AIP AQ Unified Scenario

From Earth Science Information Partners (ESIP)

<Back to AQ Pilot Scenario Workspace

This scenario describes how air quality and related earth observations could be used to inform a wide spectrum of decision making; it is structured around three decision-making end users:

  1. A policy-maker, needing synthesized information on the importance of intercontinental pollutant transport
  2. An air quality compliance manager, who needs to assess whether a regional pollution event was due to an "exceptional event"
  3. The public, needing information about air quality now and in the near future to make activity decisions

In general, the scenario envisions GEOSS facilitating two broad goals: building connections to facilitate movement of data between actors, and developing interoperable tools for intercomparison and fusion of a wide variety of atmospheric data. A number of the actors in the scenario have overlapping roles, and in reality the same individuals will serve several downstream decision makers. In fact, the common need for integrated atmospheric observations is a primary motivation for the structure of this scenario.

Overview

Air pollution is a global problem that causes premature mortality and morbidity, damages crops and ecosystems, and contributes to climate change. In the United States alone, poor air quality is estimated to cause tens of thousands of deaths and cost society more than $100 billion annually. Globally, air pollution contributes to the deaths of more than 800 thousand people per year, most in the developing world.

This scenario describes how air quality and related earth observations could be used to inform a wide spectrum of decision making; it is structured around three decision-making end users:

  1. A policy-maker, needing synthesized information on the importance of intercontinental pollutant transport
  2. An air quality compliance manager, who needs to assess whether a regional pollution event was caused by transport of pollution from a distant fire, dust storm, etc.
  3. The public, needing information about air quality now and in the near future to make activity decisions

Air pollution does not respect jurisdictional boundaries and is affected by sources and processes over local, regional, intercontinental, and global scales. Understanding the causes of specific instances of air pollution and predicting air quality in any area therefore requires descriptions of atmospheric processes valid and useful on a wide range of scales. Given the wide variety of relevant observations at many scales, each of the above decisions ultimately needs an array of observations and models to describe the atmosphere.

Earth observing data are needed from a wide variety of sources:

  • ambient monitors, measuring the concentrations of pollutants near the ground
  • radiosondes and other instruments which profile an atmospheric column
  • chemical transport models
  • satellite measurements, which report either column densities of pollutants or, with limited vertical resolutions, 3-D fields
  • meteorological data and models
  • emissions data and models
  • demographic and economic information

Decision makers, and those that inform them, need access to these data. Moreover, each data source above is significantly limited and not able to broadly document the state of the atmosphere. Therefore, synthetic fusion and intercomparison of the data will allow analysts to produce a far more complete and accurate description of the atmosphere than obtainable from any one data source, or any one class of data. There are a number of scientific approaches to this challenge, but the technical tools for intercomparison, fusion, and processing of air quality data are not operationally available. Such tools would not make data fusion and intercomparison automatic, but could make it operationally feasible.

Therefore, this scenario a) documents actors' need for access to existing near-real time and historic information, which GEOSS will facilitate and b) describes needed integrated air quality information, which is not operationally available, but which could be facilitated with interoperable, service-oriented tools for data fusion and intercomparison that will be promoted by GEOSS. To describe the needed information and tools, the scenario also documents the upstream users, and the information they need, in the data value chain.

Note that this scenario is consistent with GEO project HE-07-03: Integrated Atmospheric Pollution Monitoring, Modelling, and Forecasting in the GEO 2007-2009 Work Plan, and with the efforts of the CEOS Atmospheric Composition Constellation; the development of the GMES Atmospheric Service, GEMS, and its planned extensions; the evolution of AIRNow, IDEA, 3D-AQS, and Datafed; and other ongoing efforts.


Context and pre-conditions

Actors in the scenario, and the information they need

A number of actors process earth observations information upstream of the decision makers, who base their decisions on highly synthesized data.

Actor Grid2.png

While presented here in a matrix, a number of these actors have overlapping roles, and in reality the same individuals will serve several downstream decision makers. In fact, the common need for integrated atmospheric observations is a primary motivation for the structure of this scenario.

