Benefits of Atmospheric Composition Monitoring and International Data Exchange

Atmospheric composition, and changes therein, have multiple impacts on our lives and the environment. For instance, rising greenhouse gas concentrations cause global warming that intensifies weather extremes and drives ocean acidification. Increasing levels of air pollution are a threat to human health, ecosystems and agricultural production. To understand the state of the air we breathe, changes therein, the resulting impacts and the responsible drivers, observations of atmospheric composition are indispensable and so is the open exchange of that data across all sectors. The new WMO Unified Data Policy is expected to help further strengthen and broaden this exchange. The Policy text includes, for the first time, atmospheric composition data as an essential discipline area for WMO activities and establishes an organizational policy for their exchange. It also clearly recognizes the symbiotic nature of research and operations and the mutually beneficial data exchange between the two communities.

The WMO Global Atmosphere Watch (GAW) assists Member States and Territories in observing atmospheric composition and sharing observational data. However, atmospheric composition data is produced by  various agencies within and outside National Meteorological and Hydrological Services (NMHSs), including national and sub-national environmental protection agencies, academia and the private sector. Hence atmospheric composition data sharing occurs far beyond the WMO Community, thus the new WMO data policy is of great interest and relevance to GAW.

To reach the goal of the Paris Agreement to limit global warming to well below 2 °C, many countries have pledged to move towards net-zero greenhouse gas emissions. Without access to and exchange of atmospheric observational data, we would not know about the increasing greenhouse gas concentrations since industrialization and we would not be able to track future progress or identify emission hotspots to take action.

Ozonesonde (Source KMA).png
Figure 1 - Ozonesonde observation at Pohang (GAW ID: POH). Ozonesondes measure ozone and meteorological variables at different altitudes as the sonde ascends, until the balloon bursts. (Source: KMA)

Important applications for atmospheric composition data

Long-term atmospheric measurements are important to inform and support policy – and to demonstrate ultimately the success of any measures taken. For example, long-term atmospheric composition data attest to the beginning of a recovery of the ozone layer – an environmental success story. Stratospheric ozone depletion was among the environmental problems that led to signing in 1987 of the Montreal Protocol to phase out ozone depleting substances. The success of the treaty can be seen through the measured recovery of the ozone layer at a rate of 1–3% per decade since 2000 in upper stratosphere regions outside the poles (WMO, 2018a). Observations of ozone depleting substances, stratospheric ozone and UltraViolet (UV) radiation provide observational evidence to support the Protocol. Observations are made using diverse techniques and instruments from the ground and from the space. Vertical ozone profile are measured using ozonesonde.  An example of ozonesonde in preparation for launch is shown in Figure 1. Analysis of the chlorofluorocarbon CFC-11 long-term and quality-controlled observations indicated a slowdown in the decline of the atmospheric concentration after 2012, connecting it to an increase in global emissions from eastern Asia (Montzka et al., 2018; WMO, 2018b).

Greenhouse gas concentrations are also well-documented through long-term measurements around the globe. Global analysis of these observations, presented in the annual WMO Greenhouse Gas Bulletin (see Figure 2), shows that carbon dioxide (CO2) passed the 400 parts per million (ppm) level throughout GAW stations in the northern hemisphere in 2014. The Bulletin reported in 2016 when remote locations in the southern hemisphere, such as the Cape Grim GAW Global station in Tasmania, also breached that mark. In 1989, when GAW was initiated, the global mean CO2 concentration was 353 ppm.

CO2 Fraction mole.pngFigure 2 - Globally averaged CO2 mole fraction. The red line is the monthly mean with the seasonal variation removed; the blue dots and blue line depict the monthly averages. Observations from 133 stations were used for this analysis. The data is available, and the analysis is carried out by the World Data Centre for Greenhouse Gases, operated by the Japan Meteorological Agency (JMA). (Source: WMO Greenhouse Gas Bulletin 16, 2020)

Atmospheric pollutants (aerosols and reactive gases) are responsible for poor air quality, which causes an estimated seven million premature deaths globally every year (World Health Organization (WHO), 2016). Data on aerosols and reactive gases are important to determine acute health threats and are used in the estimates of the Global Burden of Disease (Shaddick et al., 2021). Observations are also used to monitor compliance with air quality regulations and to track changes in the abundance of pollutants in response to policy (UNECE, 2016). Delivering such data in near real time is crucial for improved forecast accuracy for early warning systems and to guide mitigation measures.

