Interdisciplinary Field, Modeling, and Laboratory Studies Session Summary

DOE Atmospheric Chemistry Annual Meeting, Nov. 1996

1. Introduction

The subject of discussion for this breakout session was cooperative work between field, modeling and laboratory scientists within the Atmospheric Chemistry Program. Our goal was to review examples of past cooperation and the potential for future interdisciplinary studies with the commensurate benefits to DOE. Within this framework we discussed

• specific ways by which models could assist in improving ACP field programs in meeting the goals of OHER
• the type of field programs that would aide in model evaluation studies, and
• laboratory measurements that would benefit both modelers and field scientists.

Our report begins with a summary of key ideas (Section 2), followed by more detailed notes of our discussions on modeling work (Section 3), field and laboratory studies, (including Mexico City) (Section 4), and concludes with a non-systematic survey of inter-disciplinary 'success stories' (Section 5).

2. Summary of key ideas

• The wide range of spatial scales covered by ACP models will necessitate the increased use of nested grids or "models within models" in order to make use of field observations.
• We need closer participation of modelers in the planning stage of field and laboratory programs.
• "Indicator species" were seen as a cross-disciplinary (laboratory, field and modeling) subject of research where ACP could make a substantial scientific contribution of practical utility in resolving the oxidant problems of the United States.
• There is, at present, no receptor modeling capability within ACP.
• Sensitivity analysis techniques appear to be very promising in providing guidance to field and laboratory programs. However, their results to date are of questionable reality because these newly developed techniques have been used only in models of limited dimensionality.
• Chemistry measurements at night, above the nocturnal surface layer are conspicuously absent in the scientific literature, hence the recommendation that ACP scientists work with the DOE Research Aviation Facility in planning a field campaign of nighttime chemistry with associated laboratory measurements to evaluate key uptake coefficients and measurements related to the fate of NO₃ radical at night. A number of specific reactions and species were identified for analysis.
• It was agreed that ACP will not be heavily involved in DOE's 1997 Mexico City aerosol study due to budgetary concerns and time limitations.
• A number of "success stories" were noted involving collaboration between ACP field, modeling, and laboratory scientists, and between ACP meteorologists, chemists and mathematicians.

3. Models

3.1 Three categories of models were identified for use in the analysis of field studies and for planning future studies.

• Process models: with their explicit descriptions of transport, chemical reactions, etc. These models are well represented within ACP.
• Techniques for sensitivity analysis: although not models in the sense of describing processes, these new methods have made great strides within the modeling community.
• Receptor models: which are being used in other research programs but not within ACP.

Comments on these three categories of models are noted below.

3.1.1 Process Models

• As a result of the broad scale of spatial and temporal scale processes spanned by the existing suite of process models there has not been extensive use by modelers of data collected during the ACP field projects. By and large, only scientists involved in the field programs use the resulting observations.
• It was noted that data collected during the long over-ocean transects carried out by the G-1 aircraft during ACP's participation in the NARE '92 and '93 field campaigns could provide data on spatial scales suitable for use within global scale models.
• Closer cooperation is needed between modelers and field scientists during the planning stage of field programs.
• Global- and regional-scale models need to make use of nested grids and "models within models" in order to make use of observations taken during the ACP field campaigns.
• "Indicator species" were seen as a cross-disciplinary tool between field observations and modelers. However, many questions about indicator species remain, including:
• How robust are these species to a variety of chemical and meteorological conditions?
• When do they work as indicators of NO_x or hydrocarbon limited oxidant production?
• How close is the agreement between observations and model predictions of indicator species? Since indicator species frequently involve the ratio of chemical species, they provide a more rigorous test of model performance than do evaluations based on individual chemical species. Hence, a model that performs well in predicting ozone may do poorly in predicting a related indicator species.

3.1.2 Receptor Models. An alternative approach to process models is based on the use of multivariate methods. These methods can be used to evaluate the source of atmospheric aerosols and the mechanisms of chemical transformations during transport. Multivariate methods have proven to be extremely effective for the analysis of field data collected under other field programs but have not been applied within the ACP. Some of the methods that have proven useful in other chemistry programs include:

• Factor analysis (FA),
• Principal component analysis (PCA), and
• Positive matrix factorization (PMF).

3.1.3 Sensitivity Models. This category of model refers to new tools that build on the results of process models; they appear to offer the potential for valuable insights into what quantities need to be measured in the laboratory and the field. The utility of sensitivity studies for the analysis and planning of field studies, and their use for understanding comprehensive chemical models, was cited as one of ACP's "success stories". Illustrative examples of the type of information these tools can provide include:

• The sensitivities of ozone to VOC reactions under polluted atmosphere were found to be a factor of 2-25X greater than the sensitivity of ozone to VOCs under less polluted or relatively clean conditions.
• The key reactions and processes associated with indicator species were found to be different than those associated with chemical species.
• The amplitudes and signs of the sensitivity coefficients suggest that the presence of clouds significantly reduce the total oxidizing capacity of the local atmosphere and potentially alters the O₃-precursor relations.

All of the sensitivity studies presented to date have involved models of limited dimensionality, e.g., box or column models. It is not clear how the results would change in comparable studies involving 3-dimensional models where such non-local processes as transport and mixing are important.

