Recommended Supporting Programmatic and Infrastructure Priorities for Advancing Atmospheric Chemistry Research
Thus far, this report has discussed the importance of the field of atmospheric chemistry (Chapters 1–4) and recommended Priority Science Areas (PSAs) and more focused key gaps that the Committee believes will drive the field forward (see Chapter 5). In this chapter, the Committee discusses important practical recommendations over the next decade for enabling the research priorities described in Chapter 5. As per the statement of task, these recommendations are directed primarily to areas of interest for the Atmospheric Chemistry Program at the National Science Foundation (NSF). We take as a given that the funding of high-risk/high-reward projects by NSF is necessary to advance the field of atmospheric chemistry.
As with the scientific areas, the Committee gathered ideas from the atmospheric community for possible logistical and programmatic recommendations. In particular, in order to understand the trends that are currently occurring in the field of atmospheric chemistry and the research emphases of different federal groups supporting work in this area, the Committee requested information from five government agencies. The full results are presented in Chapter 6.1 and Appendix C.
During the Committee’s interactions with members of the atmospheric chemistry research community (see Appendix B), the great majority of participants indicated their appreciation for the NSF staff who have managed this area. There is a sense that proposals have generally been dealt with efficiently and fairly. Appreciation was expressed for the system of using a combination of independent peer review and Program Officers’ judgment, rather than relying heavily on panel reviews as is the case increasingly in other parts of NSF. Having a base of permanent staff, with some rotating visitors, has provided continuity of interaction with the community.
The Committee recognizes that an important fraction of the NSF-funded research involving atmospheric chemistry research occurs through support to the National
Center for Atmospheric Research (NCAR) from the Facilities Section of NSF. While the Statement of Task did not explicitly include an assessment of NCAR’s role, NCAR is part of the infrastructure of atmospheric chemistry in the United States and the peer-reviewed Atmospheric Chemistry Program at NSF relies on partnering with NCAR. We thus also comment on NCAR’s role in the field.
To enable the recommended science priorities in Chapter 5, support of various activities that include the development of “tools”; the collection, analysis, and archiving of data; and collaborations within atmospheric chemistry and with other communities are all important. These are shown schematically in Figure 6.1 and discussed in more detail below in Sections 6.2 through 6.5.
The field of atmospheric sciences as a whole has been growing over the past several decades. One way to illustrate this trend is by tracking the number of graduate students in the field. NSF reports that a range of 140–170 PhDs in Atmospheric Science and Meteorology are conferred each year in the United States and that this number is slightly increasing through time.1 Although it was not possible to discern what fraction of the total is atmospheric chemistry due to the interdisciplinary nature of the field, and this number does not capture the atmospheric chemists graduating with PhDs in other fields (e.g., Chemistry, Environmental Engineering), it is the sense of the Committee that atmospheric chemistry is also growing in terms of numbers of researchers. While workforce issues were not within the scope of the Statement of Task, the Committee notes that the engagement, recruitment, and retention of traditionally underrepresented and underserved groups continues to be an important responsibility of the atmospheric chemistry community. More detailed data gathering on the atmospheric chemistry–related PhD degrees awarded and the career paths of atmospheric chemistry PhDs would allow for better characterization of the challenges facing the field as a whole.
A related question is whether the resources for the field of atmospheric chemistry have also been growing. Recent trends in spending by the major atmospheric chemistry research funding agencies, including NSF, National Oceanic and Atmospheric Administration (NOAA), U.S. Environmental Protection Agency (EPA), Department of
1 In the 1990s, the number of PhDs awarded in atmospheric sciences and meteorology was approximately 145 annually, and in 2013 the number was 167 graduates. Information was obtained through NSF’s Survey of Doctorate Recipients under the Scientists and Engineers Statistical Data System (NSF SESTAT) Metadata Explorer: http://ncsedata.nsf.gov/metadataexplorer/metadataexplorer.html.
Energy (DOE), and National Aeronautics and Space Administration (NASA) are shown in Figure 6.2. Overall, most of the agency funding shown here has been relatively flat over the past decade (with the possible exception of DOE and NOAA Oceanic and Atmospheric Research, OAR), especially when inflation is taken into account (dashed lines in Figure 6.2 show funds adjusted to 2015 dollars2). For NSF, the total number of atmospheric chemistry research projects funded, the inflation-adjusted combined
2 The funding amounts from each agency for each year available were converted into 2015 dollars using the U.S. Inflation Calculator, based on Consumer Price Index data from the Bureau of Labor Statistics: http://www.usinflationcalculator.com.
The Committee also examined how support from the various agencies is divided among types of research approaches, including laboratory, theory, field, and modeling studies. Although there is often overlap of types of research within a given study, estimates over the past decade from the NSF Atmospheric Chemistry Program show that it has typically spent approximately half of its budget on field projects, 25 percent on lab studies, 15 percent on modeling, 5 percent on instrument development, and 5 percent on “other” projects (see Figure 6.3). The larger proportion used to support field work is likely due to the nature of field work, which generally is personnel intensive and requires resources for instrumentation and personnel deployment. Data from EPA and NOAA also indicate that the largest fractions of their support are spent on field studies, with variable smaller amounts on lab and modeling studies (see Figures C.3a and C.3b). For example, NOAA and EPA external funding appears to spend relatively more on modeling compared to laboratory studies.
In summary, the judgment of the Committee is that the field of atmospheric chemistry has been expanding for the past several decades, but the amount of funding has not increased substantially. This has put pressure on the field as a whole to do substantially more with less, while maintaining a balanced portfolio of research approaches required by the breadth of research described in Chapters 3 and 5. While the Committee does not recommend major changes to this balance now, the appropriate balance in the future will depend on the nature of the issues being addressed, which is likely to change with time. It is therefore important to have flexibility in the distribution of efforts and support, with the recognition that any initiatives to provide more support in one area will have negative effects in another unless additional support is found.
As discussed throughout this report, building toward a predictive capability for understanding the chemistry of the atmosphere requires contributions from laboratory studies, theory, field research, satellite measurements, and modeling approaches. And as described in the previous section, the current support from the federal agencies, including NSF, generally supports a balance of the various types of approaches. These are the fundamental tools of the field of atmospheric chemistry, and the Committee believes that NSF plays an essential role in fostering the development of the next generation of many of these tools. Contributions to this by industry and other government funding agencies could also be important.