Ozone damage to crops.png

Figure 3 - Ozone damage to crops. Damage increases when ozone exposure continues: initially, the level of damage is small (left), then symptoms get worse (centre and right) (Source: K Sharps, ICP Vegetation)

In addition to its health implications, air pollution has a substantial impact on agriculture due to excessive deposition of nitrogen and sulfur constituents and ozone. Surface ozone is one of the main air pollutants affecting crop yields, with global losses for staple crops (wheat, rice, maize and soybean) estimated to be in the 3–16% range, or US$ 14–26 billion annually (Avnery et al., 2011; Mills et al., 2018). The mechanisms whereby ozone affects plants and crops are qualitatively well understood but poorly quantified. Figure 3 illustrates crop damage due to ozone.

A number of short-lived pollutants also have climate impacts, for example, ozone and aerosol. Among the different effects, they contribute to radiative forcing. For example, wildfire smoke aerosols impact on radiation and, thereby, the weather forecast (demonstrated in the WMO Aerosol Bulletin (WMO, 2021b)). To increase our understanding of the different effects they have on the climate system, observations of short-lived pollutants are also of utmost importance.

Atmospheric composition data sources and requirements

As stated earlier, atmospheric composition data are produced within and outside NMHSs. Normally observations of regulated pollutants are made by environmental protection agencies. High-quality observations for research purposes, including time-limited measurement campaigns, are made by research institutions and universities. Ground-based in situ measurements are complemented by aircraft-based in situ measurements (e.g. IAGOS), as well as by remote sensing, both ground-based and from satellite. Recently, new data sources related to citizen science have emerged, and those data are increasingly generated using low-cost sensor devices (WMO, 2021a).

Observational data must be collected in a way that ensures that the data from different sources are comparable in order to produce globally consistent products and to understand spatial and temporal variations in atmospheric composition. To this end, GAW provides measurement guidelines and quality assurance/quality control tools. The data summary of the World Data Centre for Greenhouse Gases provides an example of information that can be derived from such consistent products. Figure 4 illustrates the temporal evolution and geographical distribution of CO2. Besides the clearly visible increase in CO2 over time, it shows lower CO2 concentrations in the southern hemisphere as well as a less pronounced seasonal cycle than in the northern hemisphere, due to the lower fraction of land area, hence less vegetation.


Figure 4 – Variation of zonally averaged monthly mean CO2 mole fractions. The zonally averaged mole fractions were calculated for each 20° zone. (Source: World Data Centre for Greenhouse Gases (JMA, 2020))

Although the GAW network of observations is growing, important gaps remain (Laj et al., 2019). In large parts of the world there is no observational infrastructure. For political, commercial and institutional reasons as well as because of a lack of capacity, it also happens that some observations are not shared with the international community. Figure 5 showcases the limited availability of measurement through a comparison between reanalysis and GAW measurements of ozone. A large effort has been made in the framework of the Tropospheric Ozone Assessment Report to collect all available data to perform a global assessment of various metrics that address different user communities (Lewis, 2017). This is an important first step towards increasing the availability of data, even if the raw data is not made available, and it hints at the large amount of existing data that is not being fully used.

