Policy on air pollution in the urban environment relies on evidence. That evidence does not come from one monitor, one model or one field campaign. It is constructed over time from measurements, experiments, models, comparisons and interpretations.
That is particularly the case at street level, where the urban environment is characterised by traffic, buildings, meteorology, turbulence and people. A reading at one street cannot be considered representative of another street. A model can give concentrations over an area, but may still fail to represent a local pathway. A field campaign can provide evidence under one set of conditions which may never happen in the same way again.
Therefore, urban pollution evidence must be pieced together from different sources. Its purpose is not just to measure concentrations, but to understand how pollution moves around the urban environment, where people are exposed and what decisions can be made with the available evidence.
Why Field Campaigns Matter
Field campaigns are important because they show what the real urban environment looks like. They provide information on the complexity of the urban environment which cannot be reproduced entirely in a model or experiment: actual traffic emissions, actual buildings, actual meteorology, actual background pollution and actual people.
The DAPPLE programme emphasised the importance of fieldwork in understanding the processes at the street and neighbourhood scale that control air quality in the urban environment. DAPPLE field campaigns combined measurements of winds, traffic, tracer release, fixed instruments, exposure and modelling to understand how air, people and pollution interact in the urban environment.
That evidence is important because we do not make decisions in the real urban environment based on evidence from perfectly uniform streets.
What Field Campaigns Reveal About the Urban Environment
A field campaign is not just about placing a monitor next to a road. In the complex urban environment, the design of the field campaign is at least as important as the instruments.
Measurements may need to be taken at several heights, from the street to higher elevations. Information on traffic may be needed to understand emissions. Meteorological information may be needed to understand dispersion. Mobile or personal exposure measurements may be needed to understand what people experience as they move through the urban environment.
That is because street-level pollution does not exhibit simple behaviour. A street junction is different from a straight street. A side road is different from a street with traffic. A building corner may generate local recirculation. A slight change in wind direction may change how people are exposed.
A field campaign is needed to show this variation rather than assume it away.
From Concentrations to Pathways
Air quality evidence is often discussed as if we were just interested in concentrations. That is important, but it is only part of the story.
In the urban environment it is important, for decision making purposes, to understand pathways as well as concentrations. Where did the pollutant come from? How did the pollution move? Did it affect specific streets, junctions, or building corners? Was exposure driven by a nearby source, a pervasive background level, or the configuration of the street network?
Tracer techniques help answer these questions because they let scientists introduce a known substance and watch its dispersal. Tracer studies provide evidence of how pollutants might flow through a street canyon, navigate around a corner, move past a junction or pass over rooftops — processes not always obvious with just background air quality monitoring.
In this way, the focus of the evidence base moves from “how much was seen” to “how did it move?” That makes a practical difference for understanding people’s exposure to urban air, designing monitoring networks, and preparing for emergency incidents.
The Need for Controlled Experiments
Field measurements are valuable for giving insight into real cities, but they are not easily repeatable. Wind speed, wind direction, traffic levels and background concentrations all vary continuously. A specific measurement today cannot be reproduced in exactly the same situation the next day.
A controlled experiment helps address this problem.
Wind tunnel experiments allow models of cities to be tested at scale under repeatable flow conditions. The relative location of a source, wind direction and the configuration of surrounding buildings can all be systematically altered to see what happens. Of course, the whole city cannot be represented, but physical processes which influence dispersal patterns can be isolated.
This is why the relationship between field data and wind tunnel data is such a central element of how evidence on urban air pollution is put together. Field evidence tells us what actually happens. Wind tunnel work can help explain why some things might happen in some circumstances.
The evidence is most powerful when both are drawn upon together.
Wind Tunnels as Process Evidence
Wind tunnels are particularly useful because they make process visible. They can help to understand, for example, if a street canyon is holding onto air, how the mixing of pollution changes at a junction, or what happens to a plume as wind direction changes.
Wind tunnel work should not be used as a substitute for field measurements. What it can do is allow physical explanations to be tested under conditions that can be repeated and compared.
For example, controlled wind tunnel experiments might be used to assess whether a high concentration observed in the field is likely to arise from lack of ventilation, the position of the source, local recirculation or the configuration of buildings in the immediate vicinity.
This process evidence is useful for planning and policy because it can be used to distinguish a one-off occurrence from a physical process that is occurring repeatedly.
Modelling as an Evidence Tool
Modelling is required because you cannot measure every location, at every time, for every scenario. Models let researchers and policy users estimate concentrations in places where you are not currently able to measure. They let you test different assumptions. You can explore how things might change as traffic patterns, buildings and weather all vary.
Model outputs on their own, however, are not evidence. What can make them useful is if they are linked to evidence from field measurements and from laboratory studies. It also depends on your understanding of model uncertainties.
It is for that reason that model validation is a policy issue rather than a purely technical matter. A model meant for transport planning, public health assessment or emergency response should be tested against evidence that is appropriate to the scale and nature of the decision making.
If a model is used for long-term averages, that does not mean it will also perform well for short-term peaks. A model that works in an open environment might not be as accurate in a dense urban street canyon. A rapid model might be useful for an emergency, if its limitations are understood.
Combining Physical and Digital Evidence
The combination of physical and digital modelling is becoming central to urban dispersion research. The controlled evidence about flow and dispersion processes offered by physical modelling can be complemented by exploring a much wider range of scenarios in a digital model, while also ensuring that real urban processes are considered with field data.
