Unequal distributions of crowdsourced weather data in England and Wales

Abstract

Personal weather stations (PWS) can provide useful data on urban climates by densifying the number of weather measurements across major cities. They do so at a lower cost than official weather stations by national meteorological services. Despite the increasing use of PWS data, little attention has yet been paid to the underlying socio-economic and environmental inequalities in PWS coverage. Using social deprivation, demographic, and environmental indicators in England and Wales, we characterize existing inequalities in the current coverage of PWS. We find that there are fewer PWS in more deprived areas which also observe higher proportions of ethnic minorities, lower vegetation coverage, higher building height and building surface fraction, and lower proportions of inhabitants under 65 years old. This implies that data on urban climate may be of lower quality or more uncertain in particular areas, which may limit the potential for climate adaptation and empowerment in those communities.

Spatially Explicit Correction of Simulated Urban Air Temperatures Using Crowdsourced Data

“…However, reliable tools have now been developed since the first use of PWSs for model evaluation by Hammerberg et al (2018) to filter dubious measurements out [e.g., CrowdQC from Napoly et al (2018) or CrowdQC+ by Fenner et al (2021) ], thus making PWS observations increasingly reliable. This does not resolve the question of the representativity of measurements, that is, “how is one PWS measurement representative of the simulated urban pixel?” Yet, the increasing density of PWSs in the urban environments begins to alleviate this uncertainty—despite a recognized unequal distribution of PWSs among a variety of environmental, socioeconomic, and demographic indicators ( Brousse et al 2023 ). For example, Venter et al (2020) found that a density of one PWS per square kilometer is optimal for predicting seasonal air temperature in Oslo.…”