Global distribution near-surface ozone.png

Figure 5 - Global distribution of near-surface ozone concentrations measured by the GAW network stations (2000–2009) superimposed on model simulated ozone concentrations from the Monitoring Atmospheric Composition and Climate reanalysis (2003–2010). Monthly mean for July. (Source: GAW Report 209 (WMO, 2013))

Data availability represents one challenge, the quality of the observational data is another. Some observations arrive without any associated quality information, which prevents their full utilization. Observational requirements are driven by the quality of the final products and the services that build on them, and they apply not only to the quality of the observations themselves but also to the timeliness with which they are made available. Requirements for atmospheric composition data are defined through three targeted application areas within the broader WMO Rolling Review of Requirement process and are included in a number of other applications. For instance, Monitoring of Atmospheric Composition covers applications related to evaluating distributions of and analyzing changes in atmospheric composition, temporally and spatially, on regional to global scales. Such applications require very small levels of uncertainty and good global or regional data coverage, whereas the timeliness requirement for the data exchanged may be rather relaxed to ensure high quality of the observations.

In contrast, Forecasting Atmospheric Composition Change and their induced environmental impacts covers applications from global to regional scales, with horizontal resolutions similar to global Numerical Weather Prediction (approx. 10 km and coarser) and stringent (near-real time) timeliness requirements. The uncertainty of observations exchanged for this purpose may be higher than for monitoring. This type of applications supports, for example, sand and dust storm warnings, haze-fog prediction and chemical weather forecasts. An example of near-real-time forecast validation is shown in Figure 6. A very specific set of requirements in terms of uncertainty, timeliness, spatial representation and density is applied to urban applications that target megacities and large urban complexes. These applications need horizontal resolutions of a few kilometres or finer and, in some cases, they have stringent timeliness requirements for data availability. A distinguishing feature of this category of applications is their emphasis on research in support of operational services, such as air quality forecasting, which use approaches such as pilot projects and feasibility demonstrations. In addition to the areas mentioned here, many other applications benefit indirectly from atmospheric composition data. For instance, atmospheric composition data improves the estimation of radiative forcing in numerical weather prediction and climate projections (see WMO, 2021b).

Near-real-time_validation_Ushuaia GAW.jpeg

Figure 6 - Near-real-time validation of the Copernicus Atmospheric Monitoring (CAMS) surface ozone forecast with data from the GAW station Ushuaia, Argentina. CAMS uses ozone data from 15 GAW stations for near-real-time validation (Eskes et al., 2021). (Source: ECMWF/CAMS evaluation of global forecasts)

Management and exchange of atmospheric composition data

The approach to data sharing for atmospheric composition largely depends on two factors: The agency that produced the data and national data-sharing policies. Observations made by government institutions using public funding are often subject to open data policies, where data are made freely available through government portals. This includes pollution data for compliance with national and international air quality regulations with reporting obligations.

For the research community, data represents intellectual property and are often made available only after relevant articles have been published, which may happen long after the measurements were taken. These data by universities, research institutes and others are usually collected over a limited time period. Overall, these data contribute significantly to the operational WMO community, although in many cases this is not the primary aim behind the data collection. The atmospheric composition community has in general adopted the so-called F.A.I.R. principles (Findable, Accessible, Interoperable, Reusable) for data sharing. However, the F.A.I.R. principles do not explicitly promote open and unrestricted access, and if this is needed, such provisions must be explicitly formulated: Data need to be both technically open (i.e., available in a machine-readable standard format to be processed by a computer application) and legally open (explicitly licensed to allow commercial and non-commercial use and re-use without restrictions).

Research data are typically stored in dedicated data repositories or cloud-based archives. Due to the affiliation of the same research infrastructure to multiple projects, initiatives and programmes, duplicative data holdings across multiple repositories represents a serious issue which is currently discussed by the community. Metadata repositories that allow multiple project/network data affiliations are considered among possible solutions to avoid multiple submissions (see Figure 7). The maintenance of these repositories also presents a challenge in terms of funding and managing rapidly increasing amounts of data. There is also broad implementation of Digital Object Identifiers (DOIs) that facilitate transparency, traceability and attribution, especially between operational, research and applications communities. Further technological developments related to the implementation of data licensing that would allow for the origin and data ownership to be recognized will support researchers to share data as freely as possible. It is recognized that customized licenses and restrictions may lead to complicated license “stacking”.

