United States      Office of Research and   EPA/600/A-95/041
Environmental Protection  Development      March 1995
Agency        Washington DC 20460
&EPA   Proceedings
         Fourth US/FRG/EC
         Workshop on
         Photochemical Ozone
  PROPERTY BF Problem and Its Control
                          European Commission
                        Science, Research and Development
       , Regional, and
Global Scale

Issues and Studies in the 1990s

Marriott Hotel, Charleston, SC, USA
June 13-1 7, 1994

                                          March 1995
          Fourth US/FRG/EC
Workshop on Photochemical Ozone
       Problem and Its Control:

          Urban, Regional, and
              Global Scale

       Issues and Studies in the 1990s

        Marriott Hotel, Charleston, SC, USA
               June 13-17, 1994
        Atmospheric Research and Exposure Assessment Laboratory
              Office of Research and Development
              U.S. Environmental Protection Agency
             Research Triangle Park NC 27711 USA
                                    Printed on Recycled Paper

               Workshop Co-Chairmen
B. Dimitriades                     D. Jost
Environmental Protection           Umweltbundesamt
Agency, USA                        Germany
                Steering Committee

              P. Roth, USA, Chairman
                K. Becker, Germany
                 H. Jeffries, USA
                  R. Derwent, UK
                 D. Kley, Germany
                  Project  Officer
                 Basil Dimitrides
Atmospheric Characterization and Modeling Division


     This Workshop Proceedings report was developed by assembling
summary  statements  prepared  subsequent  to the  Workshop by  the
Steering Committee with the help of assigned Workshop participants.
Although  the  statements  reflect the  Workshop  discussions  in
general, they have not been subjected to review, and, therefore, do
not  necessarily  reflect  a   consensus  of  viewpoint  among  the

     Although the research described in this report has been funded
wholly or  in part by the United  States  Environmental Protection
Agency, it has not been subjected  to Agency review and, therefore,
does not necessarily reflect the views  of the  Agency, and  no
official endorsement should be inferred.


     The Workshop was organized at the  instigation of Drs. Basil
Dimitriades (USA)  and Erich Weber  (Germany).  The Workshop program
and agenda were developed by Drs. Dimitriades and Becker (Germany) .
Workshop arrangements were made by Research & Evaluation Associates
on  contract  with USEPA.    The Workshop  proceedings report  was
prepared based on presentations and follow-up discussions.  Summary
statements were prepared by the Steering Committee with the help of
assigned participants.  The contribution of these individuals and
of all Workshop participants are acknowledged with gratitude by the
Workshop organizers.   Last, but not least,  the financial support by
the  USEPA, Bundesministerium  fur Umwelt  und  Reaktorsicherheit
(BMU),  German Ministry  for  Science  and  Technology  (BMFT),  and
European Commission  (EC), is gratefully acknowledged.


     This Workshop was  held within the framework  of cooperation
under the  US-German Environmental  Agreement  (Agreement  Number:
Oil, Project  Number:   002).   It  contributes  to high  priority
activities identified by Minister Toepfer and Administrator Thomas
during their December 1987  meeting.  As of to date, such Workshops
have been held as follows:

     First Workshop:      Cologne,  Germany,  May 4-6, 1988

     Second Workshop:    Chapel Hill,  NC,  USA, June 5-8,  1990
                         (Held in coordination with the EC)

     Third Workshop:      Lindau/Lake Constance, Germany,
                         June 30-July 3, 1992 (Held  in coordination
                         with BMFT and EC)

     Fourth Workshop:    Charleston, SC,  USA,  June 13-17,  1994
                         (Held in coordination with BMFT and EC)

                      TABLE OF CONTENTS




                 ENVIRONMENT PROGRAMME	   3
          1.3.2  EUROTRAC	   4

          2.1.1  BIOGENIC HYDROCARBONS	   5
          2.1.2  PEROXY RADICAL CHEMISTRY	   6
          2.1.3  NIGHTTIME CHEMISTRY	   6
     2.6  REACTIVITY ISSUES	  11
     2 .8  CONCLUSIONS	  11

                 INVENTORY DATA	  12
          3.1.2  RESEARCH/DEVELOPMENT NEEDS	  13
     3.2  SUMMARY	 17

      EVALUATION	 17
      4 . 2  PANEL DISCUSSION	 19



      7.1  PANEL DISCUSSION	  26



                         WORKSHOP AKNODKCEMENT
                     FOURTH US/FRG WORKSHOP

                  Issues & Studies in the 1990s

[To be held within the  framework of  cooperation under the United
States-German  Environment  Agreement.   It  contributes  to  high
priority   activities   identified   by   Minister   Toepfer   and
Administrator  Thomas  during their December  1987 meeting.  It is
supported by the USEPA, the German Government (BMU and BMFT), and
the European Commission  (EC).]

Charleston, SC                          June 13-17,  1994
United States of America

     This  is to announce  the  organization of the  Fourth US/FRG
Workshop on the  photochemical  ozone  pollution problem, scheduled
for  June  13  -  17,  1994,  in  Charleston, SC,  USA.   Dr.  Basil
Dimitriades, USEPA, Research Triangle Park, NC, USA,  and Dr. Dieter
Jost, UBA,  Berlin, Germany, are the Workshop Co-Chairmen.  As with
the last two Workshops,  the scope of this Fourth Workshop also has
been expanded  to include participation of other countries within
the European Union (EU).

     The purpose and  scope of the Fourth Workshop relate to common
concerns of  the  United States  and Europe  about  the  pervasiveness
and  effects of  photochemical  air pollution  and also about  the
relatively  low  confidence with  which  past  and on-going  ozone
control efforts in the US have  been received.  Consistent with the
recommendations from  the Third  Workshop, the Fourth Workshop seeks
to  focus on specific  issues  that are  expected to  dominate  the
research scene in the decade of 1990s, in the areas of atmospheric
photochemistry,  emission  inventories,   instrumental  analysis  of
pollutants, and ozone air quality modeling. Some currently thought
to be key  issues to  be discussed are the  status of  Process-Based
Modeling (PBM)  science and prospects for future advances, questions
of how to deal with the PBM uncertainty problem and how promising
are  the  observational approaches to studying the  photochemical
ozone problem.  Furthermore, because of the strong current interest
in the regulatory policy-science inter-connection, discussions are
also to be included on policy developments and their influence on
research  in  the various  countries,  and  on  experiences  from
regulatory applications of air quality  models.  An outline of the
program planned, in terms of Workshop Session titles, is attached.
Specifics  on intended Workshop topics and  on  the input  to  be
solicited from the Workshop participants are explained next.

     In the atmospheric photochemistry area,  the dominant issues
ontinue  to be  those  related  to  disagreements  between   model
predictions and smog chamber data and/or field data on ozone yield,
on NOX inhibition of  ozone,  and on  composition of NOx-oxidation
products.   A  major  cause of such  disagreement is thought  to be
errors in the  chemical mechanism of aromatic and biogenic Volatile
Organic Compounds (VOCs) and in the  mechanism of ozone formation
under low  concentration and VOC/NO  conditions.   Testing of such
mechanisms against data from special  smog chamber tests provides a
powerful tool for researching the problem, but such  smog chamber
data  are not  currently  in existence, or  are  not  comprehensive
enough for that purpose,  and, also, are not easy to obtain.  These
and other issues in the voc photochemistry, ozone-olefin reactions,
and  radical  reactions  areas  are  Workshop  topics  for  which
presentations and discussions are invited.

     The Workshop agenda includes for the first time, the closely
related subject of  interactions  of the ozone  problem  with other
photochemical air pollution problems.  Studies of such interactions
are  essential  to understanding  the  effects  of ozone  related
controls   on   other  photochemical  pollutants   (e.g.,   aerosol
particles, acid pollutants, etc.), and vice  versa.   Furthermore,
expending  the  scope  of photochemical  ozone models  to  treat
formation  and transport of  other  photochemical  pollutants will
greatly  enhance the  value and  credibility   of  models.    First,
attention will be given to the coupling of photochemistry and fine
particles.     It   is  frequently   observed  that  high  ozone
concentrations   are   associated   with   high   fine   particle
concentrations.  Aside from their  health and visibility effects,
particles by virtue of their large surface area also act as sinks
of free radicals thereby slowing the photochemical process.  Such
issues  also  are  Workshop  topics  for which presentations  and
discussions are solicited.

     In the emissions inventory area, the bulk of on-going research
effort continues to be on studies seeking to improve measurement of
sources  and  emissions with  emphasis  on mobile sources,  area
sources, and natural VOC emissions,  where  the  largest uncertainties
are  thought  to  exist.   A  new  development is  an  increased
appreciation of the importance of natural NOX emissions and of the
large uncertainties  associated with  mobile source NOX emissions.
However,  the  crucially important issue  of  inconsistency between
emission  inventory  and ambient concentration  data  has attracted
only sporadic research efforts, mainly for development of methods
for assessing degree of consistency,  and very little for remedying
the problem (e.g., direct measurement of pollutant fluxes  in urban
plumes).    Discussions  of  these  issues  and  suggestions  for
encouraging both new  and needed research are invited.

     Numerous field studies in the US and other countries have been
completed  since the Third Workshop  in 1992, and results are being
reported and interpreted both in terms of evidence pertaining to
local problems, and in terms  of evidence  advancing the science in


the areas of ozone-to-precursors dependencies,  ozone and precursors
measurement and distribution within the boundary layer, and model
testing methodology.  The need has been stressed for a concentrated
and coordinated effort to analyze all such evidence,  define areas
of agreement and disagreement,  and  ultimately identify issues of
national and international scope to  be the subjects of recommended
future research.  Special attention also needs to be directed (a)
to instrumental analysis methods for hydrocarbons, carbonyls, NO ,
NO ,  H2O2, HN03,  HN02, organic peroxides, and OH,  for  use in field
studies,  (b) to the utility of  such and other observational data
alone or in conjunction with grid models as an  approach to studying
the photochemical ozone problem, and  (c) to the significance of the
BNL-free troposphere exchange process in accumulation and transport
of ambient ozone.   An effort is  made  by the Workshop organizers to
solicit presentations and discussions on recent field studies and
especially on the utility of observational data as a complement to
grid models,  and to derive conclusions with respect  to lessons
learned  and implications  in  terms   of  new  study approaches  and
specific research needed.

     In the model  evaluation area,  the  dominant  issues are those
related to  disagreements  between Eulerian (regional/urban)  model
predictions and field data on levels  of NO  in the atmosphere, on
NO^ oxidation and  oxidation  products,  ana  on the efficiency of
oxidant production. Also at issue is how much of the disagreement
is due to chemical  mechanism errors  and  how much is due to failure
of the model to resolve  scales  of  chemical and  meteorological
interactions in the Eulerian framework.  Process-level testing of
the models at rural concentration levels is an  important complement
to smog chamber testing, but requires special, difficult to make,
chemical  measurements.  Another  broader scope new  issue is  the
question of how close the PBM modeling science is to the point at
which  further  development efforts  will likely have diminishing
returns.    The Workshop  organizers solicit presentations  and
discussions on  new measurements to probe the model  chemistry in
situ on diagnostic analyses of  model behavior, to illuminate its
chemical workings or the dependency  of the chemistry on scale, and
on  viewpoints  regarding  potential  for  further  improvement  of
process-based models.

     A significant  part of the Workshop effort will be dedicated to
discussing  topics  relevant  to the  application of science  to
regulatory programs and to the  interdependence of regulatory and
research programs.  Workshop participants  are  invited to report on
recent regulatory policy developments, experiences from regulatory
applications of models,  and the  influence that regulatory programs
are having on future research in the US,  EU and the other countries
represented in  the Workshop.    Of particular  interest  are topics
specifically  related  to  the  issue  of VOC  vs.  NO  control  for
ambient ozone reduction, to the use of ambient monitoring data in
assessing  and/or  diagnosing  ozone  episode  problems  and  most
importantly  to  the  treatment  of   the  uncertainty  factor  in
development of model-derived ozone strategies. Presentations and


discussions are invited  that would help compare  experiences and
viewpoints from different countries and identify issues of common

                         FOURTH  D8/FR6/BC  WORKSHOP
                   In Urban, Regional, and Global Scale
                    Issues & Studies in the 1990s
Marriott Hotel
Charleston, B.C.
                 June 13 - 17, 1994
Sunday, June 12

6:00 PM - 9:00 PM
Hotel Lobby

Mondav. June 13
7:00 AM - 8:00 AM
Hotel Lobby
Monday. June 13

8:00 AM
           B.Dimitriades, USA
Envir. Protection Agency

             E.Weber, Germany
Department of Environment

              H. Ott, Belgium
EC—Envir. Research Programme

          W. Schoett, Germany
German Embassy, Wash. D.C.
8:15 AM
Keynote Address
                 G.Foley, USA
           Acting AA/ORD, EPA
         SESSION I.  Regulatory 6 Research Developments
                                    Co- Chairs;  B.Dimitriades, USA
                                                 K.Becker, Germany
8:45 AM  Review of the EU Regulatory Work in the
         Field of Photochemical Pollution

                           P. Hecq. Belgium

9:00 AM  Results from Monitoring and Modeling Studies  J.  Pankrath f  Germany
         of Ozone Reduction Strategies for Germany
9:15 AM  AM Photooxidant Research within the CEC
         Environmental Programme

9:30 AM  EUROTRAC:   A Coordinated Project for
          Tropospheric Research

9:45 AM  The North American Research Strategy for
         Tropospheric Ozone (NARSTO):   An Update
                                                       G.Anaeletti.  Belgium

                                                       P.  Barrellf  Germany

                                                       K.  Schere. USA
10:00 AM
        SESSION II.   Atmospheric Photochemistry

                                        Chair;   H.Jeffries,  USA
10:15 AM  Evaluation of Chemical Mechanism for
          Ozone Formation in Europe

10:40 AM  Critical Mechanisms of the Photooxidant
          Chemistry and Their Evaluation in European
          Simulation  Facilities

11:05 AM  Aromatic VOC Chemistry Uncertainties and
          Their Impacts Upon Model-Derived Ozone
          Control Strategies

11:30 AM  Chemical Mechanism of the Oxidation of
          Aromatic Hydrocarbons

12:55 PM                         LUNCH

1:30  PM  Uncertainties in Chemical Mechanisms for
          Urban and Regional Scale Air Quality

1:50 PM  Simulation of Tropospheric Chemistry for a
         New Generation of Atmospheric Models.

