Gert Jan Steeneveld
Part of
Keywords
Go to knowledge network
Crowdsourcing, gathering large amounts of data through the Internet, has been incredibly useful in urban meteorology. While it has been used to collect urban air temperature, air pressure, and precipitation data from sources like mobile phones or personal weather stations (PWSs), wind data have not been extensively researched. Urban wind behavior is complex and hard to measure accurately due to variations in location and equipment. Crowdsourcing can provide a dense network of wind observations and help understand the spatial patterns of urban wind.
In this study, we assess the accuracy of the popular "Netatmo" PWS anemometer compared to a reference instrument in both rural and urban settings. Then, we analyze wind speed data collected from 60 PWSs in Amsterdam, the Netherlands, to study wind speed distributions across different Local Climate Zones (LCZs).
We found that the Netatmo PWS anemometer consistently underestimates wind speed, and its accuracy is affected by rain or high humidity. To address these issues, we developed a quality assurance (QA) protocol to correct PWS measurements. With this protocol, PWS data significantly improved, allowing us to accurately estimate the probability density distribution of wind speed for a city or neighborhood.
We discovered that urban wind speed distributions are best described by a combination of two Weibull distributions, rather than the single Weibull distribution typically used for rural wind speed observations. However, the Netatmo PWS anemometer struggles to measure near-zero wind speeds, affecting the QA protocol's performance during periods of very low wind. Despite this limitation, the QA protocol provides satisfactory results for year-long wind speed climatology and shorter periods with higher wind speeds.
Droste, A. M., Heusinkveld, B. G., Fenner, D., & Steeneveld, G. J. (2020). Assessing the potential and application of crowdsourced urban wind data. Quarterly Journal of the Royal Meteorological Society, 146(731), 2671-2688.
Image credits
Icon image: Windmolens - Pixabay.org