Urban fluxes for free: Estimating urban turbulent surface fluxes from crowdsourced meteorological canyon layer observations

Crowdsourcing and citizen science data have gained insight in the urban heat island effect and intra-urban heat patterns in many cities. However, while the urban energy balance is key in understanding the urban climate, professional urban surface flux measurements are relatively scarce. Here we develop a method to estimate urban fluxes of sensible heat, latent heat and momentum using solely crowdsourced temperature, humidity and wind speed observations in the urban canopy through Netatmo amateur weather stations. Also, the spatial variance of temperatures recorded in a network of Netatmo stations (varT) appears to be a good predictor for the incoming solar radiation. The proposed flux method is evaluated against eddy covariance flux estimates in Amsterdam (The Netherlands), and appears to have a median absolute error of 46.3 Wm − 2 and 22.8 Wm − 2 for sensible and latent heat flux respectively. When applying varT these values drop to 30.5 and 17.5 Wm − 2 respectively. These scores compare well with schemes driven by professional observations. Hence, we offer a meaningful flux scheme that runs purely on free observations. • We develop a method to estimate urban turbulent fluxes based on Netatmo weather observations in the urban canopy layer. • Incoming solar radiation can be estimated via the spatial variance of Netatmo based air temperatures within a city. • Flux estimates based on purely free data perform well for sensible heat and momentum fluxes, but performs less for latent heat flux.