2:15 PM  Development and Evaluation of Advanced
         Chemical Reaction Solvers for Use in
         Photochemical  Models

2:35 PM  Tests of some Reduction Hypothesis in
         Photochemical  Mechanism for Air Quality

3:00 PM                          BREAK
                                                       D. Derwentf UK

                                                       K.H.  Becker. FRG

                                                       R.Atkinsonf USA

                                                       I. Barnes. FRG
                                                       J. Milford, D.Gao,
                                                        and Y-J Yang, USA
                                                       M. Gerv. USA;
                                                       H. Jeffries,

                                                       B. Aumont. France

3:15 PM  Recent Development of the Gas-Phase Chemical   W. Stockwell. FRG
         Mechanism for the Reg. Acid Dep. Model

3:40 PM  Recent Laboratory Kinetic Studies Related to   G. Le Bras. France
         VOC Oxidation (NO3+RO2, OH+Oxygenated VOCs)

4:05 PM  Peroxy Radical Chemistry: Recent Studies on   G. D. Hayman. UK
         Reactivity Patterns                           T. P. Murrells, UK

4:25 PM  Impact of Heterogeneous Reactions Involving   J. Pleim. USA
         HO, and N2O5 on Mesoscale Photochemistry—     F.Binkowski
         A Modeling Study

4:45 PM  Indoor Smog Chamber Measurements of Incremental
         Reactivity Factors                            N. Kellyf USA
5:05 PM                  BREAK FOR DINNER
7:30 PM  Panel Discussion on Ozone Mechanism Issues

               Moderator:  H. Jeffries, USA
           Panel Members:  R. Derwent, UK
                           G. Le Bras, France
                           K. Becker, Germany
                           R. Atkinson, USA

     —   What  are current  major  scientific  issues  (e.g.,  chemistry  of
          aromatics/of biogenics,  ozone chemistry under  certain precursor
          concentration conditions, carbonyl photochemistry, etc.)?

     —   What are their relative importance?

          What are current major research needs and what are their relative
          priorities (e.g., chemical mechanism issues, mechanism evaluation

     —   methods against smog chamber data/against field data,  chemical
          solvers, computer methodology, etc.)?

     —   What  plans  are there  for  such  research  in  the  US,  Europe,

 9:00 PM                 ADJOURN

Tuesday.  June 14

           SESSION III.  Emission Studies

                                           Chair;  J. Pankrath, Germany

8:00 AM  Emission Inventory Verification Studies:      D.Lavson. USA
         The California Experience

8:20 AM  Comparison of Observed and Predicted Vehicle  R. Zweidinaerr
         Emission Factors for the Fort McHenry and     K. Knapp, W. Ray,
         Tuscarora Mountain Tunnels                    and N. Robinson, USA

8:35 AM  Rationale, Experimental Design and            J. China. USA
         Methodology for Determining the Overall
         Accuracy of Whole City  VOC and NOX Inventories

8:55 AM  Hydrocarbon Emission Estimates Based          M. Roemer.
         on Aircraft and Ground Level Data             The Netherlands

9:15 AM  Advances in Biogenic Emissions Research       P. 7.-innn«=»rparfr USA

9:35 AM                   BREAK

9:45 AM  Regional Natural Volatile Organic Compound    A. Guenther and
         Emission Modeling and Measurements            P. Zimmerman, USA

10:05 AM  A Second-Generation Biogenic Emissions       T. Pierce. USA
         Inventory System for Ozone Modeling           C.Geron

10:20 AM  C14 Estimates of Ambient Biogenic VOC Ratios C. Lewis
          — Methodology & Preliminary Results         W. Lonneman. USA
10:35 AM  Panel Discussion on Emission Inventory Issues

               Moderator:   K. Becker, Germany
            Panel Members:  R. Friedrich, Germany
                            D. Lawson, USA
                            P. Zimmerman, USA
                            M. Rogers, USA

          What  are  current  estimates  of  Emission  Inventory  (El)  data
          uncertainties   for  mobile   source  VOC  and  NOX,  area  source
          anthropogenic VOC,  isoprene and non-isoprene biogenic  emissions,
          and biogenic NOX emissions?

     —   How promising/useful are observational data  for improving El data


11:45 AM
          What are current: emissions-modeling and El  research needs?  What
          are their relative priorities?

          What plans  are there  for such  research in  the United  States,
          Europe, and elsewhere?

                          Break for Lunch
                    (French Quarter Restaurant)
        SESSION IV.  Photochemical Grid Model Development & Evaluation

                                   Chair;   K. Becker, Germany
2:45 PM  The Air Quality Management Approach in the
         US: Modeling Expectations and Application
         USA Realities.  Part I

3:10 PM  The Air Quality Management Approach in the
         US: Modeling Expectations and Application
         Realities.  Part II

3:35 PM  Estimating Model Uncertainty through
         Alternative Base Case Analyses
4:00 PM
4:15 PM  Effect of Eulerian Model's Vertical and
         HorizontalResolutions on the Prediction
         of Ozone Concentration over the Rural and
         Urban Areas

4:35 PM  On the Coupled Use of Photochemical Grid
         Models and Observations-Based Approaches

5:00 PM  Intercomparison Results of Urban Airshed
         Model Versions with Different Meteorological
                                                      K. Demeriian. USA
                                                       P.Roth. USA
                                                       K. Demerjian
A. Han sen,. USA

D. Byun. USA

R. Dennis, USA

J. Godowitchf USA
5:20 PM
                     BREAK FOR DINNER

8:00 PM  Panel Discussion on Photochemical Grid Modeling Issues

               Moderator;  K. Schere,  USA
            Panel Members: R. Derwent, UK
                           R. Dennis,  USA
                           K. Dermerjian, USA
                           F. Fiedler, Germany
                           D. Simpson, Norway
                           P. Roth,  USA

     —   Where  in the  modeling method  (used  for  development of  ozone
          strategies)  lie the major sources of uncertainty [e.g., dispersion
          module  (spatial resolution),  chemistry,  EI-,  wind  field-,  and
          boundary  conditions-inputs,  etc.]?    What are  their  relative

     —   What are current major research/development needs?

          What are their relative priorities?

          What  plans  are there  for  such research  (in the  US,  Europe,
9:30 PM
Wednesday. June 15
     SESSION V.  Model Application and Field Studies — Europe

                                   Chair;  G. Le Bras, France

8:00 AM  Photochemical Pollution in Southern Europe    M. Millan. Spain
         Experimental Results from EC Research
         Proj ects

8:30 AM  Nitrous Acid Occurrence in Urban Areas and    I. Allearini. Italy
         Its Role  in Atmospheric Oxidation Processes

9:00 AM  Trend Analysis of Ozone Concentration
         between 1975-93 in the Cologne Area

9:30 AM  The Influence of Automobile Speed on the
         Oxidant Concentration in Cities During
         Smog Episodes
                             W. Dulsonf Germany

                             D. Kley,. Germany
10:00 AM

10:10 AM Emission Reduction Scenarios for VOC
         and NOX  and their  Effect on Tropospheric
         Ozone — a Case  Study for the State  of

10:40 AM Summer Smog Development in South-Western
  Germany:  Observation and Numerical

11:10 AM  Relations between Ambient Ozone
         Concentrations in Germany and  Precursor
         Emissions in Central Europe

11:40 AM Modeling the Role of Isoprene  Emissions
         in Europe: Impact on Control strategies
12:10 PM
2:00  PM The Impact of Different European Emission
         Inventories on Air Quality Modeling with
         the EURAD Model
                                                      R.  Friedrich. Germany
                                                      F.  Fiedlerr  Germany
                                                       W.  Fricker  Germany
                                                       D.  Simpson.  Norway
                                                       H.  Hassf  Germany

                                   Chair;   J.  Ching, USA
2:30 PM  Intercomparison of Scientific Findings from
         20 Regional Air Quality Studies in Europe
         and N.  America

3:00 PM The United States Photochemical Assessment
         Monitoring Stations (PAMS)  Program for
         Ozone Non-attainment Areas
                                                       E.  Cowling.  USA
                                                       N.  Gerald.  USA
                                                       W.  Hunt,  Jr.
3:30 PM
3:40 PM  Photochemical Ozone in Atlanta,  Georgia:
         Influence of Mobile Source Emissions
                                                       M. Rodgers, USA
                                                   R.  Dubose,  M. Fogelson,
                                                   M.  Meyer,  C. Ross,
                                                   M. Saunders, A.Gilliland,
                                                   G. Grodzinski,  S. Rhudy,
                                                   D.  Jager,  M. Cannon,
                                                   T.  Barker,  X. Gong,
                                                   and H. Honnecutt,

4:05 PM  Ozone Reactivities of Real-World Emissions    J. Sagebiel. USA
         of Organic Species from Motor Vehicles    A.Gertler,  W.Pierson
         Determined in the Ft. McHenry and         B.Zielinska
         Tuscarora Mountain Tunnels
4:30 PM

         (Need to congregate at the dock at 6:00 PM
Thursday*  June 16

     SESSION VI (eont'd):  Field t Model Aplication studies — USA

8:00 AM Assessing the Importance of Polar VOCs        P.  Milne. USA
        as Ozone Precursors:  Observations        A.Bernado-Bricker
        from the SCION Sites                      D.Riemer, R.Zika

8:25 AM Production of Ozone in Rural Georgia          L.  Kleinman. USA
        During the 1992 SOS Campaign              Y-N Lee. S. Springs ton,
                                                   L. Nunnermacker,
                                                   X. Zhou, and L. Newman

8:50 AM Ozone, Hydrocarbons,  NO  and Oxygenated      Y-N LeeP  USA
        Organics at a Rural Southeastern        S.Springston, L. Kleinman,
        U.S. Site                               X. Zhou,   L. Nunnermacker,
                                                 J. Lee, J. Weinstein-Lloyd,
                                                 L. Newman, M. Rodgers,
                                                 C. Stoneking, J. Pearson,

9:15 AM Reflections on a Regional Oxidant Model       S.Roselle. USA
        Study — Predicted Response of Regional       K.Schere
        Ozone to Across-the-Board Reductions in
        Anthropogenic VOC and NOX Emissions

9:40 AM               BREAK

9:55 AM Ozone Strategies for the Northeast U.S.:       E.  Meyers. USA
        Effectiveness of NOX Controls            N.Possiel, R. Wayland

10:25 AM  Ozone Strategies for the Northeast U.S.:     G.  Wolff. USA
         Effectiveness of VOC Controls
11:00 AM                LUNCHEON
              (Trawler Seafood Restaurant)

     SESSION VII.  Observational Models/Methods

                                   Chair;  P. Roth, USA

2:15 PM  The Use of Observation-Based-Models for       W. Chameides. USA
         Studying Ozone Precursor Relationships

2:40 PM  Observation-based Approaches to Determining   S. Sillmanf USA
         the Sensitivity of Ozone to VOC and NOX:
         The "Indicator Species" Approach

3:05 PM  Spatial Mapping of NOx-Limiting  Conditions     C. Blanchard. USA
         Adaptation of the Integrated Empirical        P.Roth. H.Jeffries
         Rate Approach

3:30 PM                   BREAK

3:40 PM  An Experimental Evaluation of the Integrated  N. Kelly,  USA
         Empirical Rate Model

4:05 PM  Atmospheric Reactivity of VOCs: AIRTRAK        M. Hurley, USA
         and the IER Model                             T.Wallington,
                                                       S.M.  Japar
4:30 PM  Panel Discussion on Observational Model/Method Issues

               Moderator:  P. Roth,  USA
           Panel Members:  W. Chameides,  USA
                           S. Sillman,  USA
                           C. Blanchard,  USA
                           N. Kelly, USA
                           S. Japar, USA

          What  applications  are there   for  which  use  of  observational
          data/methods is or could be recommended?

          What  are  current  major  scientific issues  in the  observational
          methods development area?  What are their relative importance?

          What  are  highest  priority  research/development  needs  in  the
          observational data/methodology area?

          What plans are there for  such  research/development  in the United
          States, Europe, elsewhere?

5:45 PM                   ADJOURN

Frida* JUB> 17
     SESSION VIII.  Wrap-up/Conclusions

                                   Chair:  P. Roth, USA

8:15 AM  Preparation of Proceedings statements by Steering Committee
         and assigned participants

               Steering Committee
               Chair:  P. Roth, USA
             Members:  K. Becker, Germany
                       H. Jeffries, USA
                       R. Derwent, UK
                       D . Kl ey , Germany

12 : 00                  ADJOURN

                    FOURTH U8/GBRMAN/EC WORKSHOP
Charleston, B.C.
June 13 - 17, 1994
                        PROCEEDINGS REPORT



     The work of the European Commission  (EC) in general and particularly in
the environment  is mainly oriented  towards the harmonization  of policies
among the twelve Member states and the integration of needs for protection of
the environment in  other  policies  (transport,  tourism, industry,  etc.)*
Photochemical  air  pollution  is  recognized as  being amongst the problems
against which action has to be taken.  In  that field, the regulations adopted
or in preparation  cover both the emission  reductions and the assessment of
ambient air quality.

     On   the    emissions    side,    four   main   sectors   are   covered:
exhaust/evaporation  from  cars,  evaporation from  the refuelling systems,
solvent usage, and combustion installations.  The "Directives" are driven by
the principles of prevention  and correction at the source,  and by the need to
fulfill the obligations resulting from the international protocols on NOX and
VOC reductions.  They are based on the use of the best available technology
with no excessive  costs.  However,  increase concerns for health protection
and ambient air quality are included.