Multi data frameworks.png

Figure 7 - Multiple associations with observing programmes/networks are considered by the WMO-sponsored metadata portals OSCAR/Surface and GAWSIS. A similar concept is implemented by the Norwegian Institute of Air Research (NILU). Among multiple frameworks NILU hosts the GAW World Data Centres for Aerosols and Reactive Gases. Data can be affiliated with multiple frameworks to avoid multiple submissions. (Source: NILU)

Access to citizen science and commercial data is much less structured and may even be restricted by subscription. Citizen science projects usually have dedicated websites. However, there is often a distinction between sharing raw data and processed data (products, plots), the latter being generally much more openly shared than the raw data. This may effectively limit the potential for evaluation of the quality of the underlying raw data.

Data from GAW observing stations are collected, quality-controlled and published by dedicated topical World Data Centres. There are also a number of Contributing Data Centres that provide the data of contributing networks. The metadata is available in the GAW Station Information System (GAWSIS), a part of OSCAR/Surface. These Data Centres have been working on the harmonization of data submission and data access procedures and will continue these efforts with the joint vision of a GAW federated data management system that will allow fully interoperable access to all GAW data. An example of information on available data in the World Ozone and Ultraviolet Radiation Data Centre (WOUDC) is illustrated in Figure 8. GAW will continue to liaise with other relevant actors (contributing and research networks, space agencies, environmental agencies and others) to harmonize metadata and data formats and thus facilitate the use of GAW and other data in various applications (WMO, 2017).

Information_on_available_data_total ozone daily obs.jpeg

Figure 8 - Information on available data in the Data Search/Download website of the World Ozone and Ultraviolet Radiation Data Centre (Source: WOUDC)

Registration and attribution are conditions that do not restrict access and reuse, but that may be critical to motivating the research community to share their data. Without citations or documented evidence of use, it is difficult for the research community to demonstrate the value of the data they create to their funding agencies. Intellectual property rights can be ensured through data licensing, which determines rights of use and provides legal security for users. This allows users from academia and the private sector to build viable use cases and business models based on concrete rights of use.

Benefits of the WMO Unified Data Policy

Broad implementation of the WMO Unified Data Policy is critical for the successful delivery of multiple services related to atmospheric composition. Similar to other cases described in article 2 and article 9, improved timely forecasting of extreme events, as well as support to environmental policy, requires open international data exchange. Implementation of the Unified Data Policy can facilitate advances for multiple applications, ranging from improvement of the air quality forecasts to support of transparency mechanisms under the United Nations Framework Convention on Climate Change (UNFCCC). To effectively use the data for specific applications, the quality of the data needs to be known. Thorough evaluation of the uncertainty of the data provides valuable additional information about data usability for particular services. Near-real time availability of the data is important for applications such as forecasts and warnings that need to be issued in a timely manner. For other application areas, for example, reanalysis and trend analysis, timeliness of data exchange is less important.

Adoption of this policy by multiple agencies within Members, beyond NMHSs, will ensure that advances in environmental services and policy are implemented in a comprehensive and cost-efficient way.

It is necessary to improve data exchange between the operational WMO community and the research community. Research projects often require access to external environmental data and services, including forecast information and observational data records, so there is an inherent inter-dependency between research and operations. The research community may not have access to or influence on the operational data (observations and model output) or data formats, which hinders data interoperability, interpretation and scientific progress. It will be desirable to harmonize data sharing protocols. WMO can offer the research community guidelines on data sharing protocols (WIGOS metadata standard and WIS infrastructure) to optimize the value of research data, though the clear benefits of this proposition have to be outlined to the research community. In return, WMO should facilitate the access of operational data to a wider community, emphasizing the mutual benefit of data sharing for both, academia and the operational communities to advance the common understanding and knowledge of the Earth System.

Current WMO policy is well-recognized in the research community, but that community requests clear guidance on licenses as well as clear definitions of the terms “core” and “recommended” data. Licensing is a likely to be a critical success factor to facilitate the sharing of data amongst WMO Members beyond the traditional NMHS community. Standardizing licenses for WMO data could help significantly with the uptake of the WMO Unified Data Policy. Alignment of the WMO data policy with existing licenses (such as the Creative Commons widely used in the research community) would help ensure acceptance in academia and the private sector and prevent practical obstacles for users.

