone is in a
coastal situation (.e.g., a sea breeze) at the time of maximum
ozone, the coastal mixing height should be used. Otherwise (even
if the EKMA/OZIPP trajectory passes over water), the inland
mixing height should be used for the maximum mixing height.
In the draft guideline, the minimum morning mixing height of
150 m was recommended instead of 250 m as now presented, which is
correct? (SF)
250 m. See the response to Questions 64 and 66.
-What are the criteria for not using mixing height program for
0800? (SF)
If a radiosonde site that is meteorologically representative of
the urban area cannot be found, due to a front between the city
and nearby radiosonde stations, the method for finding the
0800 LCT mixing height, as explained in the Guideline document
should not be used and a value of 250 meters should be used for
EKMA/OZIPP. (.See pages A-l through A-4 in the Guideline
document.)
Is this method strictly for urban use? (SF)
The method for estimating mixing heights as described in
Appendix A of the EKMA Guideline document and the mixing height
computer program (EPA-450/4-81-022) can be used at rural as well
as urban sites. The only information needed at the surface is
the measured air temperature and an estimate or measurement of
the surface air pressure (as explained in Appendix A). A problem
occurs when a surface-based stable layer is found by this method.
The 250 meter minimum value for an urban mixed layer at 0800 LCT
is too deep for rural use. A rural "mixed" layer of 100 meters
has been used by Schere and Demerjian (A Photochemical Box Model
for Urban Air Quality Simulation. Proceedings, 4th Joint Con-
ference on Sensing of Environmental Pollutants, published by
American Chemical Society 1978) as an estimate of the vertical
dispersion of surface-emitted pollutants due to wind turbulence
and diffusion.
How many areas were analyzed in regard to urban temperature
profiles? (SF)
The work of Godowitch, et. al, (Dissipation of the Nocturnal
Inversion Layer at an Urban and Rural Site in St. Louis, Missouri.
Preprints, 4th Symposium on Turbulence, Diffusion and Air Pol-
lution, January 15-18, 1976, Reno, NV. Published by American
Meteorological Society) using St. Louis RAPS helicopter data was
used along with an analysis of six helicopter temperature and
16
-------
pollutant soundings taken in Philadelphia on clear mornings
around 0800 LOT. The limited data from Philadelphia show an
urban mixed layer similar to the layer found by Godowitch,
et. al, in St. Louis. Since the urban mixed layer is generally
not detected by the recommended objective method described in
Appendix A of the EKMA Guideline document or by sodar, we suggest
using a 250 meter urban mixing height as a minimum value at 0800
LCT in EKMA/OZIPP. If low level sounding data (slow ascent
radiosondes, helicopter soundings, etc.) are available in an
urban center, we can work with a local or regional group to
estimate the height of the urban mixed layer for a given day.
74. Question: Is it typical that urban areas will generally be unstable in the
lowest 100 m? (SF)
Reply: A typical urban temperature sounding will have a less stable
layer under the urban inversion layer due to heating from
buildings and other urban surfaces and mechanical mixing due to
greater surface roughness in the urban area. According to
Godowitch, et. a!. (1979), the urban mixed layer has an average
minimum height of 160 meters.
75. Question: What does one have to do if there is no upwind surface measure-
ments or direct measurements aloft? What about using 2/3 value
of carryover from previous day? (SF)
Reply: The need for upwind measurements has been identified for several
years. A city for which there are np_ upwind data should there-
fore constitute a special case which will need to be addressed on
an individual basis with the appropriate EPA Regional Office. It
is not possible to say whether or not the procedure proposed in
this question is appropriate without being familiar with the
particulars of the individual case.
76. Question: How constant are the ozone values aloft? Both spatial or time
variation? Would it be more representative to use average ozone
aloft instead of a value on a specific day? (SF)
Reply: Values of ozone aloft (above the mixed layer) vary slightly
in time and space. The variation depends on the history of the
air at various altitudes and locations. These differences are
caused by changes in wind speed and direction with height and
time. However, the average ozone above the mixed layer (i.e.,
not due to local precursors) on a given day can have a signi-
ficant effect on the control requirement for that day (see
Section 3.1.3 of the EKMA Guideline document) in some cases.
Thus, using the average value for a set of days could result in
violations of the ozone standard after the application of controls
if the actual concentration of ozone aloft on the controlling day
is greater than the average concentration of ozone aloft.
17
-------
77. Question:
Reply:
78. Question:
Reply:
79. Question:
Reply:
80. Question:
Reply:
Concerning sensitivity to ozone, if one says that ozone surface
transport is not important and then one says OZIPP is sensitive
to dilution and emissions, aren't these inconsistent? (SF)
No. The impact of ozone transported in the surface layer is to
immediately convert initial NO to N02. However, this occurs in
any event as the result of photochemistry. Sensitivity tests
conducted thus far suggest that realistic concentrations of ozone
transported in the surface layer in the early morning hours have
little effect on peak ozone levels estimated with the model or on
calculated control requirements.
In severe control cases, would one expect ozone surface transport
to be important? (SF)
After imposition of severe controls, boundary conditions become
more important determinants of peak ozone. Therefore, it is
conceivable that ozone surface transport could be more important
after most of the city's emissions have been eliminated.
Does one assume ozone aloft will be affected by control strategies?
(SF)
Yes. A procedure for doing this is described on pages 52-57
of the March 1981 Guideline.
EPA has used 1100-1300 estimate for surface ozone but not for
precursors, what is the rationale for doing this? (SF)
We suggest that 1100-1300 Local Daylight Time surface ozone
readings taken upwind of the city be used as an indicator of
ozone transported aloft. It is also suggested that 6-9 a.m. LOT
readings be used to estimate ozone and precursors transported in
the surface layer. The difference in recommended times is a
result of the presence of the nocturnal surface-based temperature
inversion which causes stable stratification of the atmosphere.