     With regard to ambient  air quality,  one standard has been fixed for the
maximum concentration of N02  on the basis of health protection criteria.  In
the framework  of the photochemical problem in  Europe:  a "Directive" on O3
pollution has been adopted, which  fixes thresholds for health and vegetation
protection  and population  information or  warning.   This regulation  is to
improve monitoring of O3  (and as  far as  possible its precursors) and to
coordinate the actions against photochemical pollution.

     New measures for reducing the  emissions of precursors will be necessary
in Europe, but they  will have to be  justified on the needs  to meet ambient
air quality  objectives.   In the future,  the assessment of  air quality in
general, and of O3 and its precursors in particular, will be a  crucial element
of the European policy towards  cleaner air.


     In Germany, several regulations for  controlling NOX  and VOC emissions
     have been initiated for compliance with:

     • United Nations Economic Commission  for Europe (UNECE) protocols on NOX
        and VOC reductions,
     •  EU directives on ozone and its  precursors,  and
     •  Amendment to the Federal Immission Control  Act.

     Emission control potentials according to this "trendscenario" have been
     estimated to be  36% NOX  reduction and 47% VOC reduction for the year 2005
     compared to 1987.   The trendscenario meets the reduction  goal of the
     UNECE protocols (30% NOX,  30% VOC  reduction).

     For further CO2 reduction, Germany  needs additional technically feasible
     measures which would result in a 50% NOX and  72% VOC reduction.  Further
     reductions cannot be achieved by technical measures alone; economic and
     fiscal measures, etc.  are then required.

—   For assessing  compliance of these national emission reduction potentials
     with air quality levels of  ozone  [e.g., EC  health  protection level of
     110 fig Oj/m3/8  hours average, UNECE accumulated exposure threshold of 40
     ppb O3 for protection of forests (AOT40) ] that guide subsequent pollution
     control efforts,  an Action Programme and  a Measure Plan  for ozone
     reduction have been launched. Model ing-based emission control scenarios
     on the European/Federal States-hierarchy levels are supposed to support
     the decision finding for the Measure Plan.



     The research activities in the field of photooxidants, promoted by the
European  Commission  (EC),  are  presently treated  within the  Environment
Research Programme 1991-1994,  which is the continuation of past programmes
started initially about 20 years ago.

     European  Commission R&D programs  on environmental protection give
particular importance to atmospheric chemistry,  which is part of the Global
Change Area.   The other areas included in the  Environment Programme are:
Technology and  Engineering  for the  Environment,  Research on Socio-economic
Aspects, and Technological and Natural Risks.

     About  15   research projects related to photooxidants  formation are
presently  supported  by the  EC.  The  contribution outlined  some selected
projects with  a brief description of  the main results  obtained, e.g., the
determination  of  rate  constants  and   mechanism for  reactions which are
relevant to ozone formation, the elucidation of the role  of natural VOCs, and
the development  of sampling and analytical techniques for the detection of

     The importance of the subject  is evident from the increase of financial
resources allocated from the previous programmes up to now.  There is need to
mention that  this  part of the  programme  is implemented also  by  concerted
actions, i.e., by coordination of national research in the Member States with
the assistance of a specific Science Panel.

     It is  foreseen to maintain the efforts in this area at  a high level
within the 4th Framework R&D Programme 1994-1998, where research in the field
of photochemical oxidants  as described in the strategic document,  "Chemistry
in the  Atmosphere", prepared by  the adhoc EC  Science Panel, is  included
through topics as the  oxidizing capacity of the atmosphere and the role of
biogenic emissions in photooxidant formation.   These items are the basis to
design and  implement a work programme starting in 1995,  part of  the next
Specific Environment Programme for which an overall allocation of 1080 HECU
has been provided,  including Climatology,  Marine Sciences,  and Technologies

     1.3.2  EDROTRAC

     EUROpean Experiment on TRAnsport and Transformation of Environmentally
Relevant Trace Constituents in  the Troposphere  Over Europe (EUROTRAC), the
European coordinated research project within the  EUREKA initiative, addresses
three major scientific problems in tropospheric research:

     •  the formation of photo-oxidants,
     •  acid precipitation, and
     •  biosphere-atmosphere exchange of trace substances.

     The project has achieved a  remarkable success since its start in 1989 in
bringing together  international groups of  scientists to work on  problems
directly  related to  the  transport  and  chemical  transformation of  trace
substances in the troposphere.

     The contribution briefly outlined the organization of the project, which
involves some 250 research groups in 23 European  countries and focused on two
particular  aspects:  the newly  formed  application project and the current
discussions on a project  to follow EUROTRAC when it finishes  at  the end of

     The Application Project (AP)  is  intended to assimilate the results from
EUROTRAC and present them in a condensed form so that they are suitable for
use by those responsible for environmental planning and management in Europe.
Its principal themes are:

     •  Formation,  distribution and trends of photooxidants in Europe,

     •  Acidification of soil and water and the atmospheric contribution to
     nutrients, and

     •  Contribution of EUROTRAC to the development of tools for the study of
     atmospheric pollution,  in  particular, tropospheric modelling,  new or
     improved instrumentation, and the provision of laboratory data.

     The first draft of the AP report will be ready at the end of 1994, and
the report will be published in 1995.

     Following surveys of  opinion  among principal investigators  and other
interested parties, prospects for formation of a proposed project to follow
EUROTRAC in 1996 are presently under discussion within the project.   It is
likely  that  the  main  goal  will  be  to  understand  and  quantify  the
source/receptor relationships for photo-oxidants and acidifying substances on
a regional  scale  over Europe.   While  the future project  will  probably be
organized similarly to the present one, there will be substantially greater
emphasis on  the work  applications  to current  environmental issues.   The
project will probably  be  formally  introduced towards the  end of  this year


     The US has  undertaken  in  1993-94,  a  major international  research
initiative entitled,  "North American Research Strategy for Tropospheric Ozone
(NARSTO)," to deal with  formation and inter-country  transport  of ozone at
problem  levels.   The  initiative  was instigated  by  recent efforts  of the
scientific and regulatory  communities  in  the  United  States (US) to rethink
the ozone problem and its  controls,  in  the face of the disturbing failure of
past controls to achieve significant improvement of ozone air quality.  It is
intended to serve policy-related  interests of  all segments  of government and
private  sectors,  and  its  scientific  scope  is  limited  to  addressing
outstanding  issues within  the areas  of  understanding  the  processes and
factors associated with problematic accumulation of ozone in ambient air, and
development  of  effective  strategies for  managing the  ozone problem.   A
significant  novelty  of  NARSTO  is  the  joint  support  of  the effort  by
governmental agencies, private industry,  and  university communities in the
form  of  public-private partnerships.   Clearly, this  is  a major initiative
dedicated as much to  supporting  the  regulatory  programs mandated by law or
governmental  agencies  as  to  advancing  the   science   underlying  the
photochemical smog formation phenomenon.

     Regulatory  or  policy-related  concerns  dealt  with  by  NARSTO  are
associated with the role and responsibility of the anthropogenic sources in
ozone  problems in  distinction  from the  natural sources,  trans-boundary
transport of ozone and its precursors, approaches to managing ozone problems,
and the  use of  scientific evidence and  uncertainties  by  the  air quality
management and  policy communities.   Key  science  questions to be addressed
range  from general  ones  pertaining to further  understanding atmospheric
processes responsible  for ozone accumulation and transport to specific ones
pertaining to effects  of emissions and their controls and of meteorological


        This session  had the largest number of contributed papers  in the
conference.  Topics included detailed kinetics studies conducted in the
laboratory, advances in the oxidation reaction mechanisms  of biogenic and
aromatic hydrocarbons, results of smog chamber studies, new formulation

principles for reaction mechanisms for use in air quality models, new fast
chemistry solvers, inter-comparison of model mechanisms produced by
different researchers, sensitivity studies of mechanisms parameters, and
studies of reactivity.  The session also included a panel discussion that
dealt with the questions of what are the major issues currently, what are
their relative importance, what major research needs do they raise, what
are their priorities, and finally what plans are there for such research in
the US, Europe, and elsewhere.



        The degradation  of natural hydrocarbon  is  initiated by reaction,
not only by OH, but also by O, and N03.  The database on the reactions is
reasonably well established at room temperature.  The mechanisms of
subsequent degradations are poorly characterized with carbon balances of
less than 60% for even the most widely studied system — the OH initiated
degradation of isoprene in the presence of NOX.  The O3 initiated
degradation of natural compounds is of particular interest because of the
uncertain fate these processes play in the production of OH, organic acids,
and hydroperoxides.

        Current studies  have observed  features  (nitro-oxy  carbonyls and
hydroxy-aldehydes) that have not been quantified.   New techniques need to
be developed for poly-functional compounds to improve the carbon balance
and our understanding of these reaction systems.


        Peroxy radicals  play an important  role  in  photochemical production
of oxidants on both the regional and global scales.  The reactions of
peroxy radicals with NO perturbs the photochemical stationary  state leading
to the production of 03.  In low NOX regions, other peroxy  radical
reactions (RO, + HO,, R02 + RO2,  and RO2 + R'O2)  become important resulting
in the formation of hydroperoxides,  organic acids, and other products.

        The self-reaction of peroxy radicals have  been more widely studied
than the reactions of peroxy radicals with HO2,  NO,  NO2, or other peroxy
radicals.  The database on the reactions of RO2 +  RO2 show  a tremendous
variation in reactivity over six orders of  magnitude, with primary R02s
showing greater reactivity than secondary RO2s  which in  turn are more
reactive than tertiary RO2s.  Activation of the RO2  is shown on
substitution by different functional groups.  Although the databases on the
reactivity of peroxy radicals with HO2 and NO are  less extensive,  the
variation of reactivity is less than an order of magnitude.  Tentative
structure-reactivity trends can be identified,  but further data are needed
to both confirm and extend the patterns.

        There  are no data (kinetic or  mechanistic)  on the  reactions of
peroxy radicals with more than one substituent.  Such peroxy radicals will
be formed in the degradation of biogenic hydrocarbons.  In the absence of

direct measurements, reliance on structure-reactivity relationships will be
needed to estimate kinetic or mechanistic data for these reactions.
Methods to develop peroxy radicals with specific structures are needed
along with instrumentation to detect peroxy radicals both sensitively and


        Based  on CH302 and  C^tlsO2  data, the NO3 + RO2 reaction can be
significant in the conversion of RO2 to RO.   In the case of peroxynitrates,
the extension of the database confirms that the acyl peroxynitrates are
thermally much more stable than the alkyl peroxynitrates.  The new
absorption cross section data of PAN gives a photolysis rate three times
higher than the currently accepted rate.

        Additional  data are  still  needed:  (1)  extension of the NO, + RO2
studies to include larger RO?s;  (2)  mechanism studies of NO3 + VOCs
(especially alkenes) to identify and quantify the products formed (nitroxy-
and carbonyl compounds);   (3) work on the formation and fate of organic
nitrates (RO2  + NO, N03 +  VOC),  particularly in relation to long range
transport of NO ; and (4)  the heterogeneous formation of HONO needs study.

        In  the area of ozonolysis  of alkenes,  rate constants of alkene  + O3
measured as a  function of temperature, and mechanistic studies are needed.
The latter needs to be especially focused on the fate of the Criegee
biradical or other intermediates besides the Criegee biradical.

        Photolysis  rates and mechanisms for carbonyl compounds have been
improved by a  new determination of the quantum yields of methyl glyoxal,
but many more  compounds need to be studied.


        All urban areas in the US, as well  as in Europe,  show a significant
contribution of aromatics (benzene, toluene, xylene isomers, and other
alkylated benzenes) to the emission of anthropogenic hydrocarbons.  Smog
chamber experiments indicate that their contribution to photooxidant
formation is also  significant.  Modeling of ozone  formation/ therefore,
requires a proper description of the impact of aromatics on the
photooxidant chemistry within the chemical module.

        K.H. Becker noted  in his presentation that one of the areas of
uncertainty in current chemical mechanisms for atmospheric chemistry-
involves the atmospheric  degradation reactions of  aromatic hydrocarbons.
As discussed by R. Atkinson, I. Barnes, and K.H. Becker, the aromatic
hydrocarbons react in the atmosphere primarily with the OH radical by two
pathways: H-atom abstraction from the substituent  alkyl groups and OH
radical addition to the aromatic ring to form a hydroxy-cyclo-hexadienyl-
type radical  (OH-aromatic adduct).

       The  subsequent reactions under atmospheric conditions  of the benzyl
radicals formed from the H-atom abstraction channel — the minor reaction
pathway — are now understood.  However, the reactions of the OH-aromatic
adduct are not well understood.  As discussed by R.  Atkinson,  the kinetic
data of Zetzsch and coworkers, and product data of Atkinson and coworkers
dealing with the formation of biacetyl from o-xylene as a function of the
NO, and NO concentrations, show that the OH-aromatic adducts  react with O2
and NO2, with the  O2 reaction dominating in the troposphere.   The products
of these reactions are not known with any degree of certainty, as discussed
by Dr. Barnes, and data are required in the presence and absence of NO .
Dr. Barnes discussed the atmospheric loss processes and products formed
from a series of oxygenated compounds observed and/or expected as products
formed from the atmospheric photooxidation of aromatic hydrocarbons,
including 1,4- and 1,6- unsaturated dicarbonyls.

       The  data presented by  I. Barnes  indicates that  cresols are not
formed in any significant yield from toluene and p-xylene in the absence of
NOX,  but that carbonyl-containing  compounds  (multi-functional  compounds
including poly-ketones and their hydrates) account for a large fraction of
the products.

       The  processes which scavenge NO , in smog chamber studies of the
oxidation of aromatics have not been identified — these processes slow
down the formation of 03 in the chambers.  No indications have been found
in the aromatic oxidation studies for typical peroxy-type reactions of the
OH-aromatic adduct, i.e., reactions of the hydroxy-cyclo-hexadienyl-peroxy
radicals with NO to form oxy radicals, with NO2 to  form peroxynitrates, and
with H02 to  form characteristic hydroperoxides.