Laboratory Experiments and Theoretical Studies
Many of the tools needed to advance laboratory studies overlap with those needed for field studies, such as development of new instrumental approaches discussed in the following section. However, some needed tools are sufficiently complex and hardware intensive that they will be useful, at least initially, primarily in laboratory situations
where power, size, and space are not as limited as in field deployments. Ultimately such techniques may be adapted to field projects to provide important insights into unrecognized atmospheric constituents and processes. Laboratory studies could also benefit from implementing new techniques as they arise from other areas such as surface, materials, and pharmaceutical sciences with the long-term goal of interrogating complex systems under atmospherically relevant conditions.
One example of needed techniques is that for the analysis and identification of individual organic species and their location in complex milieus such as secondary organic aerosol particles (SOA) and surface films, in real time without sample collection and workup that can introduce artifacts. Techniques to probe interfaces of soft materials such as SOA and bioparticles as well as species on solid surfaces such as dust particles or the built environment are also needed in order to develop accurate mechanisms and models that faithfully represent the fundamental underlying principles and explain and guide laboratory and field studies.
Computational chemistry has undergone remarkable advances in the last decade and is able to address based on first principles simpler model systems relevant to the atmosphere. There continues to be a need to develop theoretical approaches that can be applied in practical terms to systems that encompass thousands of chemical species and cover a range of particle sizes and complexity from several to hundreds of millions of molecules, and that encompass both physical and chemical processes. Theoretical advances, supported and evaluated by both laboratory and field investigations, are required and essential to improving predictive capability.
Instrument and Instrument Platform Development
Accurate and specific measurements of trace gases and particles are central to atmospheric chemistry understanding and to developing predictive capabilities that inform societal choices (see Chapter 4). Many of the advances in the field of atmospheric chemistry that have been made over the last decades as described in Chapter 3 have arisen largely because of the ability to measure more species in air and at ever decreasing concentrations. Identification of new species and lowering detection limits for known species are key to testing the applicability of results from laboratory and theoretical studies to ambient air as well as providing essential data for testing model predictions.
As discussed in Chapter 5, new analytical techniques, instruments, and instrument platforms are needed to support the PSAs of the next decade. For example, new analytical techniques are needed to identify sources and sinks of compounds and radical
species in both gas and condensed phases that cause detrimental health outcomes, and to quantify their concentrations and distributions (see PSA 4). Further, understanding the role of aerosol particles on climate and weather will require tools to characterize particle composition, morphology, and phase over a broad size range from nanometer to micrometer scales, preferably with the ability to differentiate bulk from the surface (see PSA 3). Similarly, understanding feedbacks between the atmosphere and the biogeochemistry of natural and managed ecosystems would benefit from the development of instrumentation and measurement strategies to quantify fluxes of many different gases and particles (see PSA 5).
The development of instruments and instrument platforms is a challenging area, since there is a wide range of spatial and temporal scales that need to be covered, with specific requirements depending on the science questions being addressed. Being able to make measurements in three dimensions of many different chemical species over varying temporal and spatial scales is key, including defining the vertical heterogeneity in composition. Concentrations can change significantly with altitude, resulting in non-linearities in the chemistry, for example the rate of the RO2 radical self-reactions, which varies with the square of the RO2 concentration, becomes competitive with the first order reactions of RO2 with NO as the latter concentrations decrease. In addition, sources and concentrations of both gases and particles are often inhomogeneous with altitude. Interpreting surface measurements correctly under conditions of strong vertical mixing also requires understanding these heterogeneities and nonlinearities. Similarly, testing model predictions would be greatly aided by the availability of atmospheric chemical observations in three dimensions over as wide an area and over many different times and meteorological conditions as possible. As a third example, the connection of atmospheric chemistry to issues of societal relevance often involves measurements of atmospheric constituents on scales relevant to a specific problem. In many cases, atmospheric composition is not directly measured at or close to the site of interest, but rather interpolated or extrapolated from fixed site monitors that can be located some distance away. The latter often have a limited suite of measurements and fixed sampling times and durations that may not be suited to addressing the specific scientific questions.
The development of new instrument platforms is also important. While the use of aircraft, balloons, and blimps has provided, and will continue to provide, important atmospheric chemistry data in three dimensions, by their nature they are expensive and generally limited to planned field campaigns carried out over limited spatial and temporal regimes. The rise of unmanned aerial vehicles (UAVs or “drones”) provides increased opportunities for atmospheric chemistry measurements (see Figure 6.4;
Everts and Davenport, 2016).3 They have the potential to cover a wider spatial region than fixed sites or towers, and provide a nimble capability that can be relatively easily moved to different locations. This would be particularly useful in the case of unforeseen events such as the Deepwater Horizon explosion and aftermath, or volcanic eruptions that require rapid response with a variety of instruments. Transmission of data to a central facility that provides quality control and standardization of formats, as well as conversion to user-friendly (including the public) formats, would help take full advantage of the data generated.
In this light, a key to taking full advantage of UAVs is the development of robust, light, small, accurate, and specific sensors that would provide data traceable to standardized methods. There are currently a number of excellent instruments that are now flown on aircraft, balloons, and blimps, but the use of UAVs will likely require lighter instruments with low power demands, and hence new approaches to measurement techniques. Furthermore, there is substantial interest in developing distributed sensor networks for estimating air pollution exposure for health studies (PSA 4). Advances in solid-state technology and materials science may be helpful in the development of such methodologies. The next step in the development of miniaturized instruments and instru-
ment platforms is vitally important and NSF could play a central role in fostering that development.
Low-power, lightweight sensors would also be useful in widespread networks, perhaps involving citizens. Although low-cost distributed sampling networks provide an exciting opportunity to characterize concentrations of gases and particles with high time resolution and fine spatial granularity, the quality of many existing approaches and datasets have not been thoroughly vetted. Even established techniques for measuring atmospheric trace species can be subject to a variety of uncertainties related to representativeness, precision, bias, detection limits, and accuracy. It is thus important that new methodologies are rigorously tested and the uncertainties in the data they generate clearly defined. Poor data quality hinders continued advancement in atmospheric chemistry and can misinform the public. Therefore, robust instrumentation with excellent precision is necessary for this testing and determination of uncertainties.