Generally, the atmosphere does not become well mixed until some-
time after 9 a.m. Hence, 6-9 a.m. surface measurements would not
necessarily be accurate indicators of ozone transported aloft.
The 11QQ-1300 LDT suggestion is based on the supposition that
th.i.s is shortly after the breakup of the nocturnal surface based
inversion, but before a large amount of ozone is formed as the
result of photochemistry. The 1100-1300 LDT recommendation is
not a hard and fast rule. If an area has more locally specific
information on when the nocturnal surface-based inversion dis-
appears, this information may be used to select some other time
for using surface data to estimate ozone transported aloft.
18
-------
81. Question: If one could eliminate local contribution and transport, could
one use surface concentration to estimate ozone aloft? (SF)
Reply:
82. Question:
Reply:
83. Question:
Reply:
84. Question:
Reply:
As discussed in the response to Question 80, surface measure-
ments can be used to estimate ozone transported aloft. However,
the ground rules described in the previous response would still
be necessary. Prior to breakup of the nocturnal inversion,
ozone in the surface layer would be subject to surface deposition,
whereas ozone transported aloft would not. Hence, use of early
morning surface measurements would probably underestimate ozone
concentrations transported aloft overnight.
If one assumes a straight line trajectory with uniform speed,
'is this what actually happens in the real world? (SF)
Probably not. However, it is. exceedingly difficult to precisely
define individual trajectories, even with a data base suitable
for use with Level I models. Indeed, the usual procedure with
Level I models is to mathematically smooth trajectories. The
trajectory used in city-specific EKMA assumes that highest
observed ozone levels are, to an important degree, impacted upon
by areas with the highest precursor concentrations. Such wind
data as do exist should be used fo establish that a monitoring
site observing high ozone is downwind from the city on the days
being modeled. Whether the exact trajectory assumed in the
modeling exercise is followed is not so important unless the
emission inventory is spatially and/or temporally resolved to
great detail. If county wide emissions are used (as recommended
on pages 28-31, in the March 1981 Guideline), failure to describe
a trajectory accurately is mitigated.\
Is emission fraction always based on initial height of column?
(SF)
The procedure described in Section 3.1.5: of EPA-450/4-80-027,
Guideline for Use of City-specific EKMA jn Preparing Ozone SIPs,
requires use of the initial mixing height in calculating emission
fractions. While other procedures are possible, the recommended
approach, is most consistent with the conceptual basis of city-
specific EKMA.
Is it valid to do back trajectory with surface wind data? (SF)
For the trajectory in EKMA/OZIPP, a column ofUir is assumed
to inoye with the general wind flow downwind of. the city. More
sophisticated trajectory methods are not recommended because:
1) the methods attempt to track each parcel of air, while,
in fact, jnany parcels of air from different locations in the
urban area .make up the column of air that has the maximum ozone
concentration,
19
-------
2) wind networks are often too sparse and sites unrepre-
sentative of the general wind flow to make a true representation
of the wind field possible,
3) during stagnation periods, local effects on wind monitors
are accented due to the variable wind direction at low wind
speeds,
4) influences such as sea or lake breezes are hard to
model properly due to their changing extent in space and time,
and because the model is not designed to explicitly handle
vertical motions or recirculation cases.
Many of these problems occurred when trajectories were constructed
from St. Louis RAPS data.
Approximate trajectories, using smoothed wind fields from
highly resolved networks, can be used to provide an indication
of whether a city is responsible for a peak ozone value.
85. Question: What does EPA define as low initial concentration? (SF)
Reply: A "low initial concentration" is one that results in peak ozone
and control requirements being almost entirely determined by
post 8 a.m. emissions. It is not possible to give a firm
definition of "low," because it depends on a number of factors,
including the amount of post 8 a.m. emissions and dilution.
However, generally initial NMOC levels less than about 0.05-
0.10 ppmC and NO levels less than about .005-.01 ppm may be
regarded as "low* for most urban areas.
86. Question: If the post 8 a.m. emissions are assumed to be uniform, how can
this be of any use in determining precise control strategies?
(SF)
Reply: Fresh emissions may affect calculated control requirements in
some instances. Thus, they are useful in helping to establish
overall control targets for the city to meet. Precise impacts
of a control strategy (e.g., what will happen at location j if
emissions are reduced by x% at location i) can only be deter-
mined using Level I models. Level III modeling is most suitable
for establishing urbanwide emission reduction targets.
87. Question: If one eliminates the post 8 a.m. emissions, is the analysis
still considered to be a city-specific EKMA application that
would be approved by EPA? Does it meet EPA's definition of what
is required? CSF)
Reply: Ordinarily, post 8 a.m. emissions should be considered. However,
there may be some occasions when this may not be necessary.
Decisions of this nature would have to be made on a case by case
20
-------
basis in consultation with the appropriate EPA Regional Office.
Generally, for cases in which very high initial precursor con-
centrations exist, high morning jnlxing heights prevail and peak
ozone concentrations occur well downwind from a city in a county
with low emission density, post 8 a.m. emissions may not be a
key factor in determining control requirements.
88. Question: If post 8 a.m. emisions are derived from the emission inventory,
shouldn't initial conditions also be derived from emission
inventory instead of from measured air quality values? (SF)
Reply: Ambient concentrations of precursors are what react to form
ozone. Whether or not initial ambient conditions are derived
from an emission inventory depends on the degree of confidence
one has in estimates of early morning emissions, trajectories,
mixing heights, and the suitability of a trajectory model for
simulating periods of the day where a large degree of windshear
is often likely. Ordinarily, we believe it is more appropriate
to base initial conditions on direct ambient air quality meas-
urements. However, we recognize that there may be some cases in
which derivation of initial conditions from emission inventories
may be a more appropriate alternative. This is more likely to
be the exception rather than the rule in our opinion, however.