A combined approach to physical and digital modelling matters because no method alone can provide a full picture. Digital models can provide a large number of scenarios, but they also require validation. Physical experiments can provide insights about dispersion in simplified conditions, but in order to apply them to urban conditions they will need extrapolating and they are unlikely to provide a full picture of complex urban conditions.
Field campaigns can provide realistic measurements, but they can be difficult to repeat and subject to significant variability.
By considering the three as part of a cohesive approach, rather than separate evidence-gathering exercises, we can better understand the evidence we need.
Sensors and the Expanding Evidence Base
Sensor networks can provide yet another layer of evidence to add to urban pollution assessments. They have the potential to add to the identification of spatial and temporal variations of pollutant concentrations and to identify plumes that will go unnoticed in smaller, static monitoring stations.
However, simply adding more sensors into a network will not lead to more accurate assessment. The placement, quality, representativeness and calibration of sensors will still be key considerations. While adding more sensors may provide more insight into the detail of a particular event, we need to know what to expect. A more detailed map of pollutant concentrations might just as well give a misrepresentation if we do not understand the uncertainties behind it.
The aim is to better define how to use evidence in specific cases: in what situations and why is this type of evidence needed?
An assessment of how sensor networks can support urban pollution monitoring can help to answer that question, by showing how sensors can be used to identify local variations, test model predictions, support emergency response planning and investigate exposure pathways.
The point is that sensors are not simply extra monitoring equipment, but part of a larger evidence framework.
Evidence for Emergency Planning
Evidence needs to be built before it is needed, and there are no surprises here. Emergency planning demonstrates what needs to be understood about pollution events to be able to make appropriate decisions under uncertainty.
An airborne release could result in a lack of time and available information. There might be a need for responders to be able to determine where the material is moving to, where potentially affected areas are and the priority for measurement and monitoring.
In complex urban conditions, a determination of wind direction alone may be insufficient to accurately determine where material is moving. In an urban environment, a decision on where to measure will be based on field observations, field trials, wind tunnel and modelling exercises, or even knowledge about sensor deployment strategies that has been developed beforehand.
Planning for urban airborne releases requires an understanding of near-source pollutant transport at street junctions, around building corners and in street canyons.
The need to build this evidence is a good example of how research evidence becomes policy relevant. The aim is not to predict accurately every incident that could happen, but to provide useful evidence that can improve decisions under uncertainties.
Translating Evidence into Policy
Transitioning from a campaign to the adoption of policy is no simple matter. Interpreting and presenting evidence is critical.
While a set of measurements may find that high concentrations have occurred in one place, the ultimate policy users must be able to understand the causes and assess the probability of reoccurrence. If a model predicts exposure, it is critical that its limitations and use cases are clear. If a network of sensors has been deployed, this must be to answer an explicitly formulated question that will aid decision-making.
The following points can be considered guidelines:
- Evidence must be multi-source. Field studies, tracer releases, wind tunnel studies, modelling and sensor networks all generate data with a different character.
- Scale matters. Evidence relevant at city scale is often not sufficient for the purpose of estimating street-level exposure or in an emergency.
- Uncertainty is to be made explicit. The ultimate user needs to know what the evidence can and cannot be used for.
- Monitoring and modelling should be coordinated. The value of measurements is often derived from how well they can be used to test and interpret model performance.
- Policy should dictate research design. Evidence is strongest if it is generated in response to a pre-defined, but open-ended question. In other words, the end decision should be specified before the start of a campaign.
These issues need to be considered so that campaigns are not simply conducted as an exercise for their own sake.
The Role of the DAPPLE Approach to Evidence
The wider contribution made by DAPPLE is not just in terms of the new measurements or models generated, but in demonstrating the necessity for an integrated approach to urban dispersion.
DAPPLE combined:
- field campaigns generating real-world observations;
- tracer releases to identify pathways;
- wind tunnel experiments to generate repeatable physical evidence;
- numerical models for wider predictions and scenario testing;
- measurement campaigns to connect human exposure and city-scale dispersion.
This remains true for a variety of reasons. The urban pollutant environment is complex and difficult to describe at a scale that is relevant to individual people. No one single monitoring point can describe the full urban environment, modelling remains an imperfect prediction method that requires validation against measurements, a single field campaign cannot be used as the basis for all scenarios and a sensor network is simply a set of data without further interpretation of the underlying physical processes that generated those values.
The real value is in the combined nature of this evidence.
Building Better Urban Pollution Evidence
The generation of urban pollution evidence can be viewed as a sequence of linked questions.
- What is being observed in the real world?
- What processes are responsible?
- Can these processes be observed in a controlled environment?
- Can these processes be simulated by a model?
- Can these processes be observed with a sensor network?
- What is the level of uncertainty associated with the above?
- What decisions are likely to be based on the result?
In other words, this is evidence that moves from observation to interpretation to policy. It is this process that is so critical to the quality of urban air quality evidence.
Decisions made on traffic and planning, public health, monitoring and emergency are all based on the quality of evidence that can be used to underpin them.
The evidence generated by field campaigns can often be used to provide a starting point, but the evidence is not useful unless it is tested, linked and used to inform decisions related to people and the environments where they live, work and travel.
