Like any research project, a citizen science initiative should strive to produce results that are robust and scientifically sound. This requires putting in place measures to ensure a minimal acceptable level of quality is met. What the acceptable threshold should be is up to each project to decide based on their field of study and desired impact. The threshold is always relative since what’s acceptable to one project may be insufficient to another. As a general rule, projects that involve citizens in pollution monitoring should aim for quality on at least these two levels: data and participation.

Quality of data

Ensuring data quality is about providing data that is representative of air pollution in the area of interest at a given point in time or over a certain period. Inaccurate data will be of no use to people who want to know about the air they breathe, not to mention public authorities and environmental agencies that need reliable data for modelling and policy making.

The use of low-cost sensors in citizen science is a double-edged sword. On the one hand, it makes citizen science more accessible. More monitors means better coverage. On the other hand, it invites criticism on accuracy grounds. Due to cheaper components, low-cost devices are more sensitive to changes in the environment (e.g. temperature, humidity) than professional-grade equipment. As a result, they tend to over- or underestimate air pollution when fluctuations happen.

Fortunately, a solution to the problem exists in the form of data calibration. In short, the approach involves comparing measurements from low-cost sensors with those of reference stations with the view to correcting and adjusting low-cost sensor performance. Calibration is an essential part of quality control at the level of data as it helps to identify and eliminate data drifts. More on that in lesson 4.2.

Quality of participation

At the level of participation, quality is achieved by ensuring the set-up is representative of stakeholder needs and priorities, by eliminating bias and making sure under-represented groups are given a voice, by providing a meaningful experience where volunteers can contribute in more ways than one, and by conducting citizen science responsibly i.e. paying special attention to ethics and privacy.

As citizen science projects often operate with limited resources, key stakeholders must be prioritised and contacted first. This pragmatic approach should be undertaken carefully to ensure that balance is still maintained whenever possible, and that engaged stakeholders provide knowledge and opinions that are representative of or are accepted by the community at large. Whenever a certain group is excluded, a risk assessment should be performed to understand any negative consequences that may result from failing to include this stakeholder in the process.

Part of the art of citizen science is to empower at-risk communities and make sure their interests are reflected in the overall ambition and measures intended to achieve it. That said, a good citizen science activity will strike a balance, as one needs support of ‘big actors’ (e.g. local, regional, national policy makers) to get credibility and some chance of success, and inputs of under-represented groups to get fairness and justice

Case studies

Athens

The history of citizen science in Athens includes many projects, however most of them focused on biodiversity. Monitoring air quality is a new avenue that the city has explored through COMPAIR. Another novelty is the participation of senior citizens. Their inclusion meant that the Athens pilot could be shaped according to the needs of a vulnerable group that usually suffers the most from air pollution. 

The research team wanted the voices of men and women to be equally represented. To this end, it recruited 54 participants from Friendship Clubs, of which more than half are women. Besides climate and air pollution, volunteers expressed concerns about citizen science particulars. They were worried about power consumption of the devices and how much radiation they emit. They asked questions about privacy, namely whether their personal life be monitored in any way while the device is in use. All this was explained at the introductory workshops and in consent forms that volunteers had to sign in order to participate.

The initial plan was to monitor particulate matter using a DIY sensor.community kit. However, many participants found this device hard to assemble, prompting the research team to switch to an easier alternative (SODAQ AIR). While the SODAQ device was easier to operate, it had to be charged frequently, a requirement that some participants forgot about. One volunteer eventually dropped out because her sensor ran out of power and stopped working for a few days. After she returned the device, her personal information and consent form were destroyed.

Senior citizens in Athens learning how to use air quality sensors

Other volunteers continued monitoring air quality using SODAQ AIR, bcMeter, and NitroSense. Measurement data from these devices was calibrated with data from automatic stations operated by the Ministry of Environment. Post-data collection, volunteers used the Policy Monitoring Dashboard to assess air quality in Kipseli and Neos Kosmos. In the next step, the CO2 Calculator was used to show participants which measures they can take to reduce their carbon footprint and, in doing so, make air quality better.

Experience from the Athens campaign

Berlin

The mobile and static campaigns both followed the structure of an introductory, intermediate and final workshop. At these events, the research team presented all the necessary information about the project so that participants could form an opinion and decide whether it’s right for them. Those who provided informed consent were invited to future workshops where they learned about technical tools and what will happen to their data while using sensors and dashboards.

Berlin volunteers assembling sensors and the their deployment outside home

In the mobile campaign, volunteers measured particulate matter (PM 2.5) with a SODAQ sensor. In the static campaign, besides PM 2.5 which was measured using the same device, participants also measured black carbon with a bcMeter sensor, and traffic flows with a Telraam sensor. Gathered data was calibrated with measurements from official monitoring stations (the BLUME network) to ensure overall accuracy and validity.

Citizen science data was then analysed using Dynamic Exposure Visualisation App (DEVA), Dynamic Exposure Visualisation Dashboard (DEV-D), Policy Monitoring Dashboard (PMD), SODAQ’s knowyourair.net platform, as well as the Telraam website. It’s worth noting that DEVA, DEV-D and PMD were not available at the start of the project; they were developed based on feedback from end users in COMPAIR pilots. So, basically, citizen scientists in Berlin were analysing the data they collected using digital tools they co-created a couple of years previously.

Experience from the Berlin campaign

Recommendations

The groundwork stage can involve several meetings and activities, from a needs-gathering exercise to a hands-on sensor workshop. How many and what kind of engagements to have will depend on the capacity of the leading team and whether participants feel confident enough to proceed with the tasks. Citizen science practitioners should use these engagement opportunities to:

• Objectives: Explain the scope of the project, expected contribution and how volunteers can benefit. In general, participants benefit by improving their environmental awareness, learning new (technical) skills, meeting new people, developing a stronger sense of belonging to their community, gaining a better understanding of air pollution and its impactsUser requirements: Gather feedback on motivations, needs and priorities of interested volunteers. Based on what is possible and feasible, accommodate all the necessary requirements and make sure they are reflected in adjustments to the project set-up made following the initial co-creation round

• Technology: Explain different sensing devices: how they work, what they measure, how to assemble and operate them. Introduce digital tools if they already exist or gather user requirements to inform prototype development

• Ethics: Explain how people’s privacy will be safeguarded throughout the project. Personal data is collected on many occasions and via different means e.g. through event registration forms and surveys, pictures and videos, sensors and dashboards. People might consent to the use of their sensor data in dashboards but would prefer not to see pictures of them circulating on social media. Projects must identify these preferences early on and secure an authorisation from each participant ex-ante

• Informed consent: Capture all the risks, benefits, expected contributions, ethical considerations and privacy statements in a single document so that participants have sufficient information to make a voluntary, informed decision about whether to participate or not. Securing consent is especially important when working with children, in which case it should come from their guardian or parent

<< Back | Next >>