Intercontinental pollution transport

  • End use decision maker: Policy maker negotiating an international agreement on long-range pollutant transport
    • Information needed: Synthetic assessment reports which quantify the impact of intercontinental pollutant transport on various nations
  • Upstream information processor: National or internation scientific advisory group
    • Information needed: Technical assessments of model experiments and synthesized historical observational datasets designed to assess long-range transport
  • Upstream information processor: Scientific task force assessing long-range transport
    • Information needed: Synthetic description of the atmosphere, using multiple observations and models (satellites, ambient, etc.)
  • Upstream information processor: Air quality data analysts
    • Information needed: Chemical transport models, satellite observations, ambient observations, emissions inventories and models, meteorological data, data integrating all of the above

Exceptional pollution event

  • End use decision maker: Air quality manager assessing pollution event: is it an exceptional event?
    • Information needed: Assessment reports which quantify the impact of transport on the region for that period
  • Upstream information processor: Air quality data analysts
    • Information needed: Chemical transport models, satellite observations, ambient observations, emissions inventories and models, meteorological data, data integrating all of the above

Member of the public planning activities

  • End use decision maker: Member of the public
    • Information needed: Air Quality Index or similar health-based index for the local area
  • Upstream information processor: Air quality forecasters / data collectors
    • Information needed: Chemical transport models, meteorological data, ambient observations when available, data integrating all of the above. For regions of the world without ambient monitors, integrated satellite-model data are particularly necessary.

Earth observations providers

Most of these earth observations are used in all three of the decision making information chains.

  • National, State/Provincial, Local Environmental Management Agencies
  • National Meteorological Agencies
  • National Space Agencies
  • National Land Management Agencies
  • Industry
  • Consultants
  • Academic and Other Research Institutes
  • International cooperative fora (e.g. WMO, CEOS, EEA, ...)

Starting Information

Information available before scenario begins:

  • Meteorological data
    • Observations from ground-based networks, satellites, sondes
    • Forecasts from numerical models at the global and regional scales
  • Geographical data (land use, demographics, emissions-related activity, ...)
  • Atmopsheric Composition (Air Quality) Observations
    • Surface Monitoring Networks
    • Satellite Observations
    • Sondes
    • Ground-based remote sensors
    • Aircraft Measurements
  • Numerical Air Quality Chemical Transport Models (at regional to global scales)

Processing and Collaboration Functionality

  • Facilities to register all the components/services needed for the execution of the scenario.
  • Convenient portals for finding, accessing, visualizing, and processing observational and modeling data by analysts in near real time and for historical analysis
  • Data sharing and integration functionality including (1) registry/catalog for finding resources (2) standard-based access to spatio-temporal data and metadata, workflow software for integrating Service Components
  • Integration of multiple observational data sets to create rich 4-dimensional descriptions of the atmosphere to improve understanding of atmospheric processes
  • Comparison of observational data to numerical model estimates to improve numerical model descriptions of historical conditions (events or long-term trends)
  • Real-time assimilation of observational data into numerical models to improve numerical forecasts
  • A workspace to support the activities of the Air Quality Community of Practice participating in the Scenario
  • Effective mechanisms for distributing (in near real time) maps/images, descriptive information, and processed data health, emergency response, and air quality management authorities; to mass media; other research and assessment communities (e.g., health); and the general public.

Scenario Events

The cyberinfrastructure envisioned by this scenario is illustrated with the following events. However, since it will enable analysts to combine wide range of air quality observations, models, and other information, it will ultimately be used to produce a broad range of decision support products for a number of different audiences.

Assessment of International and Intercontinental Transport of Air Pollution

Assessment of long range pollutant transport is currently underway by several bodies. The Task Force on Hemispheric Transport of Air Pollution (TF-HTAP) has organized a series of cooperative analysis efforts including a model intercomparison exercise, which to date involves more than 25 global modeling groups and for which a data server has been established at FZ Juelich; a compilation of relevant surface observations, which is being developed by NILU as a component of EBAS; a compilation of relevant aircraft campaign observations, which is being developed by NASA Langley; and an updated version of the EDGAR global emissions inventory.

These cooperative efforts would be enhanced by:

  • constructing linkages between the various databases and other existing air quality-related data hubs (e.g., Datafed and GIOVANNI) and
  • developing and linking tools to facilitate comparison of models, observations, and emissions data. Such visualization and analysis tools may build upon existing tools (e.g., AMET, RSIG, and HemiTap Tool).

As in the other events, these capabilities are used to form more complete and accurate descriptions of the atmosphere than available from any current type of atmospheric observation. In the TF-HTAP example, Earth observations will then be used to inform policy makers via the following value chain:

  1. Model experiments and analyses of historic datasets are used to generate state-of-the-science quantitative estimates of the importance of intercontinental transport of air pollutants.
  2. Research efforts will then be compiled into a detailed report of the task force.
  3. This report is then used as the basis for an synthesis report and executive summary which will be delivered to policymakers to inform their decision making process as international conventions consider initiatives to address long range pollutant transport.