NSF currently provides instrument development support through several programs outside of its Atmospheric Chemistry Program, such as the Major Research Instrumentation (MRI) program, which has an instrument development component, and the NSF Small Business Innovation Research (SBIR) and Small Business Technology Transfer Research (STTR) programs. These programs have produced valuable breakthrough technologies that have enabled improvements in atmospheric chemistry research (NASEM, 2015; NRC, 2005b). However, there are limitations with respect to atmospheric chemistry needs. For example, only two proposals per campus are allowed to be submitted to the MRI program for larger universities, and these proposals cover all areas of science. SBIR topics have been described as being narrow as a means to pare down the number of applications (NASEM, 2015). As a result, there are generally few opportunities for high-risk, high-reward proposals for instrument development in atmospheric chemistry at NSF.
The Committee believes that there will be continuing instrument development needs related to atmospheric chemistry in the future. We encourage the Atmospheric Chemistry Program to consider mechanisms for providing more support for instrument development work. For example, the Atmospheric Chemistry Program at NSF could work more closely with other NSF programs, such as the MRI and SBIR programs as well as the Chemistry Division in the Mathematical and Physical Sciences Directorate at NSF and the Chemical, Bioengineering, Environmental, and Transport Systems Division in the Engineering Directorate, to ensure that atmospheric chemistry research needs are fully integrated into the priorities of these other programs. In addition, it is important that viable mechanisms be available within the Atmospheric and Geospace
Sciences Division (AGS) to submit proposals for new instruments and techniques that often take substantial amounts of time to adequately develop and test before they are ready for use in the field or lab. Many other federal agencies have historically been reluctant to fund such instrument development outside of their own laboratories unless there is a specific set of “deliverable” measurements by the end of the grant, which is typically 3 years. Given the challenges for instrument development, it may be that cross-directorate support or advancing a cross-directorate initiative, for example, with the Mathematical and Physical Sciences and/or Engineering Directorates, would be beneficial. Another potential mechanism might be collaboration with scientists at a national center (see Section 6.5).
There is a wide variety of available atmospheric chemistry modeling tools that range in spatial scale, temporal scale, and technical approach. The utilization of model type depends on the scientific application, and a diversity of approaches is needed to develop a broad toolbox to understand the complex problems described in Chapters 4 and 5. From the spatial perspective, tools range from point-based approaches such as box models and one-dimensional column models to three-dimensional models that range from very high resolution (on the order of meters; e.g., large eddy simulation models) to regional (e.g., continental scale) and global scales. From the temporal perspective, box models typically run on the scales of hours, whereas coupled chemistry-climate models run on the scale of decades to centuries. Methodological approaches also vary, including using Lagrangian plume models to fixed grid Eulerian models. Within fixed grid models, there are a number of new approaches that can help to answer key scientific questions. For example, data assimilation techniques can help integrate models and observations and improve predictability (e.g., Bocquet et al., 2015), adjoint model approaches can assess model sensitivities (Chai et al., 2006; Vautard et al., 2000), inverse modeling can improve understanding of sources and sinks (Kasibhatla et al., 2013), and reduced-form models can operate with reasonable verisimilitude and efficiency to be incorporated in socio-economic or decision-making tools (Kerl et al., 2015).
Modeling is the key element for developing a predictive capability for atmospheric chemistry and addressing the areas of priority science identified by the Committee in Chapter 5. For this predictive capability to develop, substantial advancements are required, including (1) acquiring more observations of chemical species across temporal and spatial scales for chemical data assimilation and model evaluation, as well as taking advantage of existing datasets (see below); (2) dedicating additional resources
to support the development of modeling software; and (3) continuing investment in high-performance computing (HPC) resources. New predictive modeling approaches can build on techniques developed by the numerical weather and seasonal to decadal forecasting communities. On shorter timescales (e.g., days), a key element influencing model skill is the organization and assimilation of observations for initial conditions. As time scales lengthen, the uncertainty in the knowledge of emissions becomes dominant. However, atmospheric chemistry models will have additional needs beyond what has been developed for the weather forecasting community, as the range of chemical lifetimes of important species provides a unique data assimilation problem.
Currently, the broad toolbox of modeling software is supported by a wide variety of federal agencies, including NSF. Despite substantial progress over the past two decades in model tools, continued investment in atmospheric chemistry model development is needed to make accurate, predictive atmospheric chemistry modeling a reality in addressing the interdisciplinary issues described in this report. However, similar to instrument development, model development faces substantial challenges in obtaining funding. A major obstacle for the atmospheric modeling community is the disparate spatial and temporal scales of atmospheric chemistry, and the resulting difficulty in building consistent modeling tools and methodological approaches that can work together and integrate across these scales. An additional and perhaps greater challenge is integrating models to develop a predictive capacity for atmospheric chemistry. NSF needs to continue its investments in model development and applications across scales, ranging from developing and incorporating theoretical chemistry to predicting global composition. There is a clear opportunity for NSF to initiate a new focus for modeling across scales to develop a predictive capacity for atmospheric chemistry. Such a research initiative could promote collaboration and coordination across agencies and across different NSF programs (e.g., theoretical chemistry and atmospheric chemistry, environmental engineering, and climate).
Currently, most three-dimensional numerical models used in atmospheric chemistry require some level of HPC, whether at a university center or a national center. Computing resource limitations and complexity of parallel architectures can present barriers for future research in atmospheric chemistry that more often than before requires HPC modeling. NSF provides the atmospheric chemistry community with vital HPC capability at NCAR, directly through HPC computing services at centers such as Yellowstone4 or indirectly through joint projects using the Community Atmosphere Model with Chemistry or the Whole Atmosphere Community Climate Model. Because atmospheric
chemistry plays a key role in climate, the urgency of these simulations needs to be understood as part of the Earth system community’s research into global change (Garcia et al., 2012; Mahlman et al., 1980; Pinto et al., 1983; Prather et al., 1987; Wild et al., 2003). In this context, NSF should ensure that principal investigators have the availability of resources to advance the science.
Summary of Development of Tools for Atmospheric Chemistry Research
The Atmospheric Chemistry Program at NSF is well positioned to lead efforts to improve the tools needed for laboratory and theory studies as well as for instrument and model development in collaboration with other directorates and programs within NSF and at other agencies. Overall, improved support for these basic tools of atmospheric chemistry research is essential for advancing the science over the next decade, and will allow the scientific community to observe, measure, and predict atmospheric chemistry processes and their interactions with other physical, biological, and human systems.
Recommendation 1: The National Science Foundation should ensure adequate support for the development of the tools necessary to accomplish the scientific goals for the atmospheric chemistry community, including the development of new laboratory and analytical instrumentation, measurement platforms, and modeling capabilities.