No matter how timely or precise the data requirements are for a given application, a commonly shared requirement is that the data are made available at all. WMO is well positioned to build on the experience gained over its long history of operational data exchange and to extend this also to atmospheric composition data.



Avnery, S., Mauzerall, D. L., Liu, J., & Horowitz, L. W. (2011). Global crop yield reductions due to surface ozone exposure: 1. Year 2000 crop production losses and economic damage. Atmospheric Environment.

Eskes, H. J., Basart, S., Benedictow, A., Bennouna, Y., Blechschmidt, A.-M., Chabrillat, S., Christophe, Y., Cuevas, E., Flentje, H., Hansen, K. M., Kapsomenakis, J., Langerock, B., Ramonet, M., Richter, A., Schulz, M., Sudarchikova, N., Wagner, A., Warneke, T., & Zerefos, C. (2021). Observation characterisation and validation methods document.

JMA. (2020). WMO WDCGG Data Summary No. 44 Volume IV-Greenhouse and Related Gases.

Laj, P., Andrade, M., Sokhi, R., Volosciuk, C., & Tarasova, O. (2019). Sustainability of Atmospheric Observations in Developing Countries. WMO Bulletin, 68(2).

Lewis, A. (2017). Tropospheric Ozone Assessment Report (TOAR): Global Metrics for Climate Change, Human Health and Crop/Ecpsystem Research. Elementa, Special Feature Collection.

Mills, G., Sharps, K., Simpson, D., Pleijel, H., Broberg, M., Uddling, J., Jaramillo, F., Davies, W. J., Dentener, F., Van den Berg, M., Agrawal, M., Agrawal, S. B., Ainsworth, E. A., Büker, P., Emberson, L., Feng, Z., Harmens, H., Hayes, F., Kobayashi, K., … Van Dingenen, R. (2018). Ozone pollution will compromise efforts to increase global wheat production. Global Change Biology, 24(8), 3560–3574.

Montzka, S. A., Dutton, G. S., Yu, P., Ray, E., Portmann, R. W., Daniel, J. S., Kuijpers, L., Hall, B. D., Mondeel, D., Siso, C., Nance, J. D., Rigby, M., Manning, A. J., Hu, L., Moore, F., Miller, B. R., & Elkins, J. W. (2018). An unexpected and persistent increase in global emissions of ozone-depleting CFC-11. Nature.

Shaddick, G., Salter, J. M., Peuch, V. H., Ruggeri, G., Thomas, M. L., Mudu, P., Tarasova, O., Baklanov, A., & Gumy, S. (2021). Global air quality: An inter-disciplinary approach to exposure assessment for burden of disease analyses. Atmosphere.

UNECE. (2016). Towards cleaner air. Scientific Assessment Report 2016. EMEP Steering Body and Working Group on Effects of the Convention on Long-Range Transboundary Air Pollution.

WMO. (2013). Guidelines for Continuous Measurements of Ozone in the Troposphere.

WMO. (2017). WMO Global Atmosphere Watch (GAW) Implementation Plan: 2016-2023.

WMO. (2018a). Scientific Assessment of Ozone Depletion: 2018, Global Ozone Research and Monitoring Project–Report No. 58 (p. 588).

WMO. (2018b). WMO Greenhouse Gas Bulletin No. 14: The state of greenhouse gases in the atmosphere based on global observations through 2017.

WMO. (2021a). An Update on Low-cost Sensors for the Measurement of Atmospheric Composition, December 2020. WMO-No. 1215.

WMO. (2021b). WMO Aerosol Bulletin No. 4: Aerosols from biomass burning.



Jörg Klausen, MeteoSwiss, Chair of WMO/GAW Expert Team on Atmospheric Composition Data Management
Claudia Volosciuk, Oksana Tarasova and Stoyka Netcheva of WMO Secretariat

Share this page