The need for modelers and field scientists to work closely together during the planning stage of field studies was a repeated topic of discussion. Sensitivity analysis was suggested as a tool that modelers could provide to make sure critical species were adequately measured in the field, and to give guidance to laboratory scientists techniques to measure key rate or uptake coefficients. As a result of one sensitivity study involving heterogeneous chemistry, it was recommended that ACP laboratory scientists work towards providing improved information on aqueous-phase kinetics so that modelers can better simulate the aqueous-phase reactions on the deliquescent aerosols and the uptake rate on both dry and deliquescent aerosols.

4. Field and Laboratory Studies

4.1 Nighttime Chemistry and Heterogeneous Chemistry

Reports of chemistry measurements at night, above the nocturnal surface layer were noted to be conspicuously absent in the scientific literature, hence the recommendation that ACP scientists work with the DOE Research Aviation Facility in a study of nighttime chemistry. Specific scientific reasons for this program are given in the following subsection. The logistics for a nighttime program are expected to be difficult as a result of Federal Aviation Administration (FAA) regulations for pilot duty-days and other restrictions related to nighttime flying; however no insurmountable problems are perceived given adequate preparation time. The scientific and practical benefits of such a measurement program are expected to be very high.

Concurrent with the establishment of such a field program is the need for laboratory studies of uptake and absorption coefficients for a variety of heterogeneous reactions. It was noted that improved instrumentation will be required in a nocturnal study of reactions of nitrate radical and ozone with monoterpenes. Specifically, high sensitivity and fast response instrumentation will need to be developed for the monoterpenes. Ozone chemiluminescent studies within ACP have indicated that this may be feasible. Rapid gas chromatographic analysis with monoterpene specific detectors may also be useful. Nitrate radical and fast response ozone monitors will also be needed. Laboratory studies of the reaction rates and products from these reactions are needed to identify marker compounds for other analytical strategies to be developed.

In the case of the heterogeneous chemistry studies of aerosols, it was generally agreed that further laboratory studies are needed and others need to be initiated to identify specific reactions and conditions where these processes are likely to be important. This work will require some time to develop, and will also need to incorporate modeling efforts before any major field efforts are initiated. It was also the group's consensus that ACP laboratory scientists have the expertise to see this work to completion.

4.2 Discussion

This gap in airborne, nocturnal measurements is particularly significant in view of NO_x loss processes suspected to occur at night, and the consequences for oxidant levels. It is common in most parts of the United States for the production of ozone to be limited by the availability of NO_x. Available NO_x may be decreased by up to 1/2 as a result of loss mechanisms associated with the well known reaction of

NO_x + O₃ -> NO₃

which proceeds at a rate of about 10% per hour for typical ozone concentrations. In order to determine whether this reaction is a true sink for NO_x it is therefore necessary to better understand the subsequent chemistry of NO₃. Reactions of suspected or known significance include

NO₃ + HNO₃ -> HNO₃
NO₃ + terpenes -> products
NO₃ -> heterogeneous mechanisms
NO₃ + NO₃ <-> N₂O₅
N₂O₅ -> heterogeneous loss

Rate constants associated with the above heterogeneous loss processes were noted to be a key source of uncertainty in our understanding of atmospheric chemistry. The distribution of formaldehyde (HCHO) and terpenes at night, above the nocturnal boundary layer, is another large source of uncertainty.

Heterogeneous processes play a significant role in determining the speciation of NO_y compounds among HNO₃, NO_x (including NO, NO₂, and NO₃), N₂O₅, HONO, and organic nitrates such as PAN. For example, the hydrolysis of N₂O₅ by H₂O to produce HNO₃, which is now routinely included in atmospheric photochemical models, reduces NO_x concentrations in the lower stratosphere and upper troposphere by shifting NO_y speciation towards HNO₃ via reaction on aerosols and clouds. In the troposphere, there has been speculation that heterogeneous reaction of HNO₃ to produce NO_x species might explain observed ratios of NO_x/HNO₃ that are higher than predicted by models. Most recently, studies of nighttime chemistry have focused on the fate of NO₃ radicals and reaction products, whose removal might reduce overall NO_x levels available for ozone photochemistry. Heterogeneous pathways for these processes may involve species such as nitrates or HONO, whose interaction with condensed surfaces have not been well studied.

Supporting these arguments for additional laboratory studies were results from a sensitivity analysis suggesting both O₃ and the indicator species O₃/NO_y are very sensitive to aerosol surface uptake processes under polluted atmospheres. The most influential species affecting this uptake was found to include: HCHO, O₃, HO₂, HNO₃ (for the indicator species O₃/NO_z), PAN (for O₃/NO_z), H₂O₂, NO_x, and HNO₃ (for O₃).