Data are an essential element of research in all fields of science, and the field of atmospheric chemistry is no exception. In particular, there are several issues related to the collection and analysis as well as sharing and communication of data that the Committee chose to focus on in this report. First, the collection of measurement data over long periods of time allows for the discernment of trends that are not possible to see in one-time field projects. Second, the answers to research questions are often apparent only after intensive data analysis or with the analysis of data from multiple projects. Third, the management of large volumes of “big data” is becoming more ubiquitous in atmospheric chemistry research (and scientific research more broadly) and mechanisms for effectively and efficiently archiving, sharing, and mining data, including making it available to the broad scientific community and the public, are needed.
Long-Term Research Sites
The Committee recognizes the central importance of long-term research sites for comprehensive atmospheric chemistry research. The need for such sites was also clearly identified by the 2001 NRC report Global Air Quality: An Imperative for Long-Term Observational Strategies (NRC, 2001). A classic example is the Mauna Loa dataset that clearly revealed long-term trends in CO2 many decades ago. Such sites are beneficial for addressing important scientific research questions, in particular those related to the spatial and temporal fluxes of elements between the atmosphere and terrestrial and marine ecosystems (see PSA 2) and understanding feedbacks between the atmosphere and natural and managed ecosystems (see PSA 5). For example, how does atmospheric chemistry modulate trends in the composition and bioavailability of nutrients and contaminants? What are the emissions from natural and anthropogenic sources across sectors and environments, and how do those emissions change with climate change, land use change, and other global changes? How efficiently do wet and dry deposition remove particles and trace gases from the atmosphere, and how are these changing?
The vision for these long-term sites is not simply to develop a network of monitoring stations, but rather to exploit research sites with core measurement capabilities and long-term knowledge about regional photochemistry, meteorology, ecosystem properties, and biosphere–atmosphere exchange processes that the atmospheric chemistry community can also use as a resource for making and interpreting new measurements. Long-term sites gain scientific value with time and provide a rich environment in which to understand changes in atmospheric chemistry driven by changing emissions, land use, climate, or other factors of societal relevance. Because the atmosphere is a global commons, it is important that sites be spatially distributed in a representative way, and that necessarily involves international cooperation and collaboration. For many scientific questions in atmospheric chemistry, research sites do not necessarily need to be created in new locations, in fact there would be great value in leveraging from existing monitoring locations. Indeed, there are several sites where such long-term research has enabled progress on important scientific questions and involve international cooperation by scientists as well as funding by governmental organizations from different countries and the UN World Meteorological Organization’s Global Atmospheric Watch.5 For example, NOAA GMD’s background monitoring sites (Mauna Loa, HI; American Samoa; Barrow, AK; the South Pole, etc.) were set up primarily for monitoring greenhouse gases, but they have been expanded to make many additional measurements and have been used extensively by non-NOAA scientists to conduct
process-oriented research.6 Another example is the AGAGE-NASA7 network which includes 11 sites measuring more than 50 gases with in situ instrumentation over the past 38 years. Beyond the existing sites of opportunity managed by international, federal, and state partners, scientific research using existing data and atmospheric modeling tools should be used to optimally site any additional measurement locations as defined by the needs of the atmospheric chemistry community.
Long-term sites have some distinct advantages over individual, one-time field projects. Developing field sites for specific campaigns is often necessary to address critical regional and/or unique problems. However, setting up new field sites for specific campaigns is typically a substantial fraction of the cost and effort associated with the experiment, and it requires a large-scale coordinated effort and commitment by the research community and funding agencies. Large-scale individual field campaigns are not always the most cost-effective approach for making scientific progress. Furthermore, it is often difficult to place the results from a single field project in a larger context unless it is built around longer-term observations and understanding. Supporting long-term sites and encouraging proposals for research at those sites enables regular introduction of new observational approaches, testing and inter-comparison of new measurement capabilities, and evaluation of new scientific understanding of emissions, transformation mechanisms, and deposition, without the need to develop novel infrastructure for costly field campaigns as frequently as is often done in the atmospheric chemistry community.
Specifically for issues of understanding biosphere–atmosphere exchange and its evolution, the atmospheric chemistry community generally lacks long-term infrastructure in an appropriate array of representative environments in the United States. NSF and other U.S. scientific agencies have developed long-term sites to understand ecological processes and the carbon cycle (e.g., NSF NEON, DOE AmeriFlux, NSF Long-Term Ecological Research [LTER]). There is potential synergy to be gained by developing infrastructure in collaboration with one or more of these existing networks, or specific long-term sites with a history of atmospheric chemistry research on which additional research facilities could be built. If sites were needed in additional locations, they may provide an appropriate focus for a recently announced initiative at NSF for mid-scale infrastructure.8 Natural emissions and deposition, and therefore atmospheric chemistry and composition, are highly dependent on phenology, meteorology, season, climate, and anthropogenic air pollution, and are extremely variable across different
ecosystems. Thus a distributed set of research sites within the context of existing knowledge and infrastructure that relates to current and future priorities in atmospheric chemistry research would be most cost-effective.
With several different agencies currently supporting various kinds of long-term sites, it would be most useful if representatives from the various agencies worked in collaboration with academic researchers to determine how existing sites could be more useful to the atmospheric chemistry research community and to decide if, where, and how many new sites might be needed. An interagency panel could help prioritize the long-term sites and determine the required infrastructure, whether the sites would be centrally managed (e.g., using resources associated with NCAR) or managed by individual principal investigators (PIs) (like AmeriFlux or LTER sites), core measurements to be included with each site, procedures for the archiving of the samples collected at these sites, and criteria and a review process to support funding decisions. Further, this panel can investigate the potential to contribute to international programs, such as the Global Atmospheric Watch program of the United Nations World Meteorological Organization, that focus on long-term observations of atmospheric chemistry around the globe.
If existing infrastructure can be used to form the backbone of a network of long-term research sites, the cost to NSF could be reasonable. In addition, NSF would be able to individually support research at these long-term sites in response to proposals from single or small groups of PIs. It could be beneficial to fund such proposals for longer than the standard 3-year period that most awards are given.
Recommendation 2: The National Science Foundation should take the lead in coordinating with other agencies to identify the scientific need for long-term measurements and to establish synergies with existing sites that could provide core support for long-term atmospheric chemistry measurements, including biosphere–atmosphere exchange of trace gases and aerosol particles.