Other efforts to assess long-range pollutant transport will use different approaches, but the connectivity and tools developed as part of this effort will be useful for those approaches too. Moreover, these capabilities will be applicable to model evaluation and analysis at the regional scale as well, ultimately benefiting a large community of air quality managers and researchers.

Exceptional Event analysis

Air quality is periodically influenced by natural and anthropogenic events, such as wildfires and dust storms. For regulatory purposes in the US, pollution episodes can be flagged as 'exceptional events' if an area would not have exceeded the pollution standard without the occurrence of a an uncontrollable and unusual natural or anthropogenic event. If it qualifies, the event is flagged and the exceedance in not considered in determination of whether the region complies with the Clean Air Act. In the US, any reasonable scientific evidence can be used to document whether the event qualifies as an exceptional event.

As with the other scenario events, GEOSS can assist these decision makers by facilitating linkages between actors and by promoting the development and distribution of service-oriented, interoperable tools for intercomparison and fusion of models, observations, and emissions data.

An event might be very noticeable or subtle. Therefore, impetus to examine a given event could come from a number of sources. However, a typical scenario would proceed in steps such as:

  1. Designated analysts at regional air quality agencies will identify a possible event. Analysts will use models, ambient observations, and satellite observations. Analysts should be open to input from the wider community, even at this early stage.
  2. Once an "interesting" event is preliminarily identified, the analysts will compile relevant data to explore the origin and evolution of the pollution, sharing their analysis on a virtual workspace.
  3. Combining observations, meteorology, and models, analysts quantify the effect of the event on the receptor region.
  4. Evidence for the influence of the event on the receptor region is compiled into a report submitted to air quality managers who assess the region's compliance with air quality standards.

Current projects (e.g. FASTNET, IDEA, SmogBlog) are significant building blocks of the needed networks and tools.

Informing the public and the health sector about air quality in real-time or near-real-time and future forecasts
(e.g. AIRNow, AIRNow-International, EPA/National Weather Service National Air Quality Forecast guidance, PHASE)

In the United States under the Clean Air Act, the EPA was required to establish a nationally uniform air quality index for reporting air quality data to the public. In 1997, the Pollutant Standards Index was redeveloped in conjunction with the latest health effects information. The resulting Air Quality Index (AQI) for the United States provides a simple, uniform system to report levels of the criteria pollutants (ground-level ozone, particle pollution (also known as particulate matter), carbon monoxide, sulfur dioxide, and nitrogen dioxide) which are federally regulated. Air quality health indices should link health to air quality concentrations, providing the public with timely, easy-to-understand information about air quality, so that citizens can personally determine their levels of health concern. EPA has assigned a specific color to each AQI category to make it easier for people to understand quickly whether air pollution is reaching unhealthy levels in their communities. Using colors to convey a health meaning instead of providing raw data has shown to be successful in many countries in communicating air quality conditions to the public. Informing, educating and protecting the public about air quality via air quality health indices facilitate the management and implementation of regulatory control strategies to improve air quality.


Current Air Quality Conditions


Consistent real-time reporting (i.e., the current hour) of air quality conditions is critical to the analytical work of air quality forecasters, managers and scientists as well as informing the public and media. The envisioned cyberinfrastructure could facilitate the reporting of air quality information worldwide. Air quality observations received by GEOSS could be disseminated to users, organizations, or applications using a variety of technologies in the architecture. This information can be distributed via mass media, internet, and telecommunications using Common Notification Protocols, including delivering location-specific air quality data to cell/smart phone devices.

A typical scenario for reporting and communicating current air quality would proceed in steps such as:

  • A continuous monitoring network of several air quality monitors are available
  • Designated technical staff at air quality agencies will monitor air quality observations. Ideally air quality observations from neighboring or regional localities are available to better understand current air quality conditions.
  • Monitored values would be ingested by a database management system. Data would be stored, quality controlled and converted to an air quality health index.
  • Processed air quality data would be assimilated and presented graphically for technical staff in real-time.
  • Approved air quality data would then be distributed to other applications, such as media channels, internal decision boards, the internet, GEOSS, and air quality models. Outreach specialists at air agencies would be engaged and ready to communicate the health effects and activity modifications that coincide with elevated air quality conditions.