89. Question: A statement was made that ozone production was not actually
sensitive to HC composition changes in smog chamber studies.
Was the change in composition and reactivity of HC considered?
CSF)
Reply: The statement which was made was that the available smog chamber
data suggest that peak ozone is less sensitive to changes in
NMOC composition than implied by some models. Changes in the
composition and reactivity of the NMOC composition were considered
in these experiments.
90. Question: Why does one sometimes get a reversal in plotting?
to double peaks? (SF)
Is this due
Reply: In some rare instances, the algorithm for interpolating the
ozone isopleths may lead to so-called reversals. The problems
may sometimes be corrected by slightly altering some of the
OZIPP inputs that control the generation of the diagram.
Examples include changing the NMOC and/or NO scales, the isopleth
lines which are generated, or the density of simulations as
controlled by the ACCURACY option. If the reversal occurs for
an isopleth which is not being used, it may simply be ignored.
21
-------
IV. Application of Isopleths in EKMA Procedure
91. Question: How do you treat ratios greater than 3Q:1? (P)
Reply: The same as other ratios. However, a ratio of 30:1 is substan-
tially higher than would be expected for an urban area. The
appropriateness of the data should be carefully checked in such
a case.
92. Question: Is it appropriate to develop an NO strategy for controlling
03?
(P)
Reply:
It may be. This would depend on such factors as the prevailing
NMOC/NO ratio, current ozone levels and the feasibility of
reducing NO emissions in the city by the needed amount.
Increasing NO to reduce local ozone levels is not an acceptable
strategy, because it may aggravate multiday transport problems.
93. Question: If one has more than one monitor operating on the same day, how
does one compute design ratios? (P)
Reply: See pages 42-43 of the March 1981 Guideline.
94. Question: How does one handle stagnation when developing isopleth diagrams?
(P)
Reply: The same as other days. See page 25 of the Guideline for
suggestions for estimating extent of transport on such a day.
95. Question: Do NMOC and NO monitors have to be collocated to generate
design ratios? (P)
Reply: Generally, yes. There may be rare occasions where State statutes
or regulations prevent this. Such cases will need to be con-
sidered on a case by case basis in concert with the appropriate
EPA Regional Office.
96. Question: Is there a limit on the base point for NMOC and/or NOV levels?
(P) X
Reply: The EKMA technique requires locating a base point on the isopleth
diagram by finding the intersection of the measured NMOC/NO
ratio line with the ozone isopleth corresponding to the observed
peak. Because of the simplifying assumptions made in the EKMA
technique, the NMOC and NO coordinates of the base point will
raost likely disagree with £he 6-9 a.m. NMOC and NOX levels
measured in the urban area (i.e., the model will not predict 03
perfectly}. If the disagreement is large, the validity of using
the technique may be questioned. As a guide to determining if
the disagreemnt is extreme, the following procedure may be
employed. The 6-9 a.m. NMOC and NO levels measured in the
22
-------
urban core may be used to predict an ozone level using either
the isopleth diagram or the CALCULATE option in OZIPP. If the
predicted ozone level differs from the observed level by more
than +50%, then the disagreement is significant (note that this
disagreement is in ozone levels, not NMOC and/or NO levels)..
In such a case, the input data used to generate the isopleth
diagram should be checked. Especially important are the inputs
affecting dilution, post 0800 emissions and 03 transported
aloft. If no improvement can be made, a decision as to the
feasibility of the EKMA technique in this individual case will
have to be made in concert with EPA.
In the above procedure, an ozone prediction differing from an
observed level by less than +50% was used as the criterion for
acceptable model performance. It must be emphasized that the
control requirement calculations are made in the relative sense
(i.e., it is the change in 0$ levels that is desired). The
factors that affect the absolute predictions are the same as
those that control the positioning of the isopleths on the
diagram, i.e., dilution, post 0800 emissions, transport, sun-
light, etc. Sensitivity studies have shown that the control
calculations are relatively insensitive to changes in these
parameters over the ranges normally encountered. Thus, if the
ozone prediction agrees with the observed level within a +50%
range, then any further improvement may have little effect on
control estimates.
97. Question: Should one take into account future emissions reductions from
past controls, i.e., 1979 SIP control requirements? (P)
Reply:
Yes.
98. Question: Must one always go through the 10-h simulation period when
developing isopleths? (P)
Reply: The OZIPP program is designed to automatically perform a 10-hour
simulation starting at 0800 LOT and ending at 1800 LOT even
though, in many instances, the observed ozone peak may occur in
the early afternoon. Simulations with city-specific EKMA and
other more sophisticated models have shown that when VOC emis-
sions are reduced, the peak ozone levels tend to occur later in
the day. Therefore, to insure that the control strategy will
reduce the peak 1-hour average 03 level to .12 ppm throughout
the day, the 10-hour simulation must be run.
99. Question: Are there any considerations for significant changes in emissions
over the three year period? (P)
Reply: If major changes in emissions have been implemented during a
three year base period, a shorter base period, reflecting
previous large reductions in emissions, should be used.
23
-------
100. Question: Is it true that the more monitors one has, the greater the
control requirements? (P)
Reply: Not necessarily, however the likelihood is increased.
101. Question: When does one decide when transport is to be considered? (SL)
Reply: The transport of 03 aloft should always be considered.
