‘We can use the cars to measure urban air pollution much more accurately’

Large-scale study of air quality in Oakland using Google Street View cars

Researchers at the IRAS are working together with Google and other partners as part of a large-scale project in Oakland, California to measure urban air pollution. Google Street View cars are currently driving around the city equipped with built-in measurement equipment which can measure air pollution at a much higher resolution than fixed measurement stations. The study was recently published in the American periodical Environment Science & Technology. “This study is much larger than any that had been done before. No one has yet been able to measure urban air quality at a street-by-street level.”

Considerable research on the effects of air quality is being conducted around the world. There are many national measurement networks to measure air quality, but these networks often have only a few monitoring stations in each city. In the Netherlands, for example, the national measurement network has ‘only’ 60 continuously operating stations. “That is enough to be able to say something about air quality in general, but not at a very fine resolution”, says IRAS researcher and co-author Roel Vermeulen. “So in our study, we often use calculation models that predict pollution levels based on detailed information about factors such as traffic. Unfortunately, we have not been able to measure the actual air quality in every street. With mobile platforms like Google Street View cars, we can measure exposure in much greater detail.”

Insight into sources of pollution

The project is the world’s first study of this scale to measure air quality over a large area. “It deals with an area of more than 30 square kilometres; absolutely the largest collection of data ever. This will give us much greater insight into the local differences due to different sources of pollution. It will also be an excellent example of what we can achieve with new technologies. For example, we may eventually be able to use these measurement sets on other vehicles, such as Post-NL or DHL delivery vans, or buses.”

A Google Street View car fitted with air pollution measurement equipment

Five times more accurate

Since the Google Street View cars are moving, the measurement results are four to five times more accurate spatially than normal stationary measurements in Oakland. In the Netherlands, the IRAS also has a measurement vehicle on the road, equipped with measurement instruments to monitor air quality. The measurements are similar, but the instruments on the Google Street View cars are more advanced due to the partnership with ACLIMA.


“So far, we’ve used our green IRAS car to monitor certain areas, after which we create mathematical models of the measurements. The difference with Google is that in Oakland we didn’t use a model; we actually conducted repeated measurements of every street. That allows us to measure air pollution in a city at a resolution that had not been possible before, which enables us to identify ‘hotspots’ in addition to the general patterns.”

Why is the study being conducted in Oakland?

“Oakland is close to San Francisco, and it has a Google location. Oakland is interesting because it has several motorways, an airport and a large seaport. But the technology we’re using could easily be used in the Netherlands as well. In fact, we’re already using it with our platforms. But it’s possible to do it on a much bigger scale, so that we could produce more detailed maps in the Netherlands as well.”

Linking to health data

The project in Oakland is still underway. The researchers are still drawing up mathematical models based on the measurements, and are considering how many measurements are necessary in order to draft good air quality maps. They also want to link the measurements to health data in Oakland. “It’s entirely conceivable that we might eventually do that in the Netherlands as well. At the moment, we are measuring ultra-fine particulates all over the country, and we will use the mobile platforms as part of the ‘Livestock Farming and Neighbours’ Health’ study (VGO).”

Read more about the Google project.