Resources for Data Analysis
Many observations of atmospheric chemical composition and processes, especially those from measurement networks and field campaigns, are not used to their full potential. Scientists historically have obtained 3 years of funding at a time, which often allows for collection of the data and publication of the most obvious trends and findings. Often the funding is insufficient to mine the data deeply for thorough analysis. In addition, questions raised by future discoveries may be answered with reanalysis of existing datasets.
Although funded research endeavors have produced important results, data streams created from these efforts present a rich resource that could be used for additional analysis to help answer evolving atmospheric chemistry and composition questions and to provide guidance for future field studies. There is considerable opportunity to exploit the treasure trove of existing measurements that represent the collective atmospheric chemistry knowledge from decades of investment in atmospheric chemistry research. For example, existing high-quality atmospheric chemistry field and environmental chamber datasets could be mined to examine chemical mechanisms, to look for trends in time, and to compare the chemistry across different chemical and dynamical regimes (see PSA 1). In addition, data mining of measurements (e.g., concentration ratios) can help constrain emissions and trends (see PSAs 2 and 5). The synthesis and analysis of existing datasets can be applied to guide future research directions and help to test models across various regimes. All previous datasets are not equally valuable, and new field campaigns that take advantage of the rapid advances in measurement technology can provide scientific information that supersedes anything already available. Nor is the analysis of previous observations a replacement for new field campaigns to address contemporary research questions. However, in the face of tighter budgets, taking advantage of existing datasets for new analysis may be a cost-effective way to continue to advance the atmospheric chemistry research agenda.
Funding is required for researchers to perform analysis on collected datasets and for collaborations that take knowledge from the lab or field into model applications. Longer grant periods or perhaps supplemental installments to afford PIs the time and effort to continue analyses may be needed to accomplish these efforts. NSF should also encourage and support new projects that use data mining to advance the science. For a fraction of the cost of another field study, NSF could dedicate some amount of funds to encourage atmospheric chemists, possibly in collaboration with computer scientists, to mine data from previous studies, performing detailed intercomparisons with satellites, measurements, and models.
Recommendation 3: The National Science Foundation should encourage mining and integration of measurements and model results that can merge and exploit past datasets to provide insight into atmospheric processes, as well as guide planning for future studies.
Data Management and “Big Data”
Establishing a predictive understanding of chemical processes in the atmosphere and their impacts as described in this report has been possible to date only through large investments in laboratory and field observations, theory, and modeling. During this process, vast and multidimensional datasets are continuously generated, particularly in field studies that require increasingly larger resources to manage. Data need to be archived in suitable formats to be useable. In particular, data archives from field measurements in atmospheric chemistry are a fundamentally important resource for the atmospheric chemistry and broader scientific and regulatory communities, and ultimately for the public. They serve as benchmarks for models, provide a rich set of observational constraints to understand atmospheric chemical processes, document changes in composition due to both natural and anthropogenic activities, and assess societally relevant impacts such as human and agricultural exposure.
Currently, the availability of these datasets varies substantially. Some are archived at data centers (with online or offline access) while others are available only upon request from individual scientists. A crucial part of facilitating the comparison and synthesis of datasets is to ensure that there is a common data formatting that allows integration between models and measurements, especially by scientists who were involved in the studies, but also by those who were not. However, the format of much of the currently archived data and the information content associated with them can vary considerably. Critical information, including expert interpretation and data quality aspects, are often not documented sufficiently. Model simulation outputs of importance for atmospheric chemistry research are often not archived or available to the general community. Other datasets generated from models or post-processing of raw data may involve codes that are insufficiently archived or documented. This diversity, together with the increasing volume and complexity of data poses great challenges for research in atmospheric chemistry.
Part of this variability in the availability of datasets is due to a lack of coordination among the federal agencies. Archives of atmospheric chemistry field campaigns and long-term measurements from research and regulatory efforts are currently maintained separately by multiple federal agencies including DOE, EPA, NASA, NCAR, NOAA, and statewide agencies (e.g., California Air Resources Board, among others), as well as private, nonprofit, and citizen scientist organizations. The atmospheric chemistry community could leverage some of these sources of data, depending on data quality. Atmospheric chemistry–related data archives are not currently coordinated at a national level between different agencies that support or conduct research and measurements in this area.
Syntheses of these large and diverse datasets may provide opportunities for breakthroughs and transformative science over the next decade. It is therefore vitally important that an effective and visionary management approach is established to facilitate the maximum use and impact of datasets, in particular those generated under NSF support. The scientific community and NSF at large have long been aware of the need for effective data management and have established guidance principles and requirements for each research PI to follow toward that goal. However, much of the support and management of datasets still rely on individual research teams, even long after funding has expired.
NSF already engages in data management efforts. Federally funded research through NSF does require data management plans, but no central coordinated data archive and sharing system exists to serve as a repository and resource for the atmospheric chemistry community and other users who need data for related societally relevant work. The use of common cyber-based infrastructures for handling and managing diverse data and facilitating information extraction and knowledge discovery could provide part of the solution to the data management problem. One such initiative established within NSF, joint between the Directorate for Geosciences (GEO) and the Division of Advanced Cyberinfrastructure (ACI), is EarthCube. EarthCube was initiated in 2011 with anticipated support until at least 2022; it is currently a community-based initiative where members can “influence how data will be collected, accessed, analyzed, visualized, shared, and archived; facilitate and participate in interdisciplinary research; and help educate scientists in the emerging practices of digital scholarship, data and software stewardship and open science” (Gil et al., 2014).9
NSF should require that future NSF-funded datasets be handled in a manner that allows ready access and comparison with previous datasets. The Committee envisions a centralized system for providing and supporting data management for atmospheric chemistry. Apart from providing facilities for data archiving, accessibility, and transparency, a centralized responsibility could assure that datasets are managed with expert preparation and fostering, accompanied with sufficient and standard documentation and metadata (such as codes for generating derivative products and links to source data). It would also ensure longevity in the storage and accessibility of datasets. A centralized system can ensure that its data resources are coordinated with other data management initiatives, such as EarthCube, and also ensure that atmospheric chemistry is an integrated discipline within these efforts. Management and financial support for such a system would have to be integrated into long-term planning and resource allocation.