Consideration of ozone transport in the surface layer is optional.
In general, the consideration of precursor transport is not
recommended, but may be considered if the necessary measurements
are available. In such cases, some sensitivity testing may be
performed to assess the importance of the precursor transport.
Appendix B of EPA-450/4-80-027, Guidelines for Use of City-
specific EKMA in Preparing Ozone SIPs, contains procedures along
with some sensitivity results that serve as a guide in conducting
the analysis.
102. Question: Should one use design value or peak value? (SL)
Reply: Model estimates should be utilized as described on pages 10-15
of the March 1981 Guideline rather than using a single peak or
design value for ozone.
103. Question: When does one use mean ratio and when does one use median
ratio? (SL)
Reply: Providing there is close agreement among ratios observed at all
sites in a city on a given day, the mean 6-9 a.m. ratio from
these sites is used to model that day. If there is only one
NMOC, NO monitoring site, or if poor agreement results among
the ratios at different sites, the median ratio from all days
selected for modeling is used. The procedure for calculating
NMOC/NO ratios is described in detail on pages 40-43 in the
March 1981 Guideline.
104. Question: Why not use mean rather than median? (SL)
Reply: The median ratio is generally less influenced by widely diver-
gent erroneous measurements.
105. Question: How does one calculate NMOC/NO ratio? (SL)
^
Reply: See pages 40-42 of the March 1981 Guideline.
106. Question: Why is CDT equal to zone 6? (SL)
Reply: It is necessary to specify "time zone" in order to accurately
define solar zenith angles. Use of a "6" for CDT is consistent
with the internal computer code for the model, as explained on
page 39 of the OZIPP User's Manual (EPA-600/8-78-014a).
24
-------
107. Question:
Reply:
108. Question:
Reply:
109. Question:
Reply:
110. Question:
Reply:
111. Question:
Reply:
112. Question:
Reply:
Why doesn't one have a start or stop time for simulation? (SL)
The OZIPP program is set to begin all simulations at 0800 LOT
and end them at 1800 LOT. The choice of these times is selected
to correspond to the conceptual basis of the model and the data
typically available for use with city-specific EKMA. If one
were to consider a different starting time, the only option
would be a time earlier than 0800 LOT since a simulation must be
started before the photochemical reactions proceed to any
appreciable extent. However, a time much earlier than 0800 LOT
would require some idea of the early morning air parcel trajec-
tory, the NMOC and NO ratio at the time and location of the
trajectory starting point,.and emissions along the trajectory.
Such an approach is more consistent with a Level II or Level I
modeling approach (i.e.* detailed trajectory analyses), and
requires far much more emissions, air quality, and meteoro-
logical data than are typically available for a city-specific
EKMA analysis. With regard to the ending time, see Question/
Response 98.
Why did EPA use default values for isopleths? (SL)
Some values had to be chosen to illustrate the isopleth diagrams
and the EKMA procedure.
What is the applicable date of future diagram? (SL)
Generally, 1987.
Why is 1987 used rather than some other year? (SL)
Because that is the year by which the 1982 SIPs are supposed
to be trying to demonstrate attainment of the ozone NAAQS.
Why doesn't one change emissions for the future? (SL)
See the responses to Questions 31 and 53. It is possible to
change emission fractions in the future if one has reason to
believe that the change in post 8 a.m. emissions will not be
proportional to the change in initial precursor concentrations
in the city. The procedure for doing this is described on pages
59 and 60 of the March 1981 Guideline.
How does one arrive at changes in ozone aloft (.06-.08)? (SL)
The procedure for estimating changes in ozone transported aloft
between the base and future periods is described on pages 53-55
of the March 1981 Guideline.
25
-------
113. Question: How does one decide if areas upwind are attainment or
nonattainraent? [SI]
Reply:
114. Question:
Reply:
115. Question:
Reply:
116. Question:
Reply:
117. Question:
Reply:
This is subjective and needs to be done in consultation with the
appropriate EPA Regional Office. A good procedure would be to
look at a map depicting upwind counties which are classified as
nonattainment versus those which are unclassified. If the emis-
sions in the "nonattainment counties" are much lower than those
in the unclassified counties, it would be appropriate to assume
that the upwind area is "nonattainment."
If one has better data on upwind transport, should you adjust
ozone transport curve? (SL)
The ozone transport curve (page 54 of the March 1981 Guideline)
should not be regarded as immutable. If there is reason to
believe that more locally applicable information exists, this
should be discussed and agreed upon with the appropriate EPA
Regional Office.
Does one need to calculate a
case? (SL)
new NMOC/NO¥ ratio for the future
/\
No. This is done implicitly whenever one applies EKMA to an
ozone isopleth diagram.
How does one calculate NO levels for the future?
(SL)
One should project future NO emissions considering such factors
as national and local NO control programs, growth and retire-
ment of older sources. Changes in ambient NO may be assumed to
be proportional to the projected change in NO emissions. One
exception to this general rule would be if a point source's
effective plume height were above the estimated afternoon mixing
height. Changes in emissions from such a source should be dis-
regarded in the analysis.
If two sites have high ozone concentrations on the same day,
does one need to do two separate analyses? (SL)
As a rule, yes. It is necessary to demonstrate attainment at
all operating sites. However, it may be possible to use the
same isopleth diagram(s) for both sites if they are located in
the same county and the peak 03 levels occur at about the same
time of day.
26
-------
118. Question:
Reply:
119. Question:
Reply:
120. Question:
Reply:
121. Question:
Reply:
122. Question:
Reply:
Are States supposed to model all the highest days at all sites?