       It was pointed out that the photooxidation of aromatics leads to
products with mutagenic properties.  The identity of the products causing
the mutagenicity is not clear, but possible candidates are nitro-aromatics
and unsaturated carbonyl compounds.

       Maleic anhydride appears to be unique to the oxidation of aromatics
and may serve as a "marker" of the extent of transformation of aromatic

       Remaining  Problems for Research:

       •  a direct analysis of the ring-cleavage step  is required and
       could be done with the use of a  suitable precursor; work in this
       direction  is ongoing in Europe,

       •  detailed chemical characterization is needed of the identified
       ring-cleavage products,  in particular,  the photolysis  and OH
       reactions  of the unsaturated carbonyl products,

       •  the toxic compounds formed  in the  aromatic oxidation need to be

        •  a comparison between the oxidation mechanisms of benzene and the
        alkylated benzenes is necessary — benzene may be different,

        •  NO  loss pathways in the aromatic oxidation need to be

            o  artifacts in smog chamber studies, and
            o  important processes in the atmosphere,

        •  without an understanding of the NO  loss,  no reliable definition
        of the  "Ozone Formation Potential (OFF)" is possible,  and

        •  formulation of a representative chemical mechanism  is required
        for models, and also OFF values.


        Smog chambers are the only way to test our understanding of the
chemistry of complex atmospheric reaction systems,  such as the dark and
photochemical oxidation of biogenic,  aromatic,  and real world urban
hydrocarbon mixtures.   Smog chambers provide macroscopic data to aid in the
formulation of reaction mechanisms based mainly on more microscopic data
from the kinetics laboratory.   That is,  chamber data show the combined
effects of all chemical processes acting together,  including ones that may
not have been studied in the laboratory and for which there are no kinetics
rate or reaction pathway data.

        Previous studies conducted in the United States, Japan, and
Australia have provided a large, yet still insufficient, smog chamber
database.  The United States data set,  mainly from the University of North
Carolina (UNC)  and the University of California,  Riverside (UCR),  has been
assembled by the United States Environmental Protection Agency (EPA)  into a
"Standard Chamber Data Base" for use in reaction mechanism testing.  Under
the EPA sponsorship,  the Commonmwealth Scientific and Industrial Research
Organization (CSIRO)  chamber data from Sydney,  Australia,  is currently
being prepared for inclusion into this database.   Also under EPA
sponsorship, additional low level NOX/VOC experiments are being performed
by the Tennessee Valley Authority (TVA)  in their indoor smog chamber.  At
present, the standard database contains over 300 dual experiments from UNC
and over 400 experiments from UCR.   In spite of this large number of
experiments, there are many shortcomings in this data set:  e.g.,  multiple
chamber data exist for only about 20 VOC compounds,  only one chamber has
run a variety of alternative fuel mixtures, and additional chamber
characterization experiments are needed in all chambers.

        Participants at this meeting suggested that new studies are needed

        •  lower NOX conditions more typical of urban and regional levels,

        •  more VOC-mix experiments, including biogenics at realistic

        •  for VOC compositions likely in the year 2005,

          more mixtures typical of alternative fuel emissions  in urban
        background VOC mixtures,

        •  a large number of chamber experiments on newly discovered
        reaction  intermediates  (e.g., hydroxy carbonyls) for which no
        kinetics  data exist and for which organic synthesis will be needed
        to produce the reactant materials, and

        •  many more experiments on heterogeneous reactions on  surfaces
        (especially chamber surfaces) and on particles.

        There is  a need for more standard and advanced equipment that  is
used with chambers.   The sensitivity of many standard monitoring
instruments has been improved and would allow work closer to ambient
conditions in chamber experiments.  Further,  spectral radiometers are now
available allowing detailed characterization of the light flux in chambers
and these should be used routinely.   Major advances in gas chromatography
and mass spectrometry (MS)  have occurred recently and new ion trap MS
detectors show very high sensitivity as well as ability to produce
molecular ion information.   New continuous instruments are also available
for continuous measurement of HONO,  an important radical source in smog
chambers that probably is formed by heterogeneous processes different in
different chambers.   Finally,  measurements of radicals (HO, and OH) in
chambers are needed to confirm our understanding of the internal radical
sources in the mechanisms.

        These new smog chamber experiments need to be conducted in both
indoor and outdoor chambers.   This is because indoor chambers offer
consistent and controlled conditions useful for verifying or testing
portions of the reaction mechanism,  but ultimately the mechanism must
function under outdoor conditions.  Indoor chambers also have a difficult
time creating the intensity and spectral distribution of sunlight and,
therefore, outdoor chambers must be used to assure that the mechanisms do
function properly under atmospheric conditions of radiation and

        Participants also agreed that a  need exists for multiple chambers
in multiple locations.   By showing that a principle mechanism,  in
combination with a chamber-specific auxiliary reaction mechanism,  is
capable of making correct predictions in at least four chambers (two
indoors and two outdoors),  we can develop faith that the mechanism is free
of compensating errors between the chamber-dependent chemistry necessary to
simulate a particular chamber and the chamber-independent chemistry
necessary to simulate all chambers and the ambient air.   Further,  in some
outdoor locations,  climatological conditions differ from those in other
outdoor locations,  and stress reaction mechanism models in unique ways.
For example,  the cooler temperatures and lower radiation intensities in

Europe will provide a different test than the typically hot and humid
summer conditions of North Carolina that are common in the UNC outdoor
chamber data set.

       Finally,  it was agreed that  all  smog chamber groups must  cooperate.
The amount of work that must be accomplished in smog chambers exceeds the
production capacity of all current chambers combined.   Further, each
chamber has unique attributes that can be exploited in testing mechanisms.
All chambers should have the same characterization of radiation intensity
and spectral conditions.   All chambers should be tested with both simple
(e.g., methane) and complex synthetic urban VOC mixtures under identical
conditions to provide a direct comparison of chamber results.  Chamber
groups should also share in quality assurance activities and inter-
comparison tests.


       The applicability  of chemical  mechanisms  for atmospheric  prediction
can be tested, and their ozone production mechanisms diagnosed, using
pathway analysis and the relative and integrated rates of radical-initiated
odd oxygen production (i.e., NO to NO2 oxidation),  and the partitioning of
odd oxygen into ozone and other species.  In such analyses, current
mechanisms generate consistent downwind ozone concentrations through
markedly different reaction pathways, generating sometimes divergent
PeroxyAcetylNitrate (PAN)  and hydrogen peroxide concentrations.

       Explicit  chemical  mechanisms addressing a large number of
hydrocarbons can be generated using structure-reactivity relationships for
the peroxy, oxy,  carbonyl, and multi-functional carbonyls generated by OH
oxidation of the hydrocarbons.

       Individual hydrocarbons  make markedly  different contributions to
episodic ozone formation depending on their OH+hydrocarbon rate
coefficients, mass emission rates, and the typical number of peroxy
radicals produced per hydrocarbon+OH reaction.

       Chemical  mechanisms in use have known  errors because  the  state of
hydrocarbon reaction kinetics knowledge is incomplete, especially for
aromatic and biogenic compounds, and some elements of the mechanisms for
higher alkanes and aromatics.  Even for the well-studied inorganic reaction
set there are still uncertainties, and better characterization of these
rate parameters by National Aeronautics and Space Administration (NASA),
International Union of Pure and Applied Chemistry  (IUPAC), and other
evaluation panels are needed.  A new evaluation currently underway is a
study of the effects of RO2 reactions on the nitrogen balance.  In this
study, the most important reactions are CH,O2  and CH,(CO)O, with HO2, CH,O2,
and with CH3(CO)O2 radicals.

       Estimated uncertainties  in rate parameters and product yields in
the Regional Acid Deposition Model 2  (RADM2) mechanism were produced by
multiple simulations over distributions of the parameters.  For surface


urban conditions, the calculated uncertainties in predicted ozone
concentrations are about 20-30 % (one sigma)  of mean predictions.  Somewhat
higher uncertainties are estimated for H2O2 and PAN.

       Formal uncertainty  apportionment results show the importance  of
improved characterization of photolysis conditions, updating the PAN
chemistry in the RADM2 mechanism, and improved understanding of aromatics
oxidation pathways and of carbonyl and radical yields in alkenes chemistry.

       An expanded version of the Regional Acid Deposition Model (RADM)
mechanism has been produced that includes more explicit input hydrocarbons
(the reaction products remain "lumped" in this version).  The mechanism is
currently being updated according to the most recent evaluated kinetics


       Ongoing  improvements in both computer  capabilities and numerical
algorithms are leading to new methods that are directly applicable to the
simulation of atmospheric chemistry.    In particular, research into
improved chemistry solution modules and expanded representation of "real"
chemistry in the models show promise of significant advances.

       Computationally  efficient solvers  are  essential for large
photochemical air quality simulation models (PAQSMs) because of the
relatively large number of grid cells and the complexity of the chemical
mechanisms (i.e., in terms of the number of species and reactions).
Historically, heuristic solvers, which attempt to exploit special features
in the set of ordinary differential equations (DDEs) that arise from the
chemical reaction mechanism rather than perusing a general and
mathematically rigorous solution, have been employed in PAQSMs.  This is
because these heuristic solvers were faster than the more rigorous solvers
by factors of 10 to 100.  In some instances,  however, these less rigorous
solvers can introduce undesirable numerical inaccuracies.  Advances in
computer technology and commensurate re-structuring of numerical algorithms
now make possible the use of the mathematically sound and more generalized
solvers.   For example, a re-casting of the Gear method, which is especially
designed to solve stiff ODE systems,  and to operate on a large block of
model cells rather than solving one cell at a time, is now considered to be
a viable solver option for PAQSMs, particularly on vector machines, such as
GRAYS.  On a CRAY C90, for example, 10,000 cells of a 40 species reaction
mechanism can be simulated in less than 4 minutes of Central Processing
Unit (CPU)  time.

       The incorporation of generalized solvers,  such  as the Gear method,
can also increase the ease with which different chemical reaction
mechanisms can be incorporated in the model.   Nevertheless,  the ever-
increasing desire to enhance spatial resolution and expanded chemical
mechanisms in PAQSMs will almost surely spur the need for even faster
solvers.   More research in this area is clearly warranted.


       At the same time the solvers are being improved, research  is
underway to expand the chemical detail represented in the reaction
mechanism (particularly,  for the intermediates and final products) without
increasing the number of ODEs that must be solved.  The first public
description of an "allomorphic" chemical reaction mechanism was given at
this workshop.   These mechanisms have both molecules and morphecules.  A
morphecule is a type of dynamic shape-shifting molecule and its allomorphs
are the variants on the morphecule.  For example, an N-ALKANE morphecule
has only normal alkanes as allomorphs.  Allomorphs have not only mass, they
have additional traits, such as OH rate coefficients and other properties.
Thus,  transport, deposit, and emission of VOCs are done using allomorphs
(non-ODE species), but the ODE solver is only operated with the molecules
and morphecules of the mechanisms.  This seemingly contradictory operation
is achieved by a type of "identity splitting" over time.  At the beginning
of a chemical timestep, the concentrations and other traits of the
allomorphic species are used to compute the concentrations and weighted
properties of the morphecule.  After a reaction step, the change in the
morphecules are "back propagated" to the individual allomorphs of the
reacting morphecule and "forward propagated" to the allomorphs of the
reaction products.  When emissions and transport of allomorphs are
computed, the changes of each allomorphs are updated.  Thus, the RO2s and
carbonyl products have time-varying character.

       This  approach has  a number of  advantages,  most notably:

          even  though chemical detail is  vastly  expanded,  the chemical
       solver can more quickly deal with  a small number of reacting

        •  the method eliminates surrogate representations  of VOC's and
       reaction products  instead of tracking the chemical  characteristics
       through  the  radical oxidation  reactions so that the "real" product
       distributions can  be  followed  and  subsequently reacted,  and

        •  such  detailed chemical representations allows for dynamic
       chemical linkages  between models or model cells of  different
       chemical scales (e.g.,  using mechanisms devised for urban,
       troposphere,  and global scales).


       Assigning reactivity  values to individual hydrocarbons is not
straightforward because they are unlikely to be  geophysical quantities
and are likely to be dependent on both chemical mechanism and environmental
conditions.  Thus, the Maximum Incremental Reactivity/Maximum Ozone
Incremental Reactivity (MIR/MOIR) reactivity values depend on
hydrocarbon/NOx ratios and care must  be taken to  use values appropriate to
each application.  More experimental testing is needed of the basic
assumptions to the various "reactivity models" and empirical models
[e.g., 6. Johnson's Integrated Empirical Rate (IER) model] based  on  limited
smog chamber studies.  For example, some reactivity models suggest that


reactivities are additive.  Many attendees have suggested that careful
testing of hydrocarbon mixtures in other snog chambers needs to be
conducted to confirm the various assertions about reactivities

        In addition,  strong temperature effects have been observed  in
captive air outdoor smog chamber studies, and model calculated reactivities
do not include these effects.  The omission of this effect leads to some
uncertainty in understanding the value of reactivities.

        In Europe, the VOC Protocol to the UNECE  International  Convention
on Long Range Transboundary Air Pollution encourages member states to use
reactivity in the development of hydrocarbon control strategies to address
those sources which contribute to ozone formation in Europe.  This work
needs to be extended in its coverage of hydrocarbon species, source
categories,  and to different areas of Europe.


        J. Pleim showed  results  of a preliminary  modeling study using  RADM,
including heterogeneous reactions involving N2O« and HO2.  Aerosols were
represented by a bi-modal method and their surface area was computed as a
function of sulfate concentration,  relative humidity (RH),  and NIL/SO4.
The heterogeneous reaction rates were computed for the diffusion limited
assumption.   However, for the HO2 reaction,  Cu(II)  is  thought to be the key
catalyst.  Therefore, the reaction cuts off above some critical RH when
Cu(II) concentration gets too low.   Probably this is an underestimate of
reality since the reaction probably continues even when it is no longer
diffusion limited.  The model results show potential impact on O3
(downward)  and H2O,  (upward) concentrations.  Clearly, these reactions need
more study,  both in the laboratory and the field.