Air Quality Forecasting


In the United States the state and local air agencies provide the public and media with current and next day AQI forecasts. AIRNow acts as the centralized clearinghouse for AQI information and distributes these data and information on a national basis. The motivation for air quality forecasting by local agencies is to provide the public with air quality information that can be used to make daily lifestyle decisions to protect health. Real-time and forecasted air quality information plays a very important role in informing the public about potentially harmful conditions. This information allows individuals to take precautionary measures to avoid or limit their exposure to predicted unhealthy levels of air quality. Most environmental agencies have an air quality meteorologist, or a team of meteorologists, using a range of forecasting tools combined with practical experience to issue predictions for peak pollutant levels expected during the current- and next-day periods. Some of the challenges typically faced by air quality forecasters in making more accurate and specific forecasts include the following:

Some of the challenges typically faced by air quality forecasters in making more accurate and specific forecasts include the following:

  • Developing better understanding and quantification of spatial and temporal emissions from both anthropogenic and biogenic sources.
  • Improving the resolution and accuracy of meteorological forecasts, particularly during weak synoptic forcing (i.e., stagnant, high pressure conditions).
  • Developing and improving air quality tools and models that provide higher spatial and temporal forecasts to provide the public with information about the location and duration of unhealthy air.


Air quality observations along with weather and other observations are a vital component to the understanding and science of air quality forecasting. Consistent real-time reporting (i.e., the current hour) of air quality conditions is critical to the analytical work of air quality forecasters, managers and scientists. Various forecasting tools exist to predict future air quality concentrations and range from simple rules of thumb to statistical methods to photochemical models. Air quality episodes are affected by local/regional meteorology and by emissions sources that vary spatially and temporally. Consequently, air quality forecasting tools and techniques have been developed to focus on these local phenomena.

The range of techniques encompassing air quality forecasting include:

  • Criteria (rules of thumb) use thresholds of forecasted weather variables to predict future pollutant concentrations. For example, high ozone concentrations often occur with hot temperatures. Historical analysis might reveal that days with temperatures above 90F tend to have peak 8-hr ozone concentrations greater than 85 ppb.
  • Parametric methods (statistics, neural networks, fuzzy logic) create objective relationships between forecasted meteorology and predicted pollutant concentrations. Multivariate regression equations have been successfully used to forecast peak ozone concentrations in many areas of the country.
  • Photochemical models are a collection of computer algorithms that simulate the three-dimensional atmospheric and chemical processes that influence pollutants including transport, dispersion, and chemistry. Key inputs to the model include an emission inventory, meteorological predictions, and initial and boundary conditions for chemistry. Organizations internationally have begun running these models in a real-time mode to produce gridded fields for predicted ozone concentrations.
  • Experience gained by day-to-day forecasting, similar to weather forecasting, plays a large part in accurately predicting air quality and helps balance the limitations of the other methods noted above.


An excellent example of multi-agency collaboration the EPA and the National Oceanic and Atmospheric Administration (NOAA) National Weather Service (NWS) have established a partnership to make full use of their respective capabilities and authorities in developing a national air quality forecasting capability. This real-time air-quality forecasting (AQF) system is based on linking the EPA’s Community Multiscale Air Quality (CMAQ) Modeling System with the NWS National Centers for Environmental Prediction’s (NCEP’s) operational North American Mesoscale (NAM) weather-prediction model. These types of numerical air quality models could provide a better tool for local environmental agencies, thereby improving their own current air quality forecast capabilities or such as model could simply provide air quality forecasts for areas where no forecasts are currently available. Numerical model results offer a real-time scientific laboratory for air quality modeling with daily model simulations that can be evaluated continuously. Having a continuous archive publically available within GEOSS could aid in the development or enhancement of air-quality climatology research, air quality surfaces for health studies, and exceptional event analyses.

These efforts to understand the environment factors in local airsheds and the tools/models developed and operated are necessary to predict future air quality. Improved air quality forecasts not only better protect public health; they also assist air quality decision and policy makers with managing air quality and implementing rules and control strategies on a local and perhaps regional and national basis. To improve the accuracy of air quality forecasting, it is necessary to build upon the real-time connectivity and model-data comparisons discussed above and to assimilate observations into numerical simulation models in real-time. There are a number of existing efforts (GEMS, RAQMS) to use satellite observations to provide the initial conditions for numerical models producing air quality forecasts for the next 24-72 hours. Developing standard approaches and protocols for such processes will help expand the use of assimilation techniques, improving air quality forecasts for the benefit of all. Assimilation of satellite observations into numerical models may also enable nowcasting” surface air quality in areas that do not have surface air quality monitors, which is the situation in much of the developing world.