(SL)
Yes. However, as indicated in the response to Question 117, it
is not always necessary to generate separate isopleth diagrams
for each site-day. For example, if two sites observed a high
ozone concentration on the same day and the other inputs (e.g.,
post 8 a.m. emissions) to the OZIPP program were similar, the
same diagramCs) could be used for both sites.
How does EPA know that the control requirement is correct? (SL)
There is no guarantee of this. However, as discussed in the
response to Question 144, the impact of control strategies cal-
culated with city-specific EKMA is generally similar to that
estimated with more sophisticated Level I models.
Why do emissions factors drop for the last three hours assuming
example on page 130 of the workbook is realistic? (SF)
The emission factors for the last three hours decrease to reflect
the movement of the air parcel from the high emission density
urban core to a lower emission density suburban area downwind of
the city.
On the base diagram in the workbook, only two curves are shown,
in practice how many curves would one construct? (SF)
In practice, the number of curves to be constructed would be
dictated by the purpose for which the diagram is to be used.
For example, if control estimates are needed, then only two
isopleths would actually be used - one corresponding to the
observed peak ozone level and one corresponding to the standard
(.i.e., .12 ppm). In some instances, however, it may be desir-
able to estimate the effectiveness of potential control strategies.
For example, the following question may be asked: "What will be
the future ozone peak if VOC emissions are reduced by 30%, 40%,
and 50%, respectively?" For this situation, it would be desir-
able to construct enough isopleths to allow reasonable inter-
polation between the observed level and the standard. Note
also, that in some situations, a single diagram may be used for
more than one measured ozone peak (e.g., two sites experiencing
high ozone levels on the same day). In such a case, ozone
isopleths for each observed level being modeled would be constructed.
Is a 10% increase in NO in the diagram in the workbook (page 132)
representative and how was it derived? (SF)
The 10% NO increase was assumed for illustrative purposes
only. Ln actual analysis, the change in NO must be projected
on a case by case basis. The second paragraph on page 47 of
27
-------
EPA-45Q/4-8Q-Q27, Guideline for Use of City-specific EKMA in
Preparing Ozone SIPs, provides some general considerations in
making such projections.
123. Question: If one has specified VOC reductions (.i.e., 55%) how would one
estimate NO reduction, wouldn't one get two values? (SF)
v*
Reply: If the VOC reduction were to be specified initially, the change
in NOX emissions that would be needed to meet the ozone standard
could be found. The procedure would be to locate the base point
on the isopleth diagram as always. The the NMOC base level
would be reduced by pre-specified VOC reduction (i.e., by 55% in
the example). The change in NO emissions needed to meet the
ozone standard would then be found on the .12 ppm isopleth at
the reduced NMOC level. For example, if the base NMOC level
were 1.0 ppmC and a 55% reduction in VOC emissions were speci-
fied, the reduced NMOC level would be 0.45 ppmC. Using the
isopleth diagram, the NO level(s) on the .12 ppm ozone isopleth
would be found at the potnt(s) at which NMOC was .45 ppmC (i.e.,
the point at which NMOC equals .45 ppmC and ozone equals .12 ppm).
The NO change would be computed as the % difference between the
new NO,, and the base point NO .
*» X
Theoretically, two possible NO solutions could exist. One in
which NO was increased and oni in which NO was decreased.
More than likely, technical limitations wilT be associated with
each. Large reductions in NO emissions may not be technologically
possible. On the other hand, large increases in NO emissions
would not be permissable due to the N02 standard, afid because
the increased NO emission would likely result in ozone problems
further downwind. Consequently, the opposite approach is recom-
mended for control strategy development: (i.e., relatively
small anticipated changes in NO should be estimated first and
then the change in VOC emissions needed to achieve the standard
should be calculated).
124. Question: The calculations only relate to peak ozone, how would one relate
an N02 standard to diagram? How could one relate the annual N02
standard to changes in NO levels to meet the ozone NAAQS? (SF)
A
Reply: The standard OZIPP program is not designed to generate peak
hourly N02 isopleths. Under most circumstances, proportional
rollback is a suitable means for estimating the impact of NO
control strategies on peak short term and annual average N02,
providing the Impacts of certain source categories are discounted.
A detailed description of suitable simplified modeling procedures
for N02 is contained in Technical Basis for Developing Control
Strategies for High Ambient Concentrations of Nitrogen Dioxide",
EPA-450/4-80-017.
28
-------
125. Question: If one has an NO problem, couldn't one run the program for
the winter case in nee NO levels are higher then and therefore
consider seasonal control of UC? (.SF)
Reply: See the response to Question 124.
126. Question: Wouldn't the impact of the Federal Motor Vehicle Control
Program affect the average NO conditions?
J\
(SF)
Reply: Yes. This should be factored in to the projections for NOX
when using the EKMA isopleth procedure.
127. Question: Should one expect much reduction from the control strategy
.by reducing ozone aloft? (SF)
Reply: An assumed reduction in the levels of ozone aloft will lower
estimates of controls necessary to meet the ozone standards.
The degree of that reduction can only be determined on a case by
case basis.
128. Question: Is there an assumption in example 1, page 142 of the Workbook,
that sites are far enough apart that input variables vary? Why
can't one just choose a design value? Can an example be provided
where one just uses a design value? (SF)
Reply: In the example problem, percent reduction estimates are assumed
to have been determined on a site by site basis. That is,
isopleth diagrams were developed for each site for each day.
Pre-1982 SIP procedures for estimating VOC emission reductions
involved the concept of a design value. Pages 8 and 9 of
EPA-450/4-80-027, Guideline for Use of City-specific EKMA in
Preparing Ozone SIPs, contain the rationale for replacing the
design value concept with the specific site/day modeling approach.