—      The  degradation  pathways of alkylated benzenes  have to  be studied
        in greater detail  to  obtain a reliable mechanism which  describes
        the  impact of the  aromatics on the photooxidant chemistry at high,
        as well  as,  low  NOX concentrations.

        The  ozonolysis of  biogenic alkenes is of  great  importance for  the
        O3 budget.  A reliable mechanism needed for modeling requires:

             •     a  better understanding  of  reaction intermediates  in  both
                  the gas  phase  and the aqueous phase,  and

             •     a  quantitative analysis of the  formation  yields of
                  different alkyl-hydroperoxides, hydroxy-alkyl-
                  hydroperoxides, and the OH radical.

—     The HONO formation before sunrise observed in -urban areas has a
       significant impact on the OH radical budget in the troposphere.
       The formation mechanism needs studying because models must include
       the processes in their chemical module.

       Different types of laboratory facilities are now available in
       Europe.  There is an urgent need to operate a larger simulation
       chamber where complex chemical mechanisms can be evaluated and
       tested under appropriate troposphere conditions.



       This  session dealt entirely with emission inventory issues.  A
summary of presentations and panel discussions on those issues follow.
Panel discussions dealt with questions on current estimates of emission
inventory data uncertainties,  the potential of observational data and
methods for reducing such uncertainties,  specific research needs in the
emissions modeling and emission inventorying areas,  and plans for such
research in the US,  Europe,  and elsewhere.

       The emission inventory  is a key component of any air pollution
control program.  It includes many types of emission sources,  quantities of
emissions, the temporal and spatial characteristics of emissions, and the
process and emission control devices used at sources.   Air pollution
control agencies use emission inventories to identify potential control
measures and sources that would be subject to controls, to determine
control program effectiveness, and to predict future air quality through
air quality simulation models.  Therefore,  accurate emission inventories
are critically essential to provide the least costly basis and most
effective pollution control strategies.


       Ambient  Reactive Organic Gases ROG/NOX measurements from more than
twenty US cities differ considerably from the corresponding ratios derived
from those cities' emissions inventories.  Although this discrepancy has
existed for years, until 1987 the difference was explained by photochemical
reactions and deposition of NOX.  Ambient CO/NOX ratios also differed
considerably from corresponding emission inventory ratios.  No independent
means to reconcile these gross differences for aggregate whole-city
inventories have been attempted to date.   The magnitude of these
inaccuracies in emission inventories, if left unresolved,  leads to air
pollution control strategies that are flawed, costly,  and misdirected.

       A series of  independent studies resulting from  the 1987  Southern
California Air Quality Study (SCAQS) have documented serious
underestimations of mobile source emissions in current US emission


inventories.  This has been demonstrated through the use of urban tunnel
studies, "top-down" comparison of ambient data with emission inventories,
and air quality simulation modeling efforts.   Since the 1987 Study, similar
efforts from San Joaquin Valley, Lake Michigan, and Southern Oxidant
Studies (SOS) have shown similar results, although these results have not
yet been published in peer-reviewed literature.  The current inventory
appears to underestimate mobile source hydrocarbons and carbon monoxide by
factors of 2-2.5 and 1.5, respectively.  It is uncertain whether NOX
emissions from mobile sources are underestimated, but the most recent
mobile source emissions estimates have shown significant NOX increases,  as
large as 60% in California.  Although it is not certain how long this
discrepancy has existed in the United States, it nay have existed for up to
twenty years.  A major reason for this underestimation is the presence of
high emitters, whose emissions are not being reduced by motor vehicle
inspection and maintenance programs.

        Findings  from two additional studies  suggest mobile  sources are  the
overwhelming source of hydrocarbon emissions in Los Angeles, even though
official inventories suggest that mobile sources are responsible for only
50% of the ROG emissions.  Detailed hydrocarbon speciation suggests a
common source/source type, and receptor modeling also suggests mobile
source/whole gasoline sources account for nearly 90% of the measured

        In  Europe,  the  quality  and  accuracy of  emissions data vary
considerably from region to region.  Only a few studies for emission
inventory verification have been carried out.  However, these indicate for
the most complex emissions models,  e.g., in Germany,  United Kingdom (U.K.),
and the Netherlands, discrepancies similar to those observed in the United
States may not be present.  Other workshop members, including Europeans,
disagreed.   Differences between US and European inventory estimates are
expected due to the following possible reasons:

        •   The mobile source  emission control program  in Europe  is  less
mature than in the US,  and modeled emission reductions in the European
fleet have not yet taken place,

           Methods  for  measuring or estimating  emission factors  from in-use
vehicles from real-world samples differ for the US and Europe,

        •   Different vehicle  fleet  is present —  only  recently have
vehicles with three-way catalysts been sold in a few countries,  and

        •   Unleaded gasoline  is less expensive  than leaded gasoline
(because the taxes on gasoline are higher for leaded than for unleaded
fuels), therefore, fuel switching should be less frequent.

        In  the United States, several studies regarding mobile source
emission verification have been carried out,  but little effort has been
directed toward understanding real-world emissions from point and area
sources in the US and Europe.  In Europe, a tunnel study was recently
conducted in Switzerland, but results are not available yet.  There was no


substantial information presented in the Workshop on ambient European
measurements to verify such inventories, so the accuracies of the European
inventories remain largely indeterminate.

       The serious underestimation of mobile source emissions, along with
possible overestimation of point and area sources in the US inventories,
suggests that a plan for addressing this issue is urgently needed.
Independent approaches to evaluate the overall accuracy of whole-city
emissions inventories are needed to determine the accuracy of the inventory
and to provide critical information from which to diagnose source(s) of
errors introduced in the bottom-up construction of inventories.  In Europe,
there may be a lack of detailed ambient hydrocarbon data along with data to
calculate ROG/NO  ratios,  and a summary  of  existing data and their quality
should be carried out.  Verification of emissions inventories in Europe is
especially challenging because there are several different countries with
differing degrees of resources.

       In the US, major changes recently have been made to  the biogenic
emissions inventory,  suggesting that major efforts are needed in this area.
European scientists report that serious deficiencies exist in their
knowledge of biogenic emissions, and these needs are discussed later in
this summary.

       Emission estimates are produced  by  scaling  a source  emission factor
by some appropriate set of activity factors to yield an overall estimate of
emissions.   Current and future modeling requirements are increasingly
demanding that emission estimates be produced on spatial scales of at most
a few kilometers and time scales on the order of an hour.

       While significant  progress has been made over  the last decade  in
evaluating emission factors, especially in the area of biogenic hydrocarbon
emissions,  proportional improvement in methods of producing localized and
hour-specific activity factors has not been achieved.   In some European
locations,  however,  remarkable progress in providing hourly emissions have
been reported from the GENeration of European EMISsion  (GENEMIS), TRAnsport
of Pollutants over Complex Terrain (TRACT), and "Sanierung der Atmosphaere
ueber den Neuen Bundeslaendern" or Changes in the Atmosphere over the
former Eastern Germany  (SANA) Projects.  Many current methods for
determining annual emissions  (e.g., annual sales of solvents in a city)
provide little information to temporally and spatially allocate these

       Improvements  in emission estimates  over  the coming decade will
require development,  testing, and methods development to produce highly
specific emission and activity factors, for all source types.  A major
problem with all inventories is the lack of scientific methods to evaluate
accuracy of the different components of the inventory.  These must be
developed.   Uncertainty analysis needs to become an integral component of


all inventories.  These emission estimates need to be evaluated for their
ability to reproduce the range of emissions inferred from ambient

       If possible, emission  inventories  in the different European
countries and the US need to be made compatible.   Speciation, temporal, and
spatial resolution need to be improved.  Quality assurance efforts on the
inventories also need to be undertaken.  The policy makers from respective
regulatory agencies should define the degree of accuracy that is needed in
the inventories according to specific needs.   Emission inventories
requiring independent validation, through the use of ambient measurements,
tunnel studies, and model calculations whenever possible.  "Top-down"
emission verification results need to agree with bottom-up approaches.

       Oxygenated  organic  emissions may be significant,  and  they will
become more important in the US with the introduction of federally mandated
alternative fuels in 1995 and Phase-2 fuels in California in 1996.   Many of
these compounds are not measured as part of the PAMS network, and must be
measured in order to quantify the influence of alternative fuels upon air
quality in regions where they are in use.   Future research requires better
analytical tools.  Research needs to focus on one or two dominant
oxygenated compounds to determine the factors which control emissions.
Surveys need to be done to test for emissions of the most likely compounds.
If possible, rapid, inexpensive analytical techniques are required to
improve flux estimates.  Virtually all hydrocarbon speciation studies are
limited by the number of samples that can be analyzed.

       The  relative importance  of biogenic vs. anthropogenic hydrocarbons
in the inventory deserves additional work.  Initial results from the
Southern Oxidant Study (SOS) show that although different approaches  (UC
counting, individual species composition,  chemical mass balance model, and
emission inventory comparison) give somewhat different results, biogenics
probably do not contribute to more than 15% of the total observed
hydrocarbon composition in Atlanta.   This work holds promise for future


The following needs were identified:

       •  In Europe,  a comprehensive  survey is needed of the number and
variety of biogenic sources.  The emission rates for important species are
needed. However, some work is in progress within the European sub-project
BIATEX to produce biogenic emission maps for Europe.

       •  Testing  needs  to continue to determine  if emission algorithms
for isoprene are universal.  Studies need to be done which include nutrient
status and other variables which can be monitored by remote sensing.  Work
needs to continue to elucidate the biochemistry of isoprene so that
taxonomic and ecological links can be developed.


        •  Terpene research needs to focus on the variables that affect
carbon allocation in plants.   Real-time measurement techniques for specific
terpenes needs to be developed.

          Aerometric  flux techniques are promising.  They require further
development to determine the size of the emission footprint they observe
and to better define the range of conditions under which valid fluxes can
be analyzed.

          Remote sensing can be a powerful tool to define vegetation
variability and annual variability.   The flux community needs to emphasize
the measurement of variables which facilitate remote sensing connections.

        •  Emphasis  needs to be made to  conduct studies which  integrate
elements of atmospheric chemistry,  remote sensing, ecology,  plant
physiology,  and plant biochemistry.   Progress in understanding biogenic
fluxes requires interdisciplinary teams working together.

        •  Emissions of nitrogen oxides  from soils may be important  in
rural areas; thus,  it deserves further investigation.


The following needs were identified:

        •  It  is recommended that real-world evaluations  of emission
inventories must continue in order to verify the accuracy of the
inventories, as well as to check for trends in emissions, and overall
effectiveness of pollution control strategies.  Direct measurements at the
source  (of all source types)  are a requirement as input for the

        •  In  both Europe and  the United States, a comprehensive survey  is
needed to summarize all emissions verification data collected to date and
to assess the quality of those data.  Then a test of the accuracy of total
emissions in the inventories needs to be performed.  This needs to done
through the use of "top-down" comparisons of both ambient and inventory
data and examination of hydrocarbon speciation.  In addition to detailed
hydrocarbon speciation, total hydrocarbon measurements are needed.  Due to
concern that mobile source NOX emissions may be underestimated,  this
requires investigation.

        •  In  Europe,  the influence  of  vehicle speed  on emissions  has been
reported, and this needs to be investigated under real-world conditions in
the United States.  The reactivity of hydrocarbon emissions from light-duty
diesels on air quality in Europe also needs to be investigated because a
higher percentage of the European light-duty fleet is diesels than  in North
America.  The real-world emissions of heavy-duty diesels on ozone formation
needs to be studied because the most recent tunnel studies have shown
rather high reactivity resulting from the diesel fleet;  in addition, the


NOX emissions  from the heavy-duty  fleet are  about twenty times those from
the light-duty spark ignition engines.

        •  Within-grid cell studies of hydrocarbon  speciation,  NO and CO
needs to be carried out, in order to verify accuracy of emissions within
grid cells.   Detailed hydrocarbon speciation along with CO and NOX
measurements needs to be carried out at several points within a grid cell
in order to test for representativeness of sampling site data as well as
variability within the cell.   Data needs to be collected at specific one-
hour intervals and compared with the inventory from that cell for the
corresponding hours.  It is important to compare ambient data with
emissions data on the microscale in order to understand whether emissions
are accurate,  and whether they are being allocated to the appropriate
source regions as input to air quality simulation models.

        •  Whole-city emission  inventory verification  and diagnostic
evaluation studies,  if feasible, are needed to provide reference checks on
overall city emission inventories and to provide information to identify
error sources as a step towards improving inventories.  A mass balance
approach that includes observing the total mass flux and measures of flux
by source categories may provide the only known independent technique for
the whole-city evaluation.

        •  Methods development  must continue in  order  to make  measurement
of oxygenated compounds in ambient conditions.  Pollutant species that are
important in point and area source inventories,  such as, aldehydes,
ketones, alcohols,  and esters need to be measured in order to verify their
presence in ambient conditions.

        •  Research  needs  to be conducted  to understand  the  "weekend
effect" that has been observed in Europe and the United States.  In Los
Angeles, ambient ozone precursor concentrations are about 10% lower on
weekends than on weekdays, but ozone concentrations are about 10% higher on
weekends than on weekdays.  In Germany,  the inventory reports 30% lower
emissions during weekends, but no decrease in ozone concentrations is
observed.  Once emissions patterns are understood on weekends, data from
this natural experiment should be used for sensitivity testing for air
quality simulation models.

        •  Within Europe,  it is critical to  obtain  accurate  emissions
inventories from all countries, especially those from eastern Europe.  As
mentioned earlier,  it would be ideal to unify inventories and inventory
procedures for both Europe and North America.

        •  Experiments must be  carried out to verify accuracy  of  the mobile
source emissions models.  For example,  changes in California's mobile
source emission model EMission FACtor Model  (EMFAC) from version EMFAC7D to
version EMFAC7G have resulted in 186%,  142%, and 32% increases of
hydrocarbons,  carbon monoxide,  and nitrogen oxides, respectively.  It is
important to verify that these models are giving the correct values for the
right reasons.



