TL;DR — Today we’re sharing the takeaways from our recent webinar on Air Quality Network Design as part of series in conjunction with the Environmental Management Bureau (EMB) of the Department of Environment and Natural Resources (DENR) in celebration of Clean Air Month in the Philippines. We discuss the important questions that should be taken into consideration when designing an air quality network, including determining your budget, the sources and receptors of pollution, the pollutants of interest, and your project goals. Watch the full webinar below, or continue reading for a recap of the key takeaways!
We recently held the second installment of our Philippines Clean Air Month Webinar Series with this webinar focused on Air Sensor Network Design. Thank you to Maggie Isied, MPH, our Application Scientist and Project Manager at Clarity. We’ve recapped the major takeaways from our webinar here.
Air quality regulatory structure in the Philippines and United States
It’s important to consider the context of a given country’s regulatory structures when designing air sensor networks so as to better who will most likely interact with the data. For this webinar, we compared the regulatory structure in the Philippines with that of the United States and found that there were considerable similarities.
The Philippine Clean Air Act of 1999 established air quality regulatory infrastructure in the country. The Philippines is divided into 22 airsheds nationwide, and air quality is managed and controlled at the airshed level.
The United States is similarly divided into smaller subdivisions when it comes to air quality regulation.
At the federal level, the U.S. Environmental Protection Agency (USEPA) has nationwide air quality standards and regulates interstate sources of pollution. The Philippines has the Environmental Management Bureau (EMBD) as the equivalent.
States also monitor and regulate air quality through smaller regulatory bodies, such as California’s Air Resources Board (CARB) that regulate air pollution sources within the state. In the Philippines, the EMB’s regional offices monitor air pollution at this regional level.
At a smaller level, local air districts like the South Coast Air Quality Monitoring District (SCAQMD) deal with air quality in the more immediate area. The Philippines’ airsheds are the equivalent.
How low-cost sensors can be used
The USEPA identifies the use of low-cost sensors for the following uses:
- Science education and research
- Conducting air monitoring projects
- Supplementing regulatory air quality measurements
- Measuring local air quality to better understand sources of pollution
In 2020, the United States Governmental Accountability Office (GAO) released a report entitled “Air Pollution: Opportunities to Better Sustain and Modernize the National Air Quality Monitoring System” which evaluates the state of the national air quality monitoring network and provides recommendations for how agencies can modernize their monitoring infrastructure, including the use of low-cost sensors. Read our blog examining the GAO reports’ recommendations for low-cost sensor use here.
Air quality monitoring projects also exist under state and local agencies. For example, the California Air Resources Board has the AB 617 bill which supports community air quality monitoring in disadvantaged communities. Read our blog here to learn more about the role of environmental justice agencies in improving air quality under AB 617.
The importance of identifying pollutants and sources in air quality network design
It is essential to identify both the relevant air pollutants and pollution sources that affect air quality when designing a monitoring network in a given location.
In the Philippines, most air pollutants come from the following sources:
- Agricultural burning
- Wooden cookstoves
- Mobile sources like traffic
- Coal-fired power plants
- Cement batch plants
How to design a low-cost sensor network
When working to design a low-cost sensor network, some crucial questions to ask are:
- Where should sensors be placed?
- How many sensors should be purchased?
- What sources and pollutants should be measured?
- What rationale goes behind where the sensors are placed?
There are two general methods when it comes to network design: qualitative and quantitative.
The qualitative aspect of network design focuses on the overall design and goals of the monitoring project. It is important to determine the question you are trying to answer by establishing an air quality monitoring network — such as whether you wish to inform air quality policy information, look at exposure assessments and answer health-based questions, better understand pollution sources, or assist in siting a reference-grade monitor.
It is also important to know to whom you are planning to disseminate the data and measurements to, whether only internally or to the public.
Determining your budget by identifying your resources and limitations is the next step, as this determines the number of low-cost sensors that the network can include. Note whether there are data gaps that you would like to better understand as well as whether community members have already identified areas that need additional monitoring.
It is important to prioritize areas that have limited monitoring data, are densely populated, or would otherwise reap the most benefits from air quality monitoring.
Next, organize a map of the area that you are planning to monitor. Diversify sensor siting locations, including points of measurement in residential, industrial, and commercial areas, as well as high-traffic areas and recreation spots.
Note how factors such as geography, topography, meteorology, pollution sources, and wind patterns may affect where you want to place your sensors.
Identify the types of pollution sources that you are interested in understanding, whether this is a mobile source such as polluting vehicles or a stationary source such as a factory. Determine the pollutants associated with these sources — such as nitrogen dioxide if you are looking at traffic-related pollution — and ensure your sensors are fit to measure these pollutants.
The quantitative aspect of network design involves the data models that can be used to determine sensor placement.
One approach, the Multiresolution Dynamic Mode Decomposition Model, uses the variation in pollutant concentration over space and time to identify where to place sensor. This approach requires both spatial data and historic air pollution data to function.
Kelp et al., a USEPA-funded pilot study, used this model to determine the optimal placement of PM2.5 low-cost sensors in the contiguous United States. By using particulate matter data from 2000 to 2016, the researcher studied the daily variation in concentrations in order to identify high-priority sensor locations.
The research suggested that there are gaps in the current EPA reference-grade monitoring network in certain areas of the country, which can be supplemented by low-cost sensors.
Another approach is known as the Gaussian Process Model. Those using this approach will train a computer model to identify good placement of low-cost sensors over space and time that are most informative about locations lacking sensors. The model also evaluates the effectiveness of sensor placement.
There are different ways to interact with the data gained from an air quality monitoring network. At Clarity, we have a cloud-based API that allows customers to interact with their data.
Our Clarity Dashboard is used by both governments and industries, allowing a viewer to quickly understand air quality at a specific site, explore data, find trends and insights, and manage the data network.
Clarity’s OpenMap is used by citizens and can be used by customers to optionally share their data with the public on a common, easy-to-use platform.
Hybrid networks in action
Hybrid networks involve both reference-grade and low-cost sensors’ air quality data that are used in conjunction to arrive at a more complete picture of air quality in a given location.
The Breathe London network was launched in January 2021 and now, after a two-year pilot, is managed by the Imperial College of London. The network integrates air quality measurements from hundreds of air sensors with reference-grade measurements.
There is a high demand for localized air quality monitoring in London, especially considering the city’s long history with air pollution. The network involves a public-private partnership, allowing citizens to engage with the issue of air quality as never before as well as allowing organizations to deploy nodes or donate nodes to communities.
Caption: With regulatory monitoring and Breathe London’s low-cost sensor network, the city of London has over 400 air quality measurement points, providing a high-resolution look at air quality in the city.
The low-cost sensor network in Quezon City, Philippines began in June 2021 and has now grown to 20 sensors around the city. The network’s aim is to help identify opportunities for clean air interventions, which can then be used in the city’s air quality management plan to help the city work to meet the World Health Organization’s air quality guidelines.
Clarity’s take on leveraging low-cost sensors
Here at Clarity, we believe that is it vital to know your objectives in establishing your air quality monitoring network, including:
- What do you hope to learn about air quality? Are there specific questions to answer?
- What are your regulatory or non-regulatory performance targets?
- What spatial and temporal resolution will be necessary to achieve your goals?
- What pollutants do you need to monitor to support your goals?
From beginning to consider existing infrastructure to planning the scope and siting of your deployment of low-cost sensors, it is important to complete each part of the planning process in order to successfully establish an effective air quality monitoring network.
Interested in learning more? Read our Guide to Leveraging Low-Cost Sensors for Air Quality Monitoring 2.0 here.