129. Question: Suppose one had two HC measurements and two ozone monitoring
stations downwind, shouldn't one look at the meteorology and
establish if there are source-receptor relationships before one
develops a control strategy? (SF)
Reply: As suggested on page 10 of the March 1981 Guideline, one should
review the wind data to ensure that only those cases in which
the city impacts the ozone monitor (i.e., the monitor is not
upwind of the city) are modeled. If a monitor is clearly not
impacted by the city on a given day, that siteday should not be
modeled, even if the observed ozone level is very high.
130. Question: If one has numerous ozone, HC and NO stations, wouldn't it be
necessary to go through more extensive trajectory analysis? (SF)
Reply: Not for a Level III analysis.
29
-------
131. Question: What is the purpose of going through procedures for determining
design values? (SF)
Reply: To determine whether or not an area is currently in compliance
with the NAAQS and to provide one means for tracking trends.
132. Question: Why does one pick control requirements on a site specific basis
as opposed to lumping all sites together and picking third
highest control value? Is more control needed with the third
highest approach? (SF)
Reply: The NAAQS for ozone implies that it should be met at all sites.
Therefore, demonstrations are needed at individual sites.
-Lumping calculated controls together for all sites and then
choosing the third highest control requirement (assuming two
years of data) could produce an estimated requirement which is
too high under some circumstances.
V. Efforts to Validate EKMA
133. Question: Does uncertainty grow with time? (P)
The uncertainty bands shown in the comparisons of EKMA with
historical trends do not necessarily grow with time.
Reply:
134. Question: What was the basis for selecting the 95th percentile? (P)
Reply^ The 95th percentile daily maximum ozone concentration was
selected as a trend parameter in the comparison of EKMA with
trends for two reasons. First, because it is at the high end of
the frequency distribution for ozone, it should be affected by
control programs. Second, it is not so subject to fluctuations
in meteorological conditions as the maximum peak daily ozone
value.
135. Question: When comparing the SAI model, does one assume the same chemistry?
(P)
Reply: In the comparisons between the SAI grid and trajectory models
and EKMA, the SAI models utilized the carbon bond II chemical
mechanism, whereas the EKMA isopleths were generated using the
Dodge propylene-butane mechanism. In other words, there was no
effort to ensure identical chemistry in the models.
30
-------
136. Question:
Reply:
137. Question:
Reply:
138. Question:
Reply:
r
139. Question:
Reply:
Is there any kind of trend between EKMA and SAI in St. Louis?
(P)
For strategies in which NMOC is reduced by less than about 60-
70%, there is no apparent trend or systematic difference between
city-specific EKMA and the SAI Urban Airshed Model in the
St. Louis studies. For strategies in which NMOC is reduced by
more than about 70%, there is a tendency for EKMA to predict
greater reductions in peak ozone values than the Airshed Model.
Were there any comparisons conducted between EKMA and Rollback?
(P)
Yes. Comparisons between city-specific EKMA, rollback and the
Urban Airshed Model have been conducted for various strategies
on three days in St. Louis. Although there are some exceptions,
the general tendency in these comparisons is for rollback to
predict greater changes in peak ozone levels than either of the
other two models for NMOC reductions less than about 60-70%.
This difference is more pronounced if the prevailing NMOC/NO
ratio is high. For large NMOC reductions (e.g., greater than
about 60-70%), the difference between EKMA and the Airshed Model
versus Rollback in the St. Louis studies becomes less. For very
large reductions [e.g., greater than 70%), agreeement between
the Airshed Model and rollback is better than between the Airshed
Model and EKMA. For these very high reductions, EKMA suggests a
greater impact on peak 03 than either of the other two models.
It is not clear whether the preceding relationship would hold
for cities in which base level peak 03 and NMOC/NO ratios are
far outside the ranges observed in St. Louis.
In looking at urban areas, does one take into account reductions
between counties? (P)
Yes. This can be done by altering post 8 a.m. emissions and
generating a second isopleth diagram as described on pages 59-60
of the March 1981 Guideline. This need not be done however if
the analyst believes the differences in emission trends among
counties to be trivial. See responses to Questions 118 and
119.
Are isopleths reproductions of smog chamber data? (SL)
No. Smog chamber data were used as the basis for the model's
assumptions about reactivity. However, cityspecific isopleths
reflect the impact of dilution, fresh emissions, transport and
sunlight intensity characteristics of the city being modeled.
31
-------
140. Question:
Reply:
141. Question:
Reply:
142. Question:
Reply:
143. Question:
Reply:
How does one account for variability in meteorology? (SI)
A given isopleth diagram reflects a constant set of assumptions
about meteorology. This is appropriate, since one is interested
in assessing the impact of precursor controls without confounding
differences in meteorology. However, it is also important to
consider several different sets of meteorological conditions/
transport, etc. This is done by generating isopleth diagrams
using day-specific inputs for several different days as recom-
mended on pages 10-15 of the March 1981 Guideline.
How well does the model do at predicting levels at or near
0.12 ppm? (SL)
Only a limited number of cases in which observed ozone concen-
trations are low have been examined. These include one day in
San Francisco, one day in St. Louis and three days in Tulsa in
which peak observed ozone levels were 0.11-0.15 ppm. For four
of the five days tested with Level II version of OZIPP, predic-
tions agreed with observations within +_30%. For all four tests
conducted with Level III OZIPP (i.e., used as the underlying
basis for city-specific EKMA), agreement between observations
and predictions were within +30%. One caveat should be kept in
mind in interpreting these results. In at least some of the
cases, the ozone concentrations are low because the meteoro-
logical conditions are unfavorable for buildup of high concentra-
tions. Good model performance under these conditions does not
necessarily guarantee good performance in the case with poor
meteorology after completion of major control programs.