Laboratory programs need to investigate the heterogeneous kinetics of these NO_y species under a range of conditions appropriate to aerosols and clouds. Even N₂O₅ + H₂O reaction kinetics have not been thoroughly investigated for aerosol compositions that can include concentrated solutions of nitrates, sulfates or sea salts. These are many possibilities for the production-reaction of HONO; as a semi-volatile species it can effectively connect gas and condensed phase processes. Recent laboratory studies have shown HONO can be produced by reaction with CH₂O with HNO₃ in concentrated H₂SO₄ solution or via second order, self-reaction of NO₂ (at high levels). Gas/condensed phase uptake rates for these processes need to be measured systematically in order to derive kinetic formulations that can be applied to models under variable gas and aerosol distributions in the atmosphere. Key to successful experiments is complete characterization of NO_y speciation in order to account for overall NO_y chemical cycles with a range of gas and aerosol mixtures.

Surface measurements indicate significant concentrations of HONO, sometimes even during the daytime. This constitutes a vapor radical source which has implications for rapid generation of ozone. The source of HONO is not known, but a heterogeneous reaction pathway is suspected. If this occurs only near the surface, the overall impact is likely to be small. However there are no measurements above the nocturnal boundary layer with the result that only by making airborne measurements will we be able to evaluate this effect.

It was noted that improved instrumentation would be required for a successful field program involving the nighttime studies of nitrate radical and ozone with monoterpenes. Specifically, high sensitivity and fast response instrumentation would be needed for the monoterpenes. Recent work within ACP indicates that ozone chemiluminescent may be a feasible technique for making these measurements. Rapid gas chromatographic analysis with monoterpene specific detectors may also be useful. Nitrate radical and fast response ozone monitors will also be needed. Laboratory studies of the reaction rates and products from these reactions are needed to identify marker compounds for other analytical strategies to be developed.

4.3 Mexico City

Ozone levels from four stations in Mexico City were presented. With values of up to 350 ppb developing over a few hours from the nighttime values of approximately 40 ppb, there clearly is much of interest from a photochemical perspective. Attention also focused on the large variation in peak values between stations that are relatively close together. Although these observations, taken in light of the interest in heterogeneous processes, would seem to make Mexico City an ideal location for an aerosol/oxidant field study, other factors argued strongly against this conclusion, including:

• The emphasis of DOE's study in Mexico City will be on aerosols, with few supporting air quality measurements.
• Chemical modeling efforts will require significantly better emission inventories for Mexico City than are currently available.
• ACP scientists face significant budgetary and time limitations; however ANL will continue its involvement for ground based PAN measurements and will also explore the characterization of a few size-fractionated aerosol samples for age dating and characterization.

5. Success Stories: Examples

A number of "success stories" were identified during the breakout session. One "success" noted by field scientists was the close collaboration between boundary layer meteorologists and chemists in the analysis of aircraft observations. This story was told in poster sessions Tuesday morning that described results of the analysis of mixing and chemistry during the Southern Oxidants Study, the use of new trajectory techniques to identify emission source regions, and changes to the profiles of reactive species associated with mixing processes. A video tape showing quite good comparison between simulated plumes calculated using Four dimensional data assimilation (FDDA) and the corresponding observations during the NARE '93 field campaigns showed the strength of ACP's interlaboratory and interdisciplinary research capability.

Another set of "success stories" came from ACP laboratory programs in support of field campaigns: among those mentioned in the breakout session were:

• Development and application of the Trace Atmospheric Gas Analyzer (TAGA) for measurements of organic acids, dimethylsulfide (DMS) and halogen compounds
• The development of a 3-channel airborne peroxide analyzer for H₂O₂, CH₃OOH (MHP) + HOCH₂OOH (HMHP), used during the SOS '95 and NARSTO '96 field campaigns
• Construction and testing of a luminol based PAN analyzer, suitable for airborne measurements
• In-situ hydrocarbon monitoring instrumentation, suitable for airborne measurements
• Development and application of airborne organic carbonyl measurement capabilities
• The continued refinement of airborne NO/NO₂/NO_y measurement capability

These successes were based on identification of a need by ACP field scientists and the development of the instrumentation to satisfy that need by their colleagues in the laboratory. Participants in the breakout session identified this type of collaborative effort as necessary to maintain and improve ACP's core competencies. Future laboratory studies were recommended to identify important heterogeneous atmospheric processes effecting oxidants and aerosols. As these processes and mechanisms are identified, the kinetic and product information obtained in the laboratory will be made available to the modelers and field scientists to aid in measurement campaigns. A series of workshops were suggested as a method to facilitate these interactions. Another area for laboratory-field campaign interaction would involve the identification of indicator or marker compounds which act to specifically measure tropospheric reaction pathways.

One interdisciplinary "success story" brought up by modelers was the use of sensitivity coefficients to examine a series of reactions within a chemical kinetics model developed to evaluate halogen measurements taken during the 1995 ACP field campaign, with additional work being done in the laboratory to identify reactants of chlorine with isoprene that could be measured in the field to serve as indicator species in evaluating the role of halogens on ozone chemistry.

These studies highlight a non-systematic and incomplete review of cooperation between a variety of research teams within ACP. They illustrate a multitude of interactions among our laboratory, field and modeling teams, and between ACP mathematicians, meteorologists, and chemists. Future progress in providing the base of knowledge needed by DOE to assess environmental problems will depend strongly on this continued cooperation, not only within ACP but within the broader environmental sciences community.