Ozone and Carbon Monoxide Measurements from Zurich's Mobile Tram Network

19.9 million ozone samples from 2012-2016 and 49.7 million carbon monoxide samples from 2014-2016 were collected by sensors mounted on 10 streetcars in Zurich, Switzerland. The dataset was created by researchers at ETH Zurich and includes GPS coordinates, timestamps, and sensor readings. Data quality requires cleaning and calibration, as the raw measurements have not been post-processed.

Use Cases

• Developing spatiotemporal air pollution models based on mobile sensor readings and GPS coordinates.
• Calibrating low-cost mobile sensors against reference data from official monitoring stations.
• Analyzing urban air quality patterns and exposure based on tram route data.
• Studying sensor drift and multi-hop calibration techniques for mobile sensor networks.

Strengths

• Large temporal scale with data collected over 4.5 years.
• High spatial granularity provided by GPS coordinates from 10 moving trams.
• Includes raw sensor voltages for carbon monoxide, enabling advanced calibration.

Limitations

• Data has not been post-processed and requires cleaning and sensor-specific calibration.
• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown for the full dataset, which may limit suitability assessment.

Provenance

Source

ETH Zurich OpenSense mobile sensor network.

Collection Method

Sensors mounted on top of 10 streetcars collected measurements.

Time Range

Ozone: 2012/02 - 2016/09; Carbon Monoxide: 2014/03 - 2016/09.

Geography

City of Zurich, Switzerland.