Another possible end goal of these efforts would be an archiving tool that would provide a user-friendly interface for accessing and manipulating (e.g., plotting) global data. This could enable the design of future field campaigns to fill the gaps identified in the more detailed analysis and comparison efforts, and also provide a highly visible outreach activity directed to the public at all levels.
Recommendation 4: The National Science Foundation (NSF) should establish a data archiving system for NSF-supported atmospheric chemistry research and take the lead in coordinating with other federal and possibly state agencies to create a comprehensive, compatible, and accessible data archive system.
The scope of problems in atmospheric chemistry that need to be addressed is broad, from those that are suitable for study by a single PI to those that require a much larger breadth and depth of expertise than a single PI or single community has. Addressing more complex problems may optimally be undertaken by collaborations that are of different sizes and include scientists from a mixture of academia, government, and the private sector. This will clearly require working across disciplines and thus across directorates at NSF, across government agencies, and across international boundaries.
Developing the comprehensive understanding of the Earth system that will allow enhanced predictive capability relies on close integration of knowledge from multiple disciplines or approaches across the Earth sciences. In addition, many of the important problems in atmospheric chemistry require interdisciplinary solutions. Atmospheric chemists need to engage with scientists from other disciplines to solve problems at these disciplinary interfaces, such as collaborations between chemists and cloud physicists to determine aerosol particle effects on clouds and radiation or chemists and toxicologists to better understand the health impacts of particulate exposure. Interdisciplinary work will be integral in addressing many of the scientific questions related to atmospheric chemistry and climate/weather (see PSA 3), human health (see PSA 4), and exchange between the atmosphere and natural and managed ecosystems (see PSA 5). For example, cooperative efforts will be required between atmospheric chemists and physical and chemical oceanographers as well as ecologists to understand the fluxes and impacts of the air/sea exchange of chemicals.
Atmospheric chemistry should also play important collaborative roles where other disciplines lead. For example, joint studies will be necessary between atmospheric chemists and the health science community to understand the key species that need to be measured and on what time and spatial scales to address the needs of the epidemiology and toxicology communities. Collaborations with specific industries or regional air-quality managers can provide better characterizations of emission fluxes to the atmosphere. To contribute to the solution of problems that span multiple disciplines, atmospheric chemists need to (1) identify the science within atmospheric chemistry required to address a particular problem and (2) build collaborations with the other disciplines whose knowledge is required. Each scientific component required to address the problem needs to have either a disciplinary or cross-disciplinary “home.” A collective assumption that a particular component is handled elsewhere may lead to persistent gaps at critical interfaces. The Committee identifies here the scientific priorities within the field of atmospheric chemistry as well as those related to disciplines with which collaboration is required.
Ultimately, the predictive capabilities developed by the atmospheric chemistry community will be used by others such as economists and social scientists in order to translate the science into policy assessments. Included in this discussion of interdisciplinary work is the need to acknowledge and reward integration within a discipline in addition to obvious cross-disciplinary work. Examples are scales of modeling from process-level to global-level, or connection of laboratory, theory, field, and model studies.
NSF sponsors a broad range of research, including social and behavioral studies that provide potential partnerships with the atmospheric chemistry research community to address environmental degradation and human impacts on community scales. The Committee commends NSF for initiatives that motivate collaboration among disciplines and NSF directorates, such as Science and Technology Centers, and programs such as the Dynamics of Coupled Natural and Human Systems. However, the Committee is concerned that mechanisms to support interdisciplinary work submitted outside such directed and relatively short-term vehicles, which have specific goals and structures, may encounter barriers due to NSF institutional and review structures. There are emerging examples of cross-directorate and interagency investment areas such as the SEES (Science, Engineering and Education for Sustainability) and INFEWS (Innovations at the Nexus of Food, Energy and Water Systems) initiatives. It is noteworthy that the INFEWS initiative “enables interagency cooperation . . . and allows the partner agencies—National Science Foundation (NSF) and the U.S. Department of Agriculture National Institute of Food and Agriculture (USDA/NIFA) and others—to combine resources to identify and fund the most meritorious and highest-impact projects that
support their respective missions, while eliminating duplication of effort and fostering collaboration between agencies and the investigators they support.”10 The Committee encourages exploration of the possibility of such interagency collaborations.
Funding larger interdisciplinary collaborations is often too expensive for individual programs, such as atmospheric chemistry, so program managers need to leverage their funds by partnering with other directorates in NSF or at other funding agencies. Another challenge is that interdisciplinary research often requires sustained long-term funding, which can be difficult to achieve using a series of the typical 3-year NSF grants. In some NSF directorates, an effective approach has been to fund, in 5-year increments, centers that draw together scientists who have different expertise and are often geographically dispersed. Such centers could be real, on-the-ground centers, or they could be virtual centers that connect multiple groups to work on a common problem.
Other options for NSF include developing funding mechanisms whereby small, focused teams can integrate expertise defined by the team across multiple disciplines, which would allow an investigator-driven, creative, and competitive approach to integration. In addition, NSF could provide funding that supports short-term training of excellent young researchers in another research group with a very different focus. NSF could also identify and alter incentive structures that discourage integration across disciplines. These factors may include the lack of rewards for individual programs to fund work that crosses boundaries, with the appearance of a “zero-sum game” in which resources devoted to interdisciplinary work are perceived as flowing outside the sphere of influence of a particular program. Fixed grant sizes and reviewer pools that evaluate an entire proposal from the perspective of a single discipline can also discourage inclusion of multiple disciplines. Finally, NSF could encourage a “cross-disciplinary integration” component of proposal evaluation, thereby acknowledging the potential and substantial contributions that lie between the traditional ranking categories of “intellectual merit” and “broader impacts.”
Overall, the Committee believes that NSF should explore these various options to develop mechanisms that facilitate integration of expertise across disciplines and across academia, institutes, government, and industry. These mechanisms need to allow an investigator-driven, creative, and competitive approach that can complement the funding of individual PI research. The Committee further encourages NSF’s Atmospheric Chemistry Program to help lead these interdisciplinary efforts to address
10 Innovations at the Nexus of Food, Energy, and Water Systems: http://www.nsf.gov/pubs/2016/nsf16524/nsf16524.htm.
problems that better integrate research within the atmospheric chemistry community and help connect atmospheric chemistry research with other disciplines.
