Can crowdsourcing increase the durability of an urban meteorological network?

Dense networks of weather stations in urban areas are now becoming an increasingly common, but expensive way of monitoring the urban climate. The expense is not only related to the initial deployment of the network, but also the ongoing maintenance which can be increasingly problematic if monitoring is needed for extended periods of time. As an alternative, crowdsourcing weather data provides an effectively free option to acquire hyperlocal weather data, but is not yet fully accepted by the user community due to data quality concerns. This paper explores an approach between these two options to assess whether a more durable (i.e. longer term, less transient) means of monitoring can be achieved by using a controlled deployment of ‘low-cost’ Netatmo weather station maintained by citizens. As a result, the novelty of this paper resides in the first assessment of the durability of crowdsourcing as opposed to just the data quality alone. The results show that expert input at the deployment stage in terms of device modification and siting improves data quality, but the length and completeness of the record of unsupported weather stations is highly variable with less than 50% of devices surviving the measurement campaign, and far fewer providing full data records. Better durability, albeit at the expense of data quality, was exhibited by pre-existing Netatmo devices procured by the public in the study area with many units producing useable data throughout and beyond the measurement campaign. Overall, these are significant findings which need to be taken into account at the design stage of an urban meteorological network and prompt a rethink in the way that equipment is deployed via living lab approaches. Fundamentally, existing approaches are all largely built on a paradigm of transient sensing methodologies which are difficult to reconcile with a science that uses thirty-year baselines.