EMS Annual Meeting Abstracts
Vol. 23, EMS2026-571, 2026, updated on 22 Jun 2026
https://doi.org/10.5194/ems2026-571
EMS Annual Meeting 2026
© Author(s) 2026. This work is distributed under
the Creative Commons Attribution 4.0 License.
Poster | Tuesday, 08 Sep, 16:30–18:00 (CEST), Display time Monday, 07 Sep, 08:00–Tuesday, 08 Sep, 18:00| TransitZone, P51
Exploring the urban effect on humidity using crowdsourced data
Emil Severiens1, Arjan Droste1, Sandra De Vries2, Marjolein Van Esch3, and Remko Uijlenhoet1
Emil Severiens et al.
  • 1TU Delft, Faculty of Civil Engineering and Geosciences, Water Management, Delft, Netherlands
  • 2PULSAQUA, Netherlands
  • 3TU Delft, Faculty of Architecture and the Built Environment, Urbanism, Delft, Netherlands

Improving our understanding of the behaviour of near-surface atmospheric properties will help us prepare for changing climatic conditions. Especially in cities, where human health risk is concentrated, and where surfaces present a radical departure from natural conditions. Research shows that air temperature is relatively high in cities as a consequence of the urban heat island effect, increasing the risk of heat stress. Humidity too affects human thermal perception and is altered by the urban environment. The urban effect on near-surface atmospheric humidity, however, is not nearly as well studied as the urban effect on air temperature.

Studies of the urban effect on humidity primarily compare measurements in proximate but contrasting urban and rural locations, aiming to establish the existence and behaviour of an urban moisture excess/deficit (also known as the urban moisture island effect). This has yet to be done with crowdsourced data. Unlike traditional stationary data collection methods, crowdsourced data is both temporally and spatially dense. Using quality controlled data from privately operated weather stations may therefore allow for a better representation of the heterogeneous urban environment than previously possible for studies of the urban effect on humidity.

From a network of personal weather stations managed by the DelftMeet citizen science project, we used 35 stations in an area of 17.5 km2 in Delft, the Netherlands. Each weather station measured inter alia air temperature, relative humidity, and air pressure approximately every 5 minutes. These measurements were aggregated to 15 minute averages and used to calculate vapour pressure values for assessing the urban moisture excess/deficit. The measurement campaign ran for a period of 24 months from 01/10/2023 to 30/09/2025. Excluding missing values this left approximately 2,379,282 vapour pressure values with which to conduct comparisons with rural reference data.  

Although the analysis is yet to be finalised, initial results show a slight moisture deficit in Delft compared to its surroundings. This deficit does seem to persist throughout the day and throughout the year, however. Such a stable but weak moisture deficit would certainly be uncommon.  

How to cite: Severiens, E., Droste, A., De Vries, S., Van Esch, M., and Uijlenhoet, R.: Exploring the urban effect on humidity using crowdsourced data, EMS Annual Meeting 2026, Utrecht, Netherlands, 6–11 Sep 2026, EMS2026-571, https://doi.org/10.5194/ems2026-571, 2026.