In addition to working across a single agency, cross agency efforts are particularly important, and NSF assistance in obtaining support for partners may be needed when funding from a single program or even research agency is not possible. There are existing mechanisms for coordination and collaboration among the various federal agencies, including the Air Quality Research Subcommittee, which is an interagency subcommittee of the National Science and Technology Council in the White House. Interagency planning needs to first identify the topical components required for progress, and then agree upon agency homes for each component, including knowledge gaps that have persistently hindered advances. The overall goal should be to incentivize close coupling of knowledge from multiple disciplines and approaches across government agencies. Cross-agency integration should focus on identifying and removing barriers in targeted subjects such as planning for long-term research sites (Recommendation 2), effective management of data archives (Recommendation 4), or broad research themes as discussed below. In contrast, persistent meetings that serve primarily administrative purposes are not encouraged.
As examples of interagency evaluation, NSF could more proactively partner with agencies like the National Institutes of Health (NIH; including the National Institute of Environmental Health Sciences [NIEHS]) and the U.S. Environmental Protection Agency (EPA) on cross-agency initiatives, such as centers focusing on the connection between air quality and health. As discussed in PSA 4, the atmospheric chemistry community can bring important research to this inherently interdisciplinary topic including the chemical characterization of air pollution that leads to adverse health effects, improved quantification of exposures to pollutants, and the characterization of indoor environments. Closer connections of the atmospheric chemistry community with the epidemiology and toxicology communities could allow for enhanced progress. Examples of previous interdisciplinary work in this area show the enormous potential for progress. For example, meningitis outbreaks in sub-Saharan Africa have been tied to local weather, air quality and pollution sources, and climatic conditions, suggesting complex interactions between the atmosphere and human impacts (e.g., Deroubaix et al., 2013; Dukic et al., 2012; Hodgson et al., 2001).
Support of cross-cutting scientific research will require participation of multiple groups, such as from different directorates within NSF and/or a variety of government agencies (e.g., Centers for Disease Control and Prevention [CDC]), DOE, EPA, NASA, NIH, NOAA, as well as cooperation with international entities. Additionally, given the societal relevance of many areas of atmospheric chemistry, the Committee believes that
it is important to consider public engagement with atmospheric chemistry research. Examples of public engagement include participation in atmospheric chemistry measurements, use of air quality indicators, and making air quality data readily accessible in a user-friendly format.
Recommendation 5: The National Science Foundation should improve opportunities that encourage interdisciplinary work in atmospheric chemistry and facilitate integration of expertise across disciplines and across academia, institutes, government, and industry. This improvement may include support of focused teams and virtual or physical centers of sizes appropriate to the problem at hand.
Capacity Building and International Collaboration
Many of the scientific challenges discussed in Chapters 3 and 5 are global in nature and affect individuals from a wide array of backgrounds. Data from a diversity of regimes or representative environments within the United States and around the globe are critical to addressing these challenges, from quantifying the flow of atmospheric carbon and nitrogen between ecosystems and the atmosphere to understanding aerosol particle properties and precursors to unravel anthropogenic influences on climate.
The international atmospheric chemistry community is strong and committed to developing a global understanding of atmospheric chemistry and its impacts on human activities. This has become increasingly essential as the importance of long-range transport is becoming evident. International cooperative efforts are necessary to make major progress in attaining the objectives of all the Priority Science Areas described previously. For example, working with the international community will help to identify locations and develop the sites for long-term measurement programs for many global scale studies (Recommendation 2). These range from individual PI–PI interactions to large-scale international research programs (see Box 6.1).
The success in understanding and applying atmospheric chemistry to improve air quality in the United States has effectively translated into success in improving air quality in some other regions, such as Mexico City. However, this success does not extend to many other areas such as parts of Africa, where indoor and urban air quality remain poor and the depth of air quality problems is essentially unknown. Even some underserved areas within the United States may be poorly characterized. While the basic principles of atmospheric chemistry and its implications for air quality in well-studied regions in the United States and other developed countries also apply in
principle to other situations, each urban area has unique environmental conditions, meteorology, and emissions. As a result, cost-effective solutions to poor air quality vary among populated regions around the globe. Scientists from developed countries have studied the atmospheric chemistry of some regions, but for many regions the understanding of the regional atmospheric chemistry is too poor to devise cost-effective
solutions. A more sustainable approach is to build the expert human capacity and observational and modeling capability at the regional level through collaborations between U.S. atmospheric chemists and scientists who focus on these regions. This approach also has the greatest chance of improving the air quality in these regions because regional governments are more likely to listen to their own atmospheric chemists than to U.S. scientists, and communities are more likely to trust and communicate with individuals who are familiar with their situations.
A complementary approach to building global atmospheric chemistry capacity is supporting peer-to-peer relationships between scientists from the United States and those in developing countries to advance and implement measurement or modeling plans in developing countries. One possibility is establishing network sites for global scientific measurements, such as the AERONET (AErosol RObotic NETwork) program,11 in developing countries or taking advantage of the Global Atmosphere Watch (GAW)12 and its Integrated Global Atmospheric Chemistry Observations (IGACO) strategy under the World Meteorological Organization. Another is the transfer of measurement and modeling capability that is common in the United States for assessing air quality to scientists in developing nations where such capability is uncommon, recognizing that this may be determined in part by export controls and other restrictions. While NSF may be less prone to such restrictions, there is a need to develop a strategy to support collaborative relationships in the face of laws in both the United States and abroad that have the potential to hinder progress. Such activities both within NSF and across multiple U.S. agencies are very important for fostering global research programs and building global capacity in atmospheric chemistry. These collaborative plans should include joint participation in maintenance, quality control, and data analysis in addition to implementation of the tools, and they can involve training students from both countries. Maintaining this capacity beyond the initial development phase is vital in producing long-term observations and modeling datasets in regions for which the air quality and atmospheric composition can be rapidly changing.
NSF already does a substantial amount to promote international collaborations, for example, through the foundation-wide PIRE (Partnerships for International Research and Education) program or the USAID PEER (Partnerships for Engaged Enhancements in Research) program. PIRE supports international partnerships to address important science and engineering problems that require international collaboration, while helping to develop an internationally-engaged workforce. PEER grants are developed
12 GAW is considered the atmospheric chemistry component of the Global Climate Observing System (GCOS); http://www.wmo.int/pages/prog/arep/gaw/gaw_home_en.html.
by scientists in developing countries to propose research with NSF-funded scientists to support research and capacity-building activities on topics with strong potential development impacts.
International cooperative activities include participation in existing and emerging international coalitions, frequent international exchange and training of students, peer-to-peer collaborations between scientists in developing countries and the United States, and sharing of atmospheric chemistry expertise from the United States with peer scientists in developing countries.
