Quantification of Aviation's Impact on Air Quality near Schiphol Airport using Low-Cost Gas Sensors

Quantification of Aviation's Impact on Air Quality near Schiphol Airport using Low-Cost Gas Sensors

Maes, Jeff (TU Delft Aerospace Engineering)

Dedoussi, I.C. (mentor)

Delft University of Technology

Aerospace Engineering

2021-06-18

A lowcost multipollutant sensor box is developed, calibrated and deployed to quantify aviationattributable air quality impacts at Amsterdam Schiphol airport. The sensor box contains four lowcost electrochemical gas sensors that respond to CO, NO, NO2 and O3 concentrations in ambient air. Additional sensors in the sensor box allow the determination of pressure, temperature, relative humidity, sound intensity, UVlight, visible light and IRlight. The pressure, temperature and relative humidity sensors assist in calibrating the gas measurements obtained by the gas sensors. Sound intensity measurements and ADSB data assist in the detection of nearby aircraft activity. Calibration of the sensor box was done by collocating it for ∼74 hours with an air quality monitor of the Dutch air quality network in Wijk aan Zee, the Netherlands. Regulatory air quality data could be obtained at 10 s resolution for CO, NO and NO2. The absence of regulatory O3 measurements results in discarding the corresponding lowcost sensor in further analysis. Next, multivariate linear regression and random forest models were constructed to transform the raw sensor box measurements into meaningful concentrations. The random forest model has a strong tendency to overfit, even though it performs better in absolute numbers. The final calibration models for each species are multivariate linear regression models with an RMSE for CO, NO and NO2 of 118.13 μgm−3, 2.38 μgm−3 and 3.96 μgm−3, respectively. Corresponding correlation coefficients are 0.94, 0.99 and 0.98. Three different types of aircraft activity have been evaluated by deploying the sensor box at Amsterdam Schiphol airport. On 22 January 2021, the air quality impacts from landings on the Polderbaan were investigated by deploying the sensor box ∼200 m from the runway. On both 17 and 19 February 2021, the impact of aircraft departures on the Aalsmeerbaan and Kaagbaan were investigated by measuring ∼300 m from the runway. Finally, on 30 March 2021, air quality impacts from taxiing aircraft were obtained by taking measurements on the airport apron with aircraft coming as close as 43 m to the sensor box. The obtained sensor box measurements represent total air quality concentrations, which contain impacts from multiple sources. Quantifying the proportion of aviationattributable concentrations is achieved via a datadriven approach in which a baseline signal helps to extract the local aviationattributable signal. It was found that this strategy results in local signals that can be linked to activity from aircraft.

air quality
low-cost sensors
aircraft
air quality sensor box
Schiphol Amsterdam Airport
air pollution
alphasense
gas sensing
airport

http://resolver.tudelft.nl/uuid:e966bc05-1660-410c-8da6-86b4dba131d6

Student theses

master thesis

© 2021 Jeff Maes