Has final Tulsa, Oklahoma report been sent to the planning
agency? (.SL)
The report on applying city-specific EKMA to Tulsa is currently
undergoing review. It should be available for general distri-
bution in September as EPA-450/4-81-005b. See the response to
Question 28 for procedures to obtain this document.
Were cityspecific EKMA comparisons made only for St. Louis? (SL)
No. Comparisons between city-specific EKMA and more sophisticated
models have been made in St. Louis, Los Angeles, San Francisco,
Sacramento, and Tulsa. Comparisons between OZIPP predictions
and observed peak ozone concentrations using a Level III data
base have been made in St. Louis and Tulsa.
32
-------
144. Question: If an elected official asks how good the model is, what should
be the response? (SL>
Reply: The results to date suggest generally good agreement between the
impact of controls predicted with city-specific EKMA and with
Level I models, particularly photochemical grid models. These
results suggest that the likelihood of underestimating the
impact of controls by more than 10% is only about 10%. The
OZIPP model's predictions of peak 03 has been compared with
observed peak ozone in St. Louis and Tulsa. Twelve of 14 avail-
able comparisons agree within +30%. Comparisons with historical
trends present more ambiguous results. About 60-70% of the
available comparisons made in the Los Angeles basin appear to
agree. For most cases where there is disagreement, the model
underpredicts the observed change in ozone. The bulk of avail-
able comparisons suggests the model performs reasonably well.
However, individual comparisons exist where the model has done
poorly. Therefore, several days and control scenarios are
simulated in the SIPs to enable better recognition of such
outliers.
145. Question: Were comparisons made using new data validation techniques
or did EPA use data as it was recorded? When was this
completed? (SF)
Reply: Four approaches have been used to evaluate EKMA:
(.1) comparison of predicted changes in ozone concentrations
with observed changes in peak and 95th percentile daily maximum
ozone concentrations;
(2) comparison of predicted changes in peak ozone concen-
trations obtained with EKMA against changes in peak ozone predicted
with sophisticated (Level I) photochemical grid and trajectory
models;
(3) comparison of peak ozone values predicted with a
modified version of OZIPP in a Level II analysis with observed
peak ozone concentrations; and
C4) comparison of peak ozone values predicted with OZIPP in
a Level III analysis with observed peak ozone concentrations.
Work is continuing with the St. Louis data base. However, work
with the Los Angeles, San Francisco, and Sacramento data bases
has been completed. There may be some additional tests with the
Tulsa data during FY-82. In addition, extensive studies of
Philadelphia will begin in FY-82.
33
-------
146. Question: Mas comparison between EKMA and other models done with actual
monitoring values or just the predicted values from each model?
CSF)
Reply: See the response to Question 145.
147. Question:
Would city specific EKMA provide the most promising results
for Los Angeles, i.e., less controls as comparied to trajectory
or airshed models? (SF)
No.
If one characterizes a day as being stagnant, should one attempt
to construct trajectories with light variable winds? (SF)
See #84.
How does one treat ozone aloft on stagnation days? (SF)
To estimate ozone transported aloft during stagnations, one
chooses a monitor which is not likely to be impacted upon by the
city and notes the ozone concentration observed shortly after
the breakup of the nocturnal inversion. The procedure and
suitable monitoring sites for this purpose are described in
greater detail on pages 23-25 of the March 1981 Guideline.
150. Question: Typical trajectory in Salt Lake City shows flow reversal and
land sea breezes, will the trajectory method work in this
situation? (SF)
Reply: See #84.
Reply:
148. Question:
Reply:
149. Question:
Reply:
VI. General Modeling Questions
151. Question: What thought has been given to the level of modeling applica-
tions used for regulatory control and those used for scientific
analysis? (P)
Reply: In an ideal world, the same level of sophistication would be
applied to both. However, it is recognized by the EPA that the
data base, computational facilities and scientific expertise
needed in a scientific analysis of photochemical pollution are
extensive. Therefore, detailed scientific analyses can be per-
formed by relatively few groups. One role of such analyses is
to identify acceptable simplifications of potential use in
regulatory applications. For regulatory applications, recogni-
tion must be given to resource and budgetary constraints on the
part of those responsible for prescribing or implementing
controls. This means that, in many cases, a compromise needs to
34
-------
152. Question:
Reply:
153. Question:
Reply:
154. Question:
Reply:
155. Question:
Reply:
be struck between the possible and the ideal. Use of city-
specific EKMA by many cities in the 1982 SIPs is the result of
one such compromise. Scientific analyses need to continue to,
among other things, identify further improvements which can be
made to models suitable for widespread use in regulatory
applications.
Does a critical point exist between sensitivity and design
values? (P)
The peak ozone level significantly affects the degree of VOC
controls needed to achieve the standard. The most sensitive
region is at ozone levels near the standard.
Should EKMA be treated as other models, that is first, make a
test run of the OZIPP curves, calibrate the model, then design
the control strategy? (P)
Yes. And, in actuality, this is what is done. As indicated
in the response to Question 96, one should first check to see
whether the model is able to predict ozone within +_ 50% of the
observed value. If such agreement is obtained, one should
proceed with the EKMA procedure. In effect, by using observed
peak ozone values in identifying the base point on the ozone
isopleth diagram, one _^s_ calibrating the model prior to its use
in calculating control needs for each site-day.
May ozone isopleth diagrams be applied to post-8:00 a.m.
emissions? (P)
Yes. The procedure for estimating control requirements applies
to post 8 a.m. emissions, as well as to pre-8 a.m. emissions.