A recent NSF Geosciences solicitation states that the geosciences “continue to lag other science, technology, engineering, and mathematics disciplines in the engagement, recruitment and retention of traditionally underrepresented and underserved minorities.”13 Engaging U.S. researchers with a diverse range of backgrounds could foster innovation by incorporating a greater understanding of local contexts into the workforce. NSF is encouraged to support activities designed to engage underrepresented groups, either through nontraditional partnerships or through recruiting and retention of minority individuals.
Recommendation 6: The National Science Foundation, in coordination with other agencies, should continue to encourage and support U.S. scientists involved in atmospheric chemistry research to engage with underserved groups, in capacity-building activities, and in international collaborations.
The scope of the scientific problems facing atmospheric chemistry and the rest of atmospheric science are broad. As a result, the research required to solve these problems ranges from single PI investigations to large, multi-investigator, multidiscipline collaborations. In addition, the required observational research capability ranges from individual PIs’ instruments in their laboratories to sophisticated field deployable research platforms, such as aircraft, containing several instruments that are engineered to meet the high standards required for field deployment. Few individual PIs have the resources or expertise to develop, maintain, and operate an array of instruments and platforms over a long term that is required to answer today’s (and tomorrow’s) science questions. Similarly, few individual PIs are able to develop complex weather, chemistry, and climate models or to have access to high performance computational resources
on which to run these models. Yet without these research platforms, models, and computational resources, the science described in this report cannot be accomplished.
A national center can be an optimal approach for providing these more complex and costly observational and computational capabilities because (1) dedicated center staff with expertise are most efficient at maintaining these complex capabilities; (2) NSF competitive processes are suitable for making support of these abilities available to individual or groups of PIs; and (3) the center can help foster the collaborative research needed to identify and solve critical science and societal problems. A national center can provide these resources and expertise to investigators in the broader community, primarily those served directly by NSF, while at the same time contributing directly to scientific advancements. This service to the community can be more than providing upon request atmospheric measurement capabilities, Earth system models, laboratory tools, or technical and scientific expertise that are difficult to maintain within universities or private companies. It can also be attracting and retaining intellectually leading scientists to the national center who are fully engaged with university faculty and their students by contributing to their research as well as involving them in research initiated within the center. At the same time, these center scientists can have opportunities to pursue their own research directions part of the time, and thus maintain their intellectual contributions at the cutting edge of the field. In this ideal center, service and science leadership are tightly connected. Rewarding staff scientists for their service to the atmospheric science community as much as their scientific and technical excellence provides visible recognition of that connection. In addition, a national center can bring the community together—by hosting a large steady flow of visitors of all stages in their careers, facilitating collaborations, holding conferences and workshops, and playing a leading role in the production of scientific assessments. This kind of partnership between the broader community and a national center that has essential, unique capabilities can be a critical component for advancing atmospheric chemistry and addressing the scientific gaps that are exposed in this report.
The National Center for Atmospheric Research (NCAR) is currently providing some of these capabilities that are needed from a national center. NCAR was established as a federally funded national center dedicated to achieving excellence in atmospheric science research (UCAR, 1959). NCAR’s mission is “to understand the behavior of the atmosphere and related Earth and Geospace systems; to support, enhance, and extend the capabilities of the university community and the broader scientific community, nationally and internationally; and to foster the transfer of knowledge and technology for the betterment of life on Earth.”14
Atmospheric chemistry is an essential part of understanding the behavior of the atmosphere, so that a national center for the atmospheric sciences such as NCAR must include a vibrant program in atmospheric chemistry research. Atmospheric chemistry research occurs within many divisions of NCAR, especially the Earth Observing Laboratory (EOL) and the Atmospheric Chemistry Observations and Modeling (ACOM) Laboratory. EOL has aircraft and ground systems for field studies and expertise in using these systems and interpreting their data. The Committee notes that an important fraction of the NSF-funded atmospheric chemistry research relies on NSF facilities administered at NCAR. NCAR’s computational capability has supported many in the NSF Atmospheric Chemistry Program. ACOM has strengths in modeling across scales, in situ and satellite observations, instrumentation, and laboratory kinetics. Scientists within ACOM and its predecessor, the Atmospheric Chemistry Division, have historically provided prominent leadership roles within the broader atmospheric chemistry community and have been an integral part of the science research successes achieved in this field over the past five decades. The ACOM Advisory Committee15 provides input from the broader academic community to NCAR. The atmospheric chemistry community regularly questions how well ACOM (as well as its predecessors) has been fulfilling this complete vision of its role.
Many on the Committee have observed that ACOM’s capabilities have diminished in the past decade or so—including departures of some prominent atmospheric chemists. This situation occurred at the same time that NCAR ACOM scientists have been pushed to provide instruments, measurements, laboratory studies, and models for numerous projects, reducing the time for them to pursue their own research and/or development interests.
Nevertheless, the Committee believes that NCAR can be an even stronger partner with the atmospheric chemistry community by continuing to move its strategic vision closer to the Committee’s vision of the roles of a national center given above and the original founding charter of NCAR. Some steps have been made within NCAR and ACOM in this direction, but in order to be the partner that the atmospheric chemistry community needs, NCAR must find its unique role in atmospheric chemistry research, one that complements and enhances the research being conducted by the broader atmospheric chemistry community and engages individual PIs from universities, federal labs, and the private sector. Visionary intellectual scientific leadership and a strategic allocation of resources will be needed from within NCAR to bring about necessary changes to fulfill this role.
In summary, the partnership between the competitively funded NSF Atmospheric Chemistry Program and the facility-funded NSF programs at NCAR needs to be improved; doing so would facilitate greater scientific advances to close the science gaps presented in this report. As such, NSF and NCAR will need to work together to develop a strategy for improving this partnership and making NCAR a vibrant and complementary partner within the atmospheric chemistry community as a whole.
Recommendation 7: The National Center for Atmospheric Research (NCAR), in conjunction with the National Science Foundation (NSF), should develop and implement a strategy to make NCAR a vibrant and complementary partner within the atmospheric chemistry community. This strategy should ensure that scientific leadership at NCAR has the latitude to set an energizing vision with appropriate personnel, infrastructure, and allocation of resources; and that the research capabilities and facilities at NCAR serve a unique and essential role to the NSF atmospheric chemistry community.