A typical scenario for improving air quality forecasts would proceed in steps such as:

  • A continuous air quality and metrological monitoring network, including data and information potentially outside the jurisdiction (such as nearby states, countries) of the air agency are available
  • Database management system available to process and manipulate the air quality and meteorological information. Assimilation of satellite observations and products could be included.
  • Inventory of available statistical, analytical and modeling tools and applications for all pertinent and applicable air quality and meteorological variables
  • Incorporate and integrate available tools and information provided by inventory and develop standard operating procedures and techniques to utilize the information.
  • Develop methodology and technique to assimilate, integrate and correlate satellite observations and measurements
  • ….
  • ….
  • ….


Involving the Health Sector


In addition to supporting the analytical work of air quality managers and scientists, the envisioned cyberinfrastructure will also facilitate the provision of useful information to the health community and the general public. By providing real-time and forecasted air quality information to the public, individuals can make decisions to protect themselves from harmful exposures. Historical, real-time, and forecasted air quality information can also help health authorities assess public health impacts of air pollution episodes and respond to extreme events. In the United States, the Centers for Disease Control and Prevention (CDC) are working with EPA, state, academic, and other partners to develop a National Environmental Public Health Tracking (EPHT) Network. The EPHT Network was developed to relate health surveillance data to environmental exposures. A pilot effort (the Public Health Air Surveillance Evaluation or PHASEis underway to understand and provide data that is most useful for exposure and health impact assessment.


A typical scenario to better understand and provide information that is most useful for exposure and health impact assessment would proceed in steps such as:


  • Identify and evaluate sources of air quality characterization which would include ambient air monitors, satellite measurements (such as the NASA Satellite Aerosol Optical Depth data), air quality forecasting models (such as the EPA/NWS air quality forecasting model guidance) or actual air agency specific forecasts, and statistically combined sources.
  • Identify, select and obtain air quality measurements for linkage, such as a 24-Hour average fine particle concentration or a 1-Hour ozone concentration
  • Identify the health effects of interest in evaluation, such as adult and pediatric asthma, acute cardiovascular endpoints, congestive heart failure and stroke
  • Identify, select and obtain sources of health effect data such as hospital discharge, mortality, and emergency department
  • Identify, select and obtain health effect measurements for linkage such as daily hospital/emergency department admissions and daily mortality counts
  • Identify and address potential public health actions such as: air quality index (AQI)/alert awareness and effectiveness, hot spots, predict hospital peaks and alert health providers for staffing and preparation, follow-up investigations and hypotheses generation
  • Determine the evaluation criteria to develop the best method of health impact assessments such as: resource requirements, ease of use, spatial and temporal coverage, compatibility with health data, correlation with actual human exposure, and uncertainty quantification


By linking forecasted pollutant concentrations to the AQI, each forecast carries a health effect message and suggests actions to reduce exposure. Ongoing health effects research is critical to the air quality forecasting program because it ensures that the latest health effects information is available for incorporation into the AQI. Efforts to capture lessons learned from these existing efforts and to develop standard approaches and protocols will help air quality management agencies expand the amount of air quality information available to the public and interested communities.


Current projects (e.g. AIRNow, AIRNow-I, PHASE)are significant building blocks of the needed networks and tools. The AIRNow-International (AIRNow-I) initiative will potentially allow other interested countries and organizations to report and distribute air quality conditions. AIRNow-I incorporates modules for data processing, quality control (QC), system diagnostic monitoring, mapping, and distribution and will be fully integrated with a geographic information system (GIS). AIRNow-I is being designed to provide flexible mapping tools applicable worldwide, including providing Open Geospatial Consortium (OGC) web services (Web Mapping Service, Web Feature Service, Web Coverage Service). When fully implemented, countries that do not currently have the capability to participate or contribute to the GEOSS effort will have the opportunity to dataflow air quality observations into the GEOSS architecture.

As stated above, the actors will benefit from constructing linkages between data sources and tools, but also from development and linking tools to facilitate comparison of models, observations, and emissions information.

Archive: Older Scenario Versions

  1. GEOSS_AIP_Pilot_-_Initial_Scenario
  2. AIP AQ Scenario A: Smoke and Dust
  3. AIP AQ Scenario B: Model - Data Synthesis
  4. AIP AQ Scenario C: forecasts
  5. AIP scenario presented in Ispra (wiki version)