The impact of solely reducing post-8 a.m. emissions can be
estimated using the procedures described on pages 59-60 of the
March 1981 Guideline.
Why is it necessary to reduce emissions over a fairly large
area (.county) when it may only be necessary to reduce emissions
in the urban core? (P)
As described on Pages 31-32 of the March 1981 Guideline, it is
ordinarily not recommended that spatial resolution finer than a
county be used in the emission inventory for city-specific EKMA.
However, the model is capable of considering grid squares as
small as 10 km x 10 km. Given model predictions, whether or not
a particular strategy is necessary or desirable is an economic,
political and institutional issue which needs to be addressed on
an individual basis.
35
-------
156. Question:
Reply:
157. Question:
Reply:
158. Question:
Reply:
159. Question:
Reply:
160. Question:
Reply:
161. Question:
Reply:
162. Question:
Reply:
Is there an updated videotape of the models? (SL).
No.
How good is the original EKMA for training purposes? (SL)
We do not recoraraend that past training material on EKMA be used".
We suggest using the March 1981 Guideline, the workbook entitled
"Workshop on Procedures Csic} to Demonstrate Attainment of the
NAAQS for Ozone in the 1982 SIPs," and these questions and
answers to prepare training materials.
Explain procedures for upwind monitoring to obtain ozone
.data aloft. (SL)
These procedures are explained on pages 23-24 of the March 1981
Guideline.
How does one account for unusual growth upwind in ozone future
transport? (.SL)
By using procedures described on pages 52-53 of the March 1981
Guideline. The only difference would be that one would assume a
higher value for transported ozone in generating the second
isopleth diagram.
What is the regression equation for Level II EKMA vs. air
quality? (SL)
A regression equation has not been derived for these comparisons.
Is there any explanation as to why the St. Louis data result in
underprediction as compared to other cities? (SL)
We believe that the major reason for poor model performance
on the days with underprediction in St. Louis is uncertainty in
the air parcel trajectories. Generally, on these days, calcul-
ated trajectories skirted or avoided the city altogether.
We are currently examining other procedures for estimating
trajectories.
Have any attempts been made to validate the standard EKMA
model? (SL)
Yes. Generally, it does not perform as well as city-specific
EKMA.
36
-------
163. Question: What is the error band around the .model predictions? (SL}
Reply: This is difficult to say, because we do not know what the
"correct" answer is with certainty. One must be willing to
assume that state-of-the-art Level I photochemical models, which
perform well in predicting ozone and precursor concentrations
during the base case, yield the correct answer. If this assump-
tion is made, results of our studies thus far (including the
Tulsa studies) could be interpreted to mean that the likelihood
is greater than 90% that the estimated impact of a control
strategy estimated with EKMA is within 15% of the "correct"
value. That is, if the "correct" reduction in peak ozone were
50%, p(.A03 = 50 j^ 15) > .90. If one is primarily concerned
about EKMA being used to prescribe controls which are too
stringent, the results suggest that the likelihood of under-
estimating the impact of controls by more than 10% is only .10.
Thus, if the "correct" reduction in peak ozone were 50%, p(A03
^60%) = .10.
164. Question: Can the resulting error band be used to determine the error
band on the control point? (SL)
Reply: See the response to Question 163.
165. Question: Is there any reduction in cost possible by reducing the number
of isopleths in OZIPP? (SL)
Reply: No cost reductions will be realized by reducing the number
of isopleths to be drawn. However, other procedures can be used
to accomplish some savings in computer time. These are described
in Section 3.2.3 (page 46) of EPA 600/8-78-014a, User's Manual
for Kinetics Model and Ozone Ispoleth Plotting Package.
166. Question: What is the cost difference between a one point run versus an
isopleth run?
Reply: With the standard OZIPP procedures, approximately 65 single
simulations are performed to generate an isopleth diagram.
Thus, one would expect a significant difference between isopleth
runs and single point runs. Example costs and times on EPA's
UNIVAC computer are shown below:
One Point Run
Isopleth Run
Computer Time/min.
0:15
3:00
Cost
$ 1.00
14.00
37
-------
TECHNICAL ht.'URT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-450/4-81-030
2.
3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
Questions and Responses at the 1981 Workshops on
Procedures To Demonstrate Attainment of the NAAQS
for Ozone in the 1982 SIP s.
5. REPORT DATE
September 1981
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
8. PERFORMING ORGANIZATION REPORT NO.
Edwin L. Meyer, Gerald L. Gipson & Robert F. Kelly
9. PERFORMING ORGANIZATION NAME AND ADDRESS
U. S. Environmental Protection Agency
Office of Air Quality Planning and Standards
Research Triangle Park, North Carolina 27711
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
12. SPONSORING AGENCY NAME AND ADDRESS
Same as above
13. TYPE OF REPORT AND PERIOD COVERED
Final
14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
16. ABSTRACT
Report presents responses to 166 questions asked at three 1981 workshops
describing the basis and application of city-specific EKMA. Subject matter
discussed includes the conceptual basis for EKMA, implications of the statistical
form of the ozone NAAQS, monitoring support needed for EKMA analyses, generation
and application of ozone isopleth diagrams and results of efforts to validate EKMA.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
ozone
State Implementation Plans
City-Specific EKMA
Air Quality Monitoring
Model Validation
b.lDENTIFlERS/OPEN ENDED TERMS
COSATl Field/Group
18 DISTRIBUTION STATEMENT
Unlimited
19 SECURITY CLASS IThis Report/
[20 SECURITY CLASS iThis page/
21 NO. OF PAGES
40
32 PRICE
EPA Form 2220-1 (Rev. 4-77) PREVIOUS ECI~>ON s OBSOLETE
------- |