       The workshop has provided a discussion forum of current issues
facing the emission inventory community, both from Europe and North
America.   It is clear that the state of affairs is different in both
regions.   For example, the mobile source emission reduction programs in
Europe do not have the same maturity as those in the United States.
Therefore, predicted decreases resulting from such programs for emission
reductions have not been incorporated into European inventories.   It is
clear that an urgent need exists to provide an overall plan for increasing
the accuracy and credibility of emission inventories,  both in North America
and Europe.   The list presented herein provides suggestions for areas of
future research focus.  Most of all, we recommend that participants in this
effort be directly involved with those who produce the inventories.


        K.  Demerjian and P.  Roth jointly  argued that too great  a  reliance
is being placed on photochemical air quality simulation models (PASQMs) in
national planning, as a consequence of the Clean Air Act (CAA).  In
particular, they stated that:

         •   The State Implementation Plan (SIP) process does not  adequately
reflect the need for interim iterations and correction,  given the United
States history of falling short of SIP projections for improvements in air

            Models are not sufficiently accurate  to justify their use
alone, without corroboration or feedback in time.  As justification, they
discussed eleven types of deficiencies and issues that now accompany model
use, including high noise-to-signal ratio, limited data availability, lack
of "hands off" testing, presence of compensating errors, need for
continuing extension of models to accommodate new understanding, increasing
spatial scales and thus changes in the nature of difficulties to be
resolved, increasing model complexity, high resource demands, inexperience
of modeling staff, sole dependency on approach, and a history of past
failures in projection.

         •   Consideration needs to be given to adopting  "a  process for
attainment demonstration" rather than confining demonstration to the
submission of a SIP.

         •   The process might consist of  de-emphasis of  the use of PAQSMs,
introduction of the use of observation-based analyses to facilitate
corroboration of the results of PAQSMs, emissions- and receptor-oriented
monitoring  (with careful prescription of the attributes of networks and
instrumentation), and assessment of the risk of incorrect estimation of


outcomes.  [An example of the methodology was presented subsequently by A.

        A. Hansen pointed out that a major issue now facing decisions-
makers concerned with attainment of the National  Ambient Air Quality
Standard (NAAQS) for ozone in northeastern United States is determining the
need for reducing NOX emissions  and, if necessary, specifying the  extent of
reductions warranted.  In order to evaluate  the merits of requiring
reductions in NOX emissions, Federal and state  regulatory agencies have
enlisted the use of PAQSMs.  Questions have  been  raised about the validity
of modeling efforts conducted to date.  Dr.  Hansen described an alternative
approach to examining the usefulness of today's models in estimating
emissions reduction requirements.  The primary attributes of the approach

         •  Applying  models of interest in a geographical area having a
data base adequate to support model performance evaluation [such data also
allows the choice of model inputs to be constrained to a narrower range
than routinely allowed by available (sparse)  data],

         •  Accepting the models for application only after adequate
performance is demonstrated,

         •  Generating generic findings for the study area, and

         •  Developing inferences, based upon what has been learned,  for
areas relying primarily on routinely available data,  such as, the

        Key elements  in the development and analysis of the proposed
approach have been termed "risk-based" due to the desire to determine the
risks of developing incorrect findings, and  are:

        •  Evaluation of model performance under the constraints imposed by
use of a sound and comprehensive data base through careful comparison of
concentration estimates and observations, diagnostic analysis,  and
modification of inputs as needed, solely based on scientifically
justifiable arguments.  Demonstration of performance adequacy is requisite
for further work as the degree of correctness of  the results of sensitivity
analyses depends on the soundness of the base case.

        •  Recognizing the  likelihood  of modeling  representations  evolving
that contain compensating errors, an approach to  develop an understanding
of the uncertainties that arise in this situation are:

          —  to purposefully create alternative  base cases that are
essentially equivalent in ozone simulation performance to a reference base
case, and

          —  to examine the response of each base case to changes in
emissions that realistically mirror contemplated  control strategies.


          If responses are similar to the reference base case findings,
which was developed using the comprehensive data base,  the model is
relatively insensitive to the presence of prevailing uncertainties and risk
of developing incorrect strategies is relatively low for that case.  On the
other hand, if responses vary dramatically with the alternative base case,
the risk of developing incorrect findings is high.  In this case, depending
on the model to "point the way" is indeed demonstrated to be risky.
Alternative base cases constructed from less constraining sparse data in
general will have a higher probability of responding differently to
emissions changes than the reference (constrained) base case.

        Demonstration  of  high risk over a  sequence of alternative base
cases is a generic finding.  Yet an effort will be made to clearly
delineate the implications of the Northeast findings.   This will be pursued

         •  degrading the  data base for the data-rich area and comparing
findings  both  with the data-rich simulations and with those of the
Northeast  (for a similar data base), and

         •   comparing  atmospheric  process  dynamics,  physical
characteristics, and attributes for the two areas in an attempt to draw
comparisons and parallels, and if feasible to estimate "bounding levels"
(e.g., a "ceiling" or "floor" on  emissions reduction requirements) for
uncertainties that attend control strategy findings.

Note:  Examples of "degrading" the data base include:

         •   reducing the  number of meteorological  sounding sites available
for Lake Michigan Ozone Study  (LMOS) to a sparse data base and re-doing the
meteorological modeling using only these data, and

            discarding all but one or two  upwind monitoring sites and
assuming that boundary concentrations derived from the retained data are
spatially  invariant.

        In  summary, the motivating elements of this approach  relative to
that being pursued in other current studies are:

        (a)  properly  evaluating model performance and  reasonably assuring
the acceptability of a reference base case,

        (b) developing alternative,  approximately  equivalent  base cases
that are consistent with available data, but contain "undetected"
compensating errors, and

        (c) examining  in  depth the impact  of these errors on  findings of
control strategy analyses.

In this way, modeling risk can be directly assessed.


       A panel discussion was held on significant issues related to
photochemical grid modeling.  Six panelists were asked to respond to any or
all of the following questions:

       •  Do the existing uncertainties in air quality modeling
    (chemical/transport/dispersion models, meteorological models,  and
emissions models)  inhibit their utility for development of ozone control

          What are the most critical uncertainties to reduce,  in order to
gain confidence in ozone predictions and estimates of needed precursor

          What are the principal modeling research directions  in the  next
5 years in the United States,  and in Europe?

       •  What is the proper role of air  quality modeling  systems  in
research studies,  and in air quality management/policy studies?

       •  How important  is  it to quantify errors/uncertainties in  model
results;   can we do this?  and

          What are the needed improvements in observational data to aid
model evaluation,  refinement,  and/or application?  What are the most
critical  data needed?  Is grid scale incommensurability (point-to-grid
comparisons)  an insurmountable problem for evaluation?

       The following principal points were made by the panel members:

P. Roth:

       Emissions are crucial to proper simulations.  Meteorology
       determines the mixing characteristics.  Land/water
       interfaces play an important role  in many serious ozone
       situations, and it is important to correctly model  these.
       We must get the HC/NOX ratios correct in the models to get
       credible results, and we must measure NOX with sufficiently
       high resolution to capture low levels.

X. Demerjian:

       There are not enough precision and accuracy in the  routine
       measurements taken of ozone and precursors.  The same issues
       in photochemical  modeling that were important in 1976 are
       still important today.  In order to avoid a backlash toward
       modeling, we need to turn the process around and use
       modeling and good observations together.

D. Simpson:
       Control strategy assessment cannot be solely based upon one
       or a few case studies, as VOC and NOX effects control vary
       from day to day.  To be responsive to policy makers, we need
       to be modeling longer time periods — preferably months or
       years — and in larger areas.  Simpler models are useful for
       these exercise type, and may provide as useful guidance as
       that from more complex models.
R. Derwent:
       Existing uncertainties do not inhibit their use for strategy
       exercises.  It is the scientists' duty to use the best
       science possible in the models at any given time.
       Meteorology is terribly important; we need to get the
       transport correct to have chances for the chemistry to work
       right.  We need research on coupling the regional ozone
       models to global scale models.  There is too long of a
       process to get current science into operational models; we
       must attempt to shorten it.  Better capability for measuring
       OH  is needed.
R. Dennis:
       Adjusting, or  "tuning", models does inhibit their utility
       for developing control strategies.  Isoprene chemistry has
       some  important errors which the regional models are
       sensitive to.  Air quality models need to be used as
       numerical laboratories to test hypotheses.  Prognostic
       meteorological models as drivers for air quality models need
       more  work in terms of their accurate depiction of
       meteorology.   We need measurement capabilities of OH, RO2,
       and HO2.  Modeling a small number of episodes is problematic
       to gaining robust understanding of model response to
       emissions reductions.

F. Fiedler:

       Better thermodynamics are needed in the meteorological
       models.  Simple models are often incorrectly applied  to
       complex situations.  We need more specific studies targeted
       to discriminate between processes.  Initialization and
       boundary condition problems continue to plague models.  More
       sensitivity studies  are required to bring out uncertainties
       and error bounds.

       A lively discussion  followed the opening panelists' statements
sparked by questions  from the audience.   The initial discussion centered on
specific weaknesses in the models'  chemical,  physical,  and meteorological
processes,  as well as in the input data errors.   The latter portion of the
discussion tackled the question of the "adequacy" of current models for


development of control strategies.  There was some disagreement among panel
members on this issue — some panelists indicated that the current science
is the best we can do at present and we need to use the models in this
vain, while other panelists indicated serious, possibly fatal, flaws exist
in current modeling systems.  Those holding the latter view indicated that
current modeling capabilities may not be adequate for developing
strategies.  The conflicts between the United States' State Implementation
Plan (SIP) process and its definitive modeling need, and the current
unacceptably high error level and uncertainty in modeling systems was


        Session V  brought  researchers together from several  European
countries who reported on:

           the analysis  of O3-network results,

        •   pollution  over  the  Iberian peninsula,

        •   observations  of HONO  in cities,

        •   box, Lagrangian,  and  Eulerian modeling studies  on both the local
        and regional  scale,  and

           emission reduction  scenarios.

        W.  Fricke  used back trajectories to  establish relationships between
O3 concentrations  at  a rural site and emission of O, precursors along the
trajectories.  Correlations among O3 and NOX and VOC, respectively, were
the best for back trajectories 5 days back in time.  The highest O3
concentrations (approx.  150 ppb) were related at the high end of NOX
emissions and VOC emissions along the path from precursors source areas to
receptor.  The Oj/VOC relation was more linear than the O3/NOX relation
which showed an O3 leveling off  for  air parcels that had traveled over the
highest NOX source areas.

        W.  Dulson  showed results from the  operational O3 monitoring at
Cologne.  Trend analysis gave a small  («1 ppb/year) decrease of O3,  based
on time since 1979.   The significance was weak at the 10% level.

        J.  Allecrrini  reported  interesting  nitrous acid (HONO)  measurements
from the cities of Milan and Rome.  He gave evidence for heterogeneous
production of this unstable compound.   Already early in the morning, the
HONO values are quite large (up to 20 ppb in March) and represent a
significant source for OH radicals.

        M.  Millan  presented results  from aircraft studies  over Spain.
Driven by inland heating and associated large-scale convergence over the
Iberian peninsula, the pollutants move away from coastal source areas
toward the interior,  reaching high into the troposphere.


       F. Fiedler reported on ozone concentrations in the state of Bader-
Wuttenberg and Eulerian modeling for an August 1990 episode.   Using a non-
hydrostatic meteorology/chemistry model, he showed that local emission
reductions have negligible affect on maximum O3 concentrations in  cities
because cities in this state are surrounded by other cities and highly
populated areas that lie within the transport-determined chemical lifetime
of ozone and precursors.

       D. Kiev, using  the EURORADMD model, showed the possible sensitivity
of maximum O3  concentration  in cities to automobile speed.  At higher
speeds, the VOC/NOX  ratios are low (VOC/NOX « 2 @ Velocity (v) - 30km/hr)
and vice versa (VOC/NOX » 3  § v  = 20 km/hr).   This causes a dramatic
increase of maximum O3  concentration for a  reduction  speed from, say,  30
km/hr to 20 km/hr.

       D. Simpson showed results from the  European Monitoring and
Evaluation Program  (EMEP) Lagrangian model.  Given the considerable
uncertainty of possibly ± 500% of isoprene emissions, he calculated the
effects of 50% NO and  50% VOC reductions on  summer-time ozone
concentrations under a range of isoprene scenarios.  Although maximum ozone
events were found to be sensitive to isoprene assumptions,  summer-time
monthly average ozone values were quite insensitive.   Thus,  presently, it
seems as if model predictions concerning summer-time ozone could give
useful policy guidance, independent of the uncertainties in isoprene

       H. Hass showed  model results from European Acid  Deposition Model
(EURAD) for the joint EURAD/TOR case (etc)  of summer 1990.   The EURAD model
is originally based on Mesoscale Meteorological Model-Version 4 (MM4)  (now
MM5) and RADM2.  In this model application, Tropospheric Ozone Research
(TOR) observations of critical parameters are compared with model results.
The change of emissions, based on different EMEP emission inventories, was
used to study the model sensitivity regarding emission changes.  For a
model sub-region consisting of Germany, Belgium, The Netherlands, and part
of northern France,  the influence of isoprene emissions was also studied.
The important result was that the additional isoprene increase brought
little increase of the 03 maximum in these  countries  when compared to
additional emission increases from anthropogenic sources.

       R.  Friederich considered a detailed emission  inventory for the
state of Baden-Wuerttemberg.  Based on current emission trends in this
state, he concluded that reductions of VOC of 61%, NO  of 43%, and CO of
70% are expected in the year 2000 compared to 1990.  He also predicted that
VOC/NO  ratios will  decrease.   Such reductions considered for Bader-
Wuttemberg are expected to reduce ozone by « 10% as compared to Fieldler's
model run for the August 1990 base case.


       This session began with an overview of salient points made in a
1991 National Academy  of Science  (NAS) report critiquing scientific


understanding underlying efforts in the United States to reduce a perceived
widespread ozone problem.  Conclusions reached at a November 1993 Air and
Waste Management Association (AWMA) Specialty Conference on Measuring and
Modeling Photochemical Pollutants were also highlighted.  Points receiving
particular emphasis were:

        •  need  for ambient data to serve as means  for assessing progress
in implementing controls and reducing ozone,

        •  need  to better recognize regional nature of ozone  incidents  in
many parts of the United States,

        •  need  to reassess effectiveness of VOC vs. NOX controls, and

        •  need  to define goals, plan  carefully, and maintain good
communications among intensive field studies participants.

        Recently promulgated  regulations to routinely obtain  some of the
ambient data identified as important were described.  The Photochemical
Assessment Monitoring Stations (PAMS)  network resulting from these
regulations will result in an estimated 94 monitoring stations being
operated during June - August by the end of 1998,  in and near 22 most
severely polluted United states cities.  Station deployment is ahead of
schedule so that almost half are expected to be operational in 1994.  Data
will be archived and be accessible to the public through the Aerometric
Information and Retrieval System (AIRS) system.  Air quality measurements
include ozone, NO, NO2,  approximately  60 organic species (including
carbonyls),  and surface meteorological measurements.  Meteorological
conditions aloft are to be measured at one site per city.  Reaction to the
PAMS regulations was generally favorable.   However, several participants
urged NOy measurements be made and efforts be undertaken to deploy NO/NO2
instruments capable of measuring concentrations as low as 5-10 ppb with
confidence.   These NO  NO, and  NOX measurements are needed to support the
use of observational approaches as corroborative techniques.   Many felt
that ozone and NO  measurements constitutes a  more  definitive means for
evaluating photochemical models'  performance than does ozone alone.
Finally, the need for better measurements aloft was identified.

        The  session next focused on selected findings from the  Southern
Oxidants Study  (SOS), an ongoing major program of field measurements and
interpretive studies undertaken in the southeastern United States during
the 1990's.   The importance of mobile sources as contributors to measured
VOC was first pointed out.  Efforts toward developing means for
characterizing mobile emission factors using more flexible models with the
ability to characterize differing mixes of driving patterns were described.
A defined goal is to have a tested approach available early next decade.
Parallel efforts to define vehicle activity levels were also noted — one
surprising finding was many more extremely short "trips" occurred than
anticipated.  A second study to characterize vehicle emission factors in
two tunnels (Ft. McHenry and Tuscarora) was described.   In these studies,
the current EPA MOBILe Emission Factor Model  (MOBILES)  was found to perform
better than an earlier California model did in characterizing Van Nuys


tunnel data.  In fact modeled VOC emissions first appeared to agree
reasonably well with observations,  whereas NO^ emissions were
overestimated.   It was speculated that this finding probably reflects use
of a driving cycle in the Federal Test Procedure (FTP)  underlying the
modeling which does not characterize the predominance of the high speed
cruise operating mode observed in the tunnels.   Need for greater
flexibility in the United States emission model  is, therefore, implied by
these findings.  Another significant finding was the importance of semi-
volatile (>C10) organic species in diesel emissions.  Given Europe's
greater prominence of diesel emissions, it was felt to be particularly
relevant for German studies.  A discussion of the importance of
characterizing reactivity of these emissions ensued, with one camp holding
that use of concepts such as maximum incremental reactivity (MIR) provided
a useful means for ranking importance of emissions, and a second camp
maintaining that such results are misleading due to the preeminence of
relatively "unreactive" emissions more than compensating for their low MIR

        Results from analyzing VOC data from  several rural  SOS sites  were
discussed next.  Complexity of resulting chromatographs and presence of
overlapping peaks results in significant uncertainties in data
interpretation.  Oxygenated compounds  (acetaldehyde, methanol, and acetone)
were typically observed in the 30-60 ppbc range in the afternoon, and
constituted about 10% of peak Non-Methane Organic Compound (NMOC)
measurements.  Other interpretative studies were undertaken to assess
relative importance of ozone generation at the earth's surface in rural
locations.  Two assessment procedures were described: a "radical budget"
approach, and the extent to which observations deviated from the
photostationary state.  The first approach provoked some controversy due to
disagreement about the lack of consideration of chain length in its
approach.  However, both methods appeared to lead to similar conclusions
that the near surface area in such locations is a net source for ozone in
the mid-afternoon.  The highest chemical formation rate for ozone at these
sites typically occurred in mid-morning.  A third presentation of rural SOS
data noted the importance of several oxygenated organics (HCHO, acetone,
and acetaldehyde) as important sources of radical production.  In all,
ozone itself appears to account for about two-thirds of the production with
the 3 oxygenates accounting for the remaining third.  This finding is
encouraging on two counts:

        •   it appears consistent with reported European findings,  and

        •   it is also consistent with the way in which  current chemical
models characterize radical production.

        Another significant finding  was methyl glyoxal  (a product of
isoprene oxidation) accounting for about half of peroxy radicals leading to
PAN formation,  with acetaldehyde constituting an important second source.

        The presentations'  focus,  next, shifted  to results  of regional
modeling studies and the trend data interpreting potential importance of
VOC and NOX controls in reducing ozone in the United States.   A "matrix" of


Regional Oxidant Model (ROM) across-the-board simulations of VOC and/or NOX
reductions in the eastern half of the United States was summarized.  For
the Northeast and Midwest, VOC control appears effective in reducing the
highest of the episode maximum ozone predictions.   More moderate
predictions appear increasingly more sensitive to NOX  reductions rather
than VOC.  In the Southeast, regional ozone generally appears more
sensitive to NOX reduction.   PAN and peroxy radical  predictions appear
similar to the ozone response, while HNO3 production is  nearly proportional
to NO .   Other regional modeling studies focusing  on the northeastern
United States, were also presented.   The regional modeling studies reported
to date feature one severe episode which occurred during the first half of
July 1988.  They are consistent in their findings which suggest:

        •  NOX control appears beneficial over a large area, but may be
counterproductive in some locations with large NOX emissions,

        •  VOC control is  primarily  effective  in reducing the  highest ozone
concentrations near the largest sources of NOX and in  mitigating
counterproductive effects of NOX control,

        •  relative effectiveness of VOC and NOX controls varies day-to-day
with meteorological conditions,

        •  effectiveness of  VOC controls appears more  sensitive to
uncertainties in biogenic emissions than do NOX control  strategies,  and

        •  emissions within  the Northeast Transport  Region appear to be the
primary  cause of high ozone in the Northeast Corridor.

        The presenters concluded that United States' strategies will most
likely have to feature significant reductions in both NOX and  VOC.

        The sessions' final  presentation noted a downward trend in  early
morning measured VOC/NOX ratios which corresponded with  downward trends in
ozone between the mid/late 1980's and early 1990's in many United States'
cities.  The trend in the ratio appears primarily due to reductions in VOC.
This latter reduction was attributed to major reductions in mobile source
emissions arising from fleet turnover and reduced Reid vapor pressure for
automotive fuels.  Discussion then ensued about whether the early morning
VOC/NOX ratio is a good indicator of effectiveness of  VOC vs.  NO controls.
Other studies, such as one in the Lake Michigan area,  were noted as
supporting use of VOC controls to reduce ozone.  Although there was no
agreement about early morning VOC/NOX ratios as a  determinant  of whether
VOC or NO control is more effective, it follows that  reducing this ratio
should make ozone more responsive to additional VOC reductions, though not
necessarily as effective as NOX controls.


        Chameides and co-workers described  the development of  an
observation-based approach for inferring ozone-precursor relationships.


This approach uses ozone and precursor concentrations rather than emission
inventories to drive photochemical model calculations.  The Observation-
Based Model (OBM) which consists of a set of mechanistically complete,
time-dependent box models, uses observed precursor and ozone concentrations
as input, and determines the sensitivity of ozone concentrations at each
location to the changes in precursor concentrations.  The Observation-Based
Application of the Urban Airshed Model (OBA-UAM) uses observed ozone and
precursor concentrations to adjust the emissions inventories that drive an
emission-based model.  A series of OBM runs using data gathered from a 1990
United States EPA study yielded the following conclusions:

        •   an urban area  similar to Atlanta may  experience both VOC- and
NOX-limited ozone episodes during the same ozone season,

        •   high-sensitivity NO  afternoon  measurements are critical  to
characterizing urban ozone photochemistry,

        •   biogenic hydrocarbons play a key role in determining ozone
sensitivities to anthropogenic VOC and NOX emissions in Atlanta, and

        •   early-morning  VOC-to-NOx ratios are a poor  indicator  of ozone-
precursor  relationships in an urban area while the afternoon H2O2-to-HNO3
ratio shows promise.

        Sillman used a  photochemical  model to identify "indicator species,"
which are species, or species ratios, that correlate consistently with VOC-
sensitive versus N0x-sensitive  ozone  in the model.   The model was exercised
using a range of assumptions to identify indicators that were insensitive
to the assumptions.  Three such indicators were identified:

        •   N0y  (or N0y - NOX),

        •   HCHO/NOy, and

        •   H2O2/HNO3.

Delta Oj/delta NO was  also investigated  as an indicator.   The  author
recommended that these species  (especially NO )  be used as a basis  for
evaluating photochemical model predictions for VOC-vs-NOx sensitivity.

        Blanchard and co-workers described the Integrated Empirical Rate
(IER) model, identified key assumptions and limitations, and developed a
set of revisions to IER equations.  Predictions from  a chemical mechanism,
Carbon Bond Mechanism-IV  (CBM-IV), and smog-chamber data from chamber
studies performed at the University of North Carolina and  the University of
California at Riverside, were used to demonstrate that the IER  expression
of maximum Smog Produced  (SP^) as a linear function of initial NOX should
be replaced with the function SP^ = 6*[NO (0)]", where a < 1.   The  authors
also showed that Smog Produced  (SP) could be expressed as
SP = 6*[NOX(0)  - NO (t) ]a, derived three equivalent expressions  for  extent
of reaction, and identified the measurements required for  calculating
extent using each expression.


       Kelly and Wang described  a  set of  smog chamber experiments
performed to test the Integrated Empirical Rate (IER) model and to evaluate
the use of the IER for determining reactivity.   Smog (NO oxidized plus O3
formed) was used as the response variable.  The authors varied initial NO
at fixed VOC concentration, and added VOC or NOX during  the course  of
irradiations.   For a constant radiation flux, NO(0) = 0.4 ppm, and VOC(O)
varying from 2.0 to 2.42 ppm propene, smog formation was not linear with
respect to irradiation time for the entire radiation.  Delta smog was a
linear function of irradiation for some compounds, but not for others.  The
rate of smog formation was independent of the initial NO concentration,
which was varied from 0.4 to 0.58 ppm NO,  for about the first half of the
irradiation period.  Addition of 0.4 ppm NO to a chamber in which VOC(O) =
2.0 ppm and N0(0)  - 0.4 ppm halted smog formation, which was interpreted as
an N02 inhibition effect.

       For Hurley  and co-workers,  Japar reported:

       •  Application of  the  IER model  (in the  form of  the AIRTRAK
instrument)  to the measurement of VOC reactivities shows promise;

       •  There appears to be significant sensitivity to  initial
conditions,  perhaps indicating a poor representation of radical sources
needed to initiate chemistry and thus sensitivity can have a serious impact
on measurements'  reproducibility; and

       •  Relationship between reactivities  calculated  from the  IER model
and airshed modeling is unclear.


       The panelists and  audience  were asked to discuss recommended
applications of observational data/methods,  existing major scientific
issues in the observational methodology area, current research needs and
their priorities,  and plans for such research in the United States, Europe,
and elsewhere.

       The major portion  of the  ensued discussion — repeated by several
individuals — was concern that observational-based approaches would be
used in place of the full dynamical/photochemical simulations.  There was
wide spread support for the concept of using observations-based approaches
in an attempt to corroborate  (or refute)  the accuracy of model applications
or to constrain the standard models.  It was pointed out that observations-
based approaches in place of current dynamical/photochemical models  (UAM,
ROM, etc.) would be problematic.   Among the concerns for overall accuracy
of the observations-based approaches were:

       •  Concentrations  of many species  (especially, primary species)  are
spatially inhomogeneous and,  especially,  have height variance.  Individual
measurements (especially,  surface measurements)  may not represent the
ozone-forming process as a whole.


        •   Some of the observations-based approaches seem too simplified
and may omit the real atmosphere complexity.

        •   Uncertainties associated with observational-based approaches
need to be identified and (if possible) quantified.  For a useful model,
the uncertainties must not overwhelm the information.

        •   Measurements do not represent "perfect information".   In  fact,
measurements often have considerable uncertainty and concern was expressed
that some approaches depend on measurements that are very difficult to

        Despite these concerns,  which point to  future  research directions,
there was widespread support for observational-based approaches as a
complement to standard models.  It was pointed out that observational-based
approaches represent a new and  (hopefully)  intelligent attempt to make use
of atmospheric measurements and identify strengths and weaknesses in the
current knowledge of the ozone-formation process.

                          LIST OF PARTICIPANTS
Dr. Ivo Allegrini
CNR-lnstituto sull'lnquinamento
Via Salaria Km 29300
1-00016 Monterotondo S. (RM)
++39-6-90625349 (Phone)
++39-6-90672660 (Fax)

Dr. Giovanni Angeletti
Direction 12
200 Rue de la Loi
B-1049 Brussels
+ +32-2-2958432 (Phone)
++32-2-2963024 (Fax)

Dr. Kurt Anlauf
Atmospheric Environment
4905 Dufferin Street
Downsview, Ontario
Canada M3H 5T4
(416) 739-4840 (Phone)
(416) 739-5708 (Fax)

Dr. Roger Atkinson
University of California-Riverside
Statewide Air Pollution Research
900 University Avenue
Fawcett Laboratory
Riverside, CA  92521
(909) 787-4191  (Phone)
(909) 787-5004 (Fax)
Dr. Bernard Aumont
Laboratoire Interunivers'rtaire
 des Systemes Atmospheriques
University Paris 7 et Paris 12
UnitS de Recherche Associ6e au
Centre Multidisciplinaire
61 Av. du General de Gaulle
F-94010 Creteil Cedex
++33-1-45171592 (Phone)
++33-1-45171564 (Fax)

Dr. Ian Barnes
Physikalische Chemie-FB 9
Bergische Universitat-GH Wuppertal
GauBstraBe 20
D-42097 Wuppertal
++49-202-4392510 (Phone)
++49-202-4392505 (Fax)

Dr. Karl-Heinz Becker
Bergische University
 Gaubstrabe 20
D-42097 Wuppertal
+ +49-202-439-2666 (Phone)
+ +49-202-439-2505 (Fax)

Dr. Charles Blanchard
956 Kains Avenue
Albany, CA 94706 .
(510)525-8964 (Phone)
(510) 528-2834

Dr. Peter Borrell
Scientific Secretariat EUROTRAC
KreuzeckbahnstraBe 19
D-82467 Garmisch-Partenkirchen
++49-8821-183140 (Phone)
++49-8821-73543 (Fax)

Mr. Joseph J. Bufalini
Atmospheric Research and Exposure
 Assessment Laboratory
Research Triangle Park, NC  27711
(919) 541-2422 (Phone)
(919) 541 -4787 (Fax)

Dr. S. Burton
Systems Applications International
101 Lucas Valley Drive
San Rafael, CA  94903
(415) 507-7179 (Phone)
(415) 507-7177 (Fax)

Dr. Nadia Butkovskaya
Institute of Chemical Physics
Russian Academy of Sciences
Kosyginstr. 4
117334 Moscow V-334

Dr. Daewon  Byun
Atmospheric Research and Exposure
 Assessment Laboratory
Research Triangle Park, NC 27711
(919) 541-0732 (Phone)
(919) 541-1379 (Fax)

Dr. William Chameides
Georgia Institute of Technology
 Earth & Atmospheric Sciences
Mail Code 0340
Atlanta, GA  30332-0340
(404) 894-3893 (Phone)
(404) 853-0232 (Fax)
Dr. Jason Ching
79 Alexander Drive
Research Commons, Bldg. 4201
3rd Floor
Research Triangle Park, NC 27711
(919)541-4801 (Phone)
(919) 541-1379 (Fax)

Dr. Ellis Cowling
North Carolina State University
College of Forest Resources
Box 8002
Raleigh, NC  27695-8002
(919) 515-7564 (Phone)
(919) 515-1700 (Fax)

Dr. Kenneth Demerjian
State University of New York
Atmospheric Sciences Research
100 Fuller Road
Albany, NY 12205
(518) 442-3820 (Phone)
(518) 442-3867 (Fax)

Dr. Robin Dennis
Office of Research & Development
Research Triangle Park, NC  27711
(919) 541-2870 (Phone)
(919) 541-1379 (Fax)

Dr. Richard Derwent
Room 156
Met O (APR)
Meteorological Office
London Road
Berkshire RG12 2SY
+ +44-344-854624 (Phone)
+ +44-344-854493 (Fax)

Dr. Basil Dimitriades
Office of Research & Development
Research Triangle Park, NC  27711
(919) 541-2706 (Phone)
(919) 541-1379 (Fax)

Dr. Willfried Dulson
Institut fur Umwelthunter
 Suchungen der Stadt Kdln
Eifelwall 7
D-50674 K6ln
+ +49-221-2217623 (Phone)
+ +49-221-2217616 (Fax)

Dr. Franz Fiedler
Institute fur Meteorologie und
Universitat Karlsruhe
KaiserstraBe 12
D-76128 Karlsruhe
011 -49-7247-82-2800 (Phone)
011-49-7247-82-4742 (Fax)

Dr. Gary Foley
Office of Research & Development
Research Triangle Park, NC  27711
(919) 541-2108 (Phone)

Dr. Wolfgang Fricke
Deutscher Wetterdienst
Meteorologisches Observatorium
Albin-Schwaiger-Weg 10
D-82383 HohenpeiBenberg
+ +49-8805-920037 (Phone)
+ +49-8805-920046 (Fax)
Dr. Rainer Friedrich
Institute fur Energiewirtschaft
 und Rationelle Anwendung
Universit§t Stuttgart
HeBbruhlstraBe 49 a
D-70565 Stuttgart
+ +49-711 -7806112 (Phone)
+ +49-711 -7803953 (Fax)

Mr. Bruce W. Gay, Jr.
Atmospheric Research and Exposure
 Assessment Laboratory
79 Alexander Drive, Bldg. 4201
Research Triangle Park, NC 27711
(919) 541-2830 (Phone)
(919) 541-1379 (Fax)

Mr. Nash Gerald
Atmospheric Research and
 Exposure Assessment Laboratory
Research Triangle Park, NC 27711
(919) 541-5652 (Phone)
(919) 541-1903 (Fax)

Dr. Michael Gery
Atmospheric Research Associates
160 N. Washington Street
Boston,  MA 02114
(617) 723-1488 (Phone)

Mr. Gerald L Gipson
Atmospheric Research and Exposure
 Assessment Laboratory
Research Triangle Park, NC 27711
(919)541-4181 (Phone)
(919) 541-7588 (Fax)

Mr. James M. Godowitch
Atmospheric Research and Exposure
 Assessment Laboratory
Research Triangle Park,  NC 27711
(919) 541-4802 (Phone)
(919) 541-1379 (Fax)

Dr. Alex Guenther
Atmospheric Chemistry Division
National Center for Atmospheric
P.O. Box 3000
Boulder, CO 80308
(303) 497-1447 (Phone)
(303) 497-1477 (Fax)

Dr. D. Alan Hansen
3412 Hiliview Avenue
Palo Alto, CA 94304
P.O. Box 10412
Palo Alto, CA 94303
(415) 855-2738 (Phone)
(415) 855-1069 (Fax)

Dr. Heinz Mass
Institut fur Geophysik und Meteorologie
Universitat zu Koln
AachenerstraBe 201 -209
D-53931 Koln
+ +49-221-4002258 (Phone)
++49-221-4002320 (Fax)

Dr. Garry D.  Hayman
B 551, Harwell  Laboratory
AEA Technology
OxonOX11 ORA
+ +44-235-432894 (Phone)
+ +44-235-432404 (Fax)
Dr. Pierre Hecq
EC, Direction 11
200 Rue de la Loi fTRMF 7148)
1049 Brussels
32-22-3687-04 (Phone)
32-22-3695-54 (Fax)

Dr. John R. Holmes
Air Resources Board
2020 L Street
P.O. Box2815
Sacramento, CA 95812
(916) 445-0753 (Phone)
(916) 322-4357 (Fax)

Dr. Steven Japar
4518 Whisper Way
Troy, Ml 48098
(313) 845-8072 (Phone)
(313) 594-2923 (Fax)

Dr. Harvey Jeffries
Department of Environmental
 Science & Engineering
C.B. #7400
University of North Carolina
Chapel Hill, NC 27599-7400
(919) 966-7312 (Phone)
(919)966-7141 (Fax)

Mr. Nelson Kelly
General Motors Research Laboratories
Environmental Research Department
30500 Mound Road
Building 1-6
Warren, Ml 48090-9055

Dr. Lawrence Kleinman
Brookhaven National Laboratory
Environmental Chemistry Division
Building 426
Upton, NY 11973
(516) 282-3796 (Phone)
(516) 282-2887 (Fax)

Dr. Dieter Kley
Forschungs-zentrum Julich
 Institut "Chemie der Atmosphere"
52428 Julich
49-2641-613741 (Phone)
49-2461-615346 (Fax)

Dr. Douglas Lawson
Desert Research Institute
5625 Fox Avenue
Reno, NV  89506
(702) 677-3193 (Phone)
(702) 677-3157 (Fax)

Dr. Georges Le Bras
Centre National de la Recherche
3C, Avenue de la Recherche
45071 Orleans Cedex 2
33-38-51-54-61 (Phone)
33-38-51-76-70 (Fax)

Dr. Yin-Nan Lee
Building 426
Brookhaven National Laboratory
P.O. Box 5000
Upton, NY 11973-5000
(516) 282-3294 (Phone)
(516) 282-2887 (Fax)

Mr. William A. Lonneman
Atmospheric Research  and Exposure
 Assessment Laboratory
Research Triangle Park, NC 27711
(919) 541-2829 (Phone)
(919) 541-4787 (Fax)
Dr. Montserrat Martin
CEAM, Plaza del Carmen 4, Palacio
 de Pineda
E-46003 Valencia
++34-6-3865212 (Phone)
++34-6-3865216 (Fax)

Mr. Edwin L Meyer
Office of Air Quality Planning
 and Standards
79 Alexander Drive, Bldg. 4201
Research Triangle Park, NC 27711
(919) 541-5594 (Phone)

Dr. Jana Milford
Dept. of Mechanical Engineering
University of Colorado
Campus Box 427
Boulder, CO 80309
(303) 492-5542 (Phone)

Dr. Millan Millan
Plaza de Carmen 4
Palacio de Pineda
46003 Valencia
++34-6-3865212 (Phone)
++34-6-3865216 (Fax)

Dr. Peter Milne
Rosenteil School of Marine
 and Atmospheric Sciences
University of Miami
4600 Rickenbacker Causeway
Miami, FL 33149
(305) 361-4786 (Phone)
(305) 361-4689 (Fax)

Ms. Christa Morawa
Bismarckplatz 1
D-14193 Berlin
++49-30-23145696 (Phone)
++49-30-89032285 (Fax)

Dr. Heinrich Ott
Directorate General XII
200 Rue de la Loi
B-1049 Brussels
++32-2-2951182 (Phone)
++32-2-2963024 (Fax)

Dr. Jurgen Pankrath
Bismarckplatz 1
14193 Berlin
+ +49-30-8903-2375 (Phone)
+ +49-30-8903-2285 (Fax)

Dr. Julia Partroescu
Universite Bucuresti, Fac. Chemie
 Cat. Chimie Analytica
Sos. Panduri,  No. 90, 7000 Bucharest
40-16310060 (Phone)
40-16310060 (Fax)

Mr. Thomas. E. Pierce
Atmospheric Research and
 Exposure Assessment Laboratory
Research Triangle Park,  NC 27711
(919) 541-1375 (Phone)
(919) 541-1379 (Fax)
Mr. Jonathan Pleim
Atmospheric Research and Exposure
 Assessment Laboratory
Research Triangle Park, NC 27711
(919) 541-1336 (Phone)
(919) 541-1379 (Fax)

Dr. Bruno Rindone
Department of Organic and
 Industrial Chemistry
University of Milano
Via Venezian 21
1-20133 Milano
++39-2-2367618 (Phone)
+ +39-2-2364369 (Fax)

Ms. Laura Roberts
Texas Natural Resource
 Conservation  Commission
P.O. Box13087
Austin, TX 78711-3087
(512) 239-1245 (Phone)
(512) 239-1605

Dr. Michael Rodgers
Georgia Institute of Technology
School of Earth and
 Atmospheric Science
923 Research Drive
Atlanta, GA 30332-0340
(404) 853-3094 (Phone)
(404) 894-9223 (Fax)

Dr. Michiel Roemer
TNO Institute of Environmental
P.O. Box 6011
NL-2600 JA Delft
++31-15-696037 (Phone)
+ +31-15-616812 (Fax)

Mr. Shawn Roselle
Atmospheric Research and Exposure
   Assessment Laboratory
Research Triangle park, NC 27711
(919)541-7699 (Phone)
(919)541-1379 (Fax)

Dr. Philip Roth
836 Fawn Drive
San Anselmo, CA  94960
(415)456-3405 (Fax)

Dr. John Sagebiel
Desert Research Institute
P.O. Box 60210
Reno, Nv 89506-0220

Richard Scheffe
Office of Air Quality, Planning,
   and Standards
Research Triangle Park, NC 27711

Mr. Kenneth Schere
Atmospheric Research and Exposure
   Assessment Laboratory
Research Triangle park, NC 27711
(919)541-3795 (Phone)
(919)541-1379 (Fax)

Dr. Wolfram Schoett
Embassy of the Federal Republic of
4645 Reservoir Road, N.W.
P.O. Box 40680
Washington, DC 20007-1998
Dr. Helmut Selinger
Kuhbachstrasse 11
D-81543 Munich
+ + 49-89-651088-63(Phone)
+ + 49-89-651088-54(Fax)

Ms. Marjorie Shepherd
Environment Canada
Atmospheric Environment Service
4905 Dufferin, Ontario M3H 5T4

Dr. Joseph Sickles,  II
Atmospheric Research and Exposure
  Assessment Laboratory
Research Triangle park, NC 27711
(919)541-2446 (Phone)
(919)541-7588 (Fax)

Dr. Sandy Sillman
Atmospheric, Oceanic & Space Science
University of Michigan
2455 Hay ward
Ann Arbor, Ml 48109-2143

Dr. D. Simpson
The Norweigian Meteorological Institute
P.O. Box 43-Blindern
+ +47-22963213(Phone)
+ +47-22963050{Fax)

Dr. William Stockwell
Fraunhofer Institute fur
  Atmospharische Umweltforschung IFU
Kreuzechbahstra6e 19
D-82467 Garmisch-Partenkirchen
+ + 49-8821-183262(Phone)
+ + 49-8821 -73573(Fax)

Mr. David Sullivan
Texas Natural Resource
   Conservation Commission
P.O. Box 13087
austin, TX 78711-3087
(512) 239-1381 (Phone)

Dr. A. F. Vilesov
MPI fur Stromvngsforschung
37073 Gottingen
•+ + 49-551-7092656{Phone)
+ + 49-551-7092607(Fax)

Dr. Erich Weber
Weberstrasse 29
D-53113 Bonn
+ +49-228-214131 (Phone)
Dr. Patrick Zimmerman
The National Center for Atmospheric
P.O.  Box 3000
Boulder, CO 80307

Mr. Roy Zweidinger
Atmospheric Research and Exposure
  Assessment Laboratory
Research Triangle park, NC 27711
                                       •kv.S. GOVERNMENT PRINTING OFFICE: »»